Seznam příloh. Příloha č. 9:

PŘÍLOHY

Seznam příloh Příloha č. 1: Řezníčková, L., Brázdil, R., Valášek, H. (2005): Early instrumental meteorological observations in the Czech Lands. In: From the Holocene to the Anthropocene: Climate of the Last 1000 Years. Swiss National Science Foundation - NCCR, Bern, nestr. Příloha č. 2: Štěpánek, P., Řezníčková, L., Brázdil, R. (2006): Homogenization of daily air pressure and temperature series for Brno (Czech Republic) in the period 1848–2005. WCDMP, WMO, Geneva, v tisku. Příloha č. 3: Řezníčková, L., Štěpánek, P., Brázdil, R. (2007): Homogenizace řad denních hodnot tlaku a teploty vzduchu v Brně v období 1848–2005. In: Česká geografie v evropském prostoru. XXI. sjezd České geografické společnosti. Jihočeská univerzita, České Budějovice, s. 1258–1263. ISBN 978-80-7040-986-2. Příloha č. 4: Řezníčková, L., Brázdil, R., Tolasz, R. (2006): Meteorological singularities in the Czech Republic in the period 1961–2002. Theoretical and Applied Climatology, 88, s. 179–192. DOI 10.1007/s00704-006-0253-5. Příloha č. 5: Brázdil, R., Krušinský R., Řezníčková L. (2008): Zprávy o počasí z let 1655–1656 v deníku Jana Františka Bruntálského z Vrbna. Časopis Matice moravské, v tisku. Příloha č. 6: Brázdil R., Černušák T., Řezníčková, L. (2008): Weather information in the diaries of the Premonstratensian abbey, Hradisko (Olomouc, Czech Republic), 1693–1783. Weather, v tisku. DOI: 10.1002/wea.264. Příloha č. 7: Brázdil, R., Řezníčková, L., Valášek, H. (2007): Počasí v Čechách v letech 1805– 1806: konfrontace vizuálních a přístrojových pozorování. Meteorologické zprávy, 60, 6, s. 187–193. Příloha č. 8: Brázdil, R., Valášek, H., Řezníčková, L., Štěpánek, P. (2008): Měření srážek v Těšíně v období leden 1777–leden 1778. Meteorologické zprávy, 61, 1, s. 26–29. Příloha č. 9: Brázdil, R., Řezníčková, L., Valášek, H. (2006): Early instrumental meteorological observations in the Czech Lands I: Ferdinand Knittelmayer, Brno, 1799–1812. Meteorologický časopis, 9, 2, s. 59–71.

Příloha č. 10: Brázdil, R., Řezníčková, L., Valášek, H. (2007): Early instrumental meteorological observations in the Czech Lands II: Andreas Sterly, Jihlava, 1816–1840 (1844). Meteorologický časopis, 10, 1, s. 3–12. Příloha č. 11: Brázdil, R., Řezníčková, L., Valášek, H., Kotyza, O. (2007): Early instrumental meteorological observations in the Czech Lands III: František Jindřich Jakub Kreybich, Žitenice, 1787–1829. Meteorologický časopis, 10, 2, s. 63–74.

Příloha č. 1

Early instrumental meteorological observations in the Czech Lands Ladislava Řezníčkováa, Rudolf Brázdila, Hubert Valášekb a

Institute of Geography, Masaryk University Brno, Czech Republic b Moravian Land Archives, Brno, Czech Republic

Early instrumental meteorological observations are the most objective source of information about the state of the atmosphere before the period of systematic meteorological records which started in the Czech Lands during the second half of the 19th century being a part of the network of Central Institute of Meteorology and Geodynamics (CIMG) in Vienna. Early instrumental records can be used for extension of existing climatological series for some places in the Czech Lands (mainly before 1850). Elaboration of the early instrumental measurements have to take into consideration some disadvantages related to this information. Problems concern relatively difficult approach to data (located in district or regional archives), short duration of observation (from some months to some years), its reading (hand-written, e.g. in Gothic script), often missing information about detail location of measurements and type of instruments used, non-standard terms of observations, some missing values, different terminology and definition of meteorological phenomena, etc. In Brno, the systematic meteorological measurements published by the CIMG started since 1848. But these measurements can be extended by many other records. Systematic instrumental observations from May 1799 to September 1812 deposited in the City Archives of Brno belong to some of them (2). A pensioned captain Ferdinand Knittelmayer (1750–1814) was their author. He observed five times per day air pressure, air temperature, wind direction and force, cloudiness, variation and property of clouds and meteorological phenomena such as rainfall, snow, fog, gale, thunderstorm etc. But no term data were preserved from his observations, only summary daily tables elaborated with respect to the Moon phases (see Figure 1). He believed that the same weather would return after the 19-year Moon cycle completion, which was, in fact, the main motivation to keep his records (1). Fig. 1. A specimen of summary records of Knittelmayer’s observation of air temperature.

Knittelmayer’s weather records have recently been digitised and logically tested. Further the missing values have been completed and interpreted in the terms of recent units and terminology. From corrected and completed observations, mothly averages and frequencies have been calculated (Figure 2). The Alexandersson standard normal homogeneity test with the reference stations Prague-Klementinum and Vienna-Hohe Warte have been used for relative homogeneity testing with a subsequent data homogenisation. Obtained results will complete existing gaps in the Brno meteorological data before 1848. Future research is supposed to be oriented not only on further known early instrumental records from Brno, but also on observations from other places in the Czech Republic (e.g. Děčín, Jihlava, Opava, Žitenice). Instrumental data will partly be combined also with systematic visual daily weather records as a special type of documentary evidence. Database of early instrumental and visual weather observations will be used for analysis of spatial and temporal climatic patterns in the Czech Lands before 1850. 1010 1005

P res s ure [hP a]

1000 995 990 985 980 975 970 1799

1800

1801

1802

1803

1804

1805

1806

1807

1808

1809

1810

1811

1812

1800

1801

1802

1803

1804

1805

1806

1807

1808

1809

1810

1811

1812

30

T em perat ure [ °C]

25 20 15 10 5 0 -5 -10 1799

1

2

3

Figure 2. Comparison of the mean monthly air pressure and air temperature in Brno (1) with the measurement at Prague-Klementinum (2) and Vienna-Hohe Warte (3) during the period May 1799–August 1812. Acknowledgement: The authors would like thank the Grant Agency of the Czech Republic for financial support giving to the grant project No. 205/05/0858. References: (1) Brázdil, R., Valášek, H., Macková, J. (2005): Meteorological Observations in Brno During the First Half of the Nineteenth Century (History of Weather and Hydrometeorological Extremes) (in Czech). Archiv města Brna, Brno, in press. (2) Meteorologische Beobachtungen in Brünn 1799–1812. Archiv města Brna, fond IV A 6. Meteorologická a přírodovědná pozorování, rkp. čís. 7263.

Příloha č. 2

HOMOGENIZATION OF DAILY AIR PRESSURE AND TEMPERATURE SERIES FOR BRNO (CZECH REPUBLIC) IN THE PERIOD 1848–2005

P. Štěpánek1, L. Řezníčková2, R. Brázdil2 1

Czech Hydrometeorological Institute, Regional Office Brno, Czech Republic Phone: +420-541 421 033, E-mail: [email protected] 2 Institute of Geography, Masaryk University, Brno, Czech Republic

Abstract Homogenization of daily meteorological series is a difficult task. Several kinds of problem have to be taken into consideration in the course of homogenization: selection of a proper homogenization method with regard to the data used, creation of reference series, completion of missing values, annual course of adjustments, and others. This paper presents an attempt to create a homogeneous series of daily air pressure and temperature readings in the city of Brno (Czech Republic). Two basic approaches were adopted: (i) homogenization of monthly series and projection of estimated smoothed monthly adjustments in annual variation of daily adjustments and (ii) homogenization of daily values in individual months and direct estimation of daily adjustments, again smoothed by low-pass filter. Differences in the results obtained from these two approaches are further discussed.

1. INTRODUCTION

In the recent years considerably more attention has been devoted to the analysis of the daily data widely recorded and stored in databases. Prior to analysis, the need to homogenize the data and check their quality arises. There is no widely accepted homogenization approach that could be generalized and applied to various meteorological elements, different climatic patterns, etc., and this will probably never be possible. This is, for example, due to the fact that the statistical properties of daily data and regional differences between them make general homogenization of daily values difficult, as well as involving more demanding data handling. During data processing, several kinds of problem have to be taken into consideration. These involve selection of a proper method for homogenization with regard to the data used, i.e. fulfilling all the conditions necessary to applying selected tests of relative homogeneity (e.g. normal distribution), creation of reference series (defining selection criteria), completion of missing values, annual course of adjustments, and others. Only a few studies, in comparison with monthly or annual data series, have been devoted to techniques addressing daily values. For example, Brandsma (2000) compared monthly adjustments, daily adjustments derived from monthly adjustments (using iterative cubic spline interpolation to preserve monthly adjustments) and daily adjustments derived from weather types. Wijngaard et al. (2003) did not use measured values, but their characteristics, such as diurnal temperature range and its annual mean, as well as the annual mean of the absolute day-to-day differences for temperature, and the annual number of wet days for precipitation. Following various homogeneity tests, these series were labelled as recommendations for further analysis.

Mekis and Vincent (2004) derived daily adjustments from monthly adjustments. These were obtained using linear interpolation between mid-month “target” values objectively chosen so that the average of the daily adjustments over a given month is equal to the monthly adjustment. This approach does not require the creation of a daily reference series or the identification of inhomogeneities in daily temperatures. Moreover, finally homogenized series of daily temperatures are compatible with homogenized monthly data-sets. The present paper is dedicated to the search for a proper methodology for daily data-set handling and its subsequent application to daily air pressure and temperature series for Brno in the period 1848–2005, with the aim of creating a homogeneous series for Brno with regard to both elements. Although there are, in general, no gaps in the Brno measurements, data are unfortunately not available from a single site, so it becomes necessary to combine different series to get one Brno series suitable for further analysis. The basic Brno stations were tested separately for relative homogeneity and, after homogenization, they were combined using overlap periods. All calculation was performed using AnClim and ProClimDB software (Štěpánek, 2006a, 2006b).

2. A BRIEF HISTORY OF METEOROLOGICAL OBSERVATIONS IN BRNO

Meteorological observations in Brno began in 1799, the work of Captain Emeritus Ferdinand Knittelmayer, but his observations for the period 1799–1812 are preserved only in the daily averages. For the subsequent years 1813–1819, the observations exist only in the form of monthly averages. On the basis of several daily readings, meteorological observations were published in the daily newspaper “Mährisch-Ständische Brünner Zeitung” from January 1820 to December 1847. In some years, parallel observations from two places in Brno were also made. Although monthly value series for air pressure, air temperature and precipitation totals have been homogenised and analysed (Brázdil et al., 2005), work with daily readings or daily averages requires further research. For this reason, the analysis provided in this paper works only with data from 1848 onwards. Meteorological observations after 1848 come from Dr. Paul Olexik (1800–1878), a physician from St. Anne’s hospital (Fig. 1). He was probably making observations from as early as the end of 1845, but it was only from 1848 that his measurements started to be published regularly in the Austrian Meteorological Yearbooks, i.e. when his station became part of the network of the Central Meteorological Institute in Vienna. He observed at 0600, 1400 and 2200 hours. On 3 December 1853 he moved the point of his meteorological observations from the hospital (204 m.a.s.l.) a short distance, to his new flat at Pekařská Street 100 (219 m.a.s.l.). Meteorological observations at this new site continued until 30 June 1878. By this time, Gregor Johan Mendel (1822–1884), abbot of the Augustinian Monastery and a pioneer geneticist, was helping to complement Olexik’s measurements, something he continued alone from 1 July 1878 in the monastery garden (204 m a.s.l.) until 30 November 1883. He began with standard readings at 0700, 1400 and 2100 hours. Alfred Lorenz (1825– 1890), a professor at the I. R. Technical University, continued meteorological observations in Brno from the university building (225 m a.s.l.), located close to the city centre, from 1 January 1884 until his death in June 1890. Upon his death, air temperature and pressure measurements definitely stopped and no new place for observation was to be found (Brázdil, 1979).

Fig. 1. Location of meteorological stations in Brno: 1 – St. Anne’s hospital; 2 – Pekařská Street 100; 3 – Augustinian monastery; 4 – I. R. Technical University; 5 – Pisárky, waterworks; 6 – Květná Street; 7 – Tuřany Airport

However, from 1 June 1890 a meteorological station at the city waterworks in Pisárky (204 m a.s.l.) (further as Brno-Pisárky) began operations, with a full observation programme up to 1937 and with air temperature measurements continuing up to 1962. Further meteorological stations in different parts of Brno were established later, of which only the two used in this paper are mentioned. The first of them was located close to the previous station on Květná Street (further as Brno-Květná), in the garden of the research agricultural institute (223 m.a.s.l.), with observations from 1 August 1922 to 31 December 1970. The second station (Brno-Tuřany) is located at the Brno airport, south-east of the city of Brno (238 m.a.s.l.), i.e. opposite to all the previously mentioned stations, which are concentrated in its western part. Observations started there on 14 April 1958. For this reason, compilation of the Brno daily temperature and pressure series is made with respect to this station. In summary, addressing knowledge of the history of temperature and pressure measurements in Brno from 1848 onwards, with respect to homogenization, it should be stressed that measurements - were provided from different parts of Brno, at different altitudes - were provided by different types of instruments - were provided in different observation terms before and after 1878 - are limited by lack of available overlap for observations predating 1890.

3. DATA USED

For outlier identification as well as for relative homogeneity testing, other stations with long-term series in the broad surroundings of Brno were also used (Fig. 2). A list of them, with basic characteristics, is given in Table 1. We have used all the measurements, i.e. not only daily averages but also separate series from individual observation hours. As has already

been mentioned, as well as standard observations times at 0700, 1400 and 2100 hours local mean time, observations were also carried out at 0600, 1300 and 2200 hours. Finally it was decided, that all the terms should further be treated as if they were 0700, 1400 and 2100 hours in the hope of disclosing possible inhomogeneities arising out of various observing times during homogeneity testing. The original observing hours were taken into consideration during decision-making about adjustments of inhomogeneities found.

Fig. 2. Geographical distribution of stations used for homogenization of the Brno series (T – air temperature, P – air pressure)

Table 1. Basic information about stations used for homogenization of the Brno series (station coordinates are given for their last or recent locations) Air temperature Station name

Latitude

Longitude

Altitude

(N)

(E)

(m.a.s.l.)

Beginning

End

Brno (various places) Brno-Pisárky Brno-Květná Brno-Tuřany Bystřice pod Hostýnem Český Těšín Holešov Jihlava Olomouc Prague-Klementinum Přerov

49º12´ 49°12´ 49°12´ 49°09´ 49°24´ 49°44´ 49°19´ 49°23´ 49°36´ 50°05´ 49°25´

16º37´ 16°34´ 16°34´ 16°ˇ42´ 17°40´ 18°37´ 17°34´ 15°32´ 17°15´ 14°25´ 17°24´

Vienna-Hohe Warte

48°13´

16°21´

Latitude (N)

Brno (various places) Brno-Pisárky Brno-Květná Brno-Tuřany Holešov Prague-Klementinum Vienna-Hohe Warte

Observing hours

225 203 223 241 315 280 224 560 215 191 203

1 Jan. 1848 1 June 1890 1 Aug. 1922 14 Apr. 1958 1 Sep. 1865 1 Jan. 1885 1 July 1895 27 July 1873 1 Jan. 1876 1 Jan. 1775 1 Apr. 1874

31 Dec. 1889 31 May 1962 31 Mar. 1970 31 Dec. 2005 31 Dec. 2005 31 Oct. 1938 31 Dec. 2005 31 Dec. 1934 31 Dec. 1960 31 Dec. 2005 31 Dec. 1979

07 (06), 14 (13), 21 (22) 07, 14, 21 07, 14, 21 07, 14, 21 07 (06), 14, 21 (22) 07, 14, 21 07, 14, 21 (22) 07 (08), 14, 21 (22) 07 (08), 14, 21 (20) 07, 14, 21 07, 14, 21

199

1 Jan. 1872

31 Dec. 2005

07, 14, 19

Longitude

Altitude

Beginning

End

(E)

(m.a.s.l.)

49º12´ 49°12´ 49°12´ 49°09´ 49°19´ 50°05´

16º37´ 16°34´ 16°34´ 16°42´ 17°34´ 14°25´

225 203 223 241 224 191

1 Jan. 1848 1 June 1890 1 Aug. 1922 14 Apr. 1958 1 Jan. 1961 1 Aug. 1787

31 Dec. 1889 31 Dec. 1937 31 Dec. 1962 31 Dec. 2005 31 Dec. 2005 31 Jan. 2002

07 (06), 14 (13), 21 (22) 07, 14, 21 07, 14, 21 07, 14, 21 07, 14, 21 14

48°13´

16°21´

199

1 Jan. 1872

31 Dec. 2005

07, 14, 19

Air pressure Station name

Observing hours

The correlation coefficients for both elements analyzed are high enough for all stations involved (Fig. 3). Their values were calculated from original data (not from series of first differences), so they are biased by inhomogeneities in a shift (the values would otherwise be higher) and also by trends (the values would be lower if the trend were removed from the series). Correlations of monthly averages are higher than those of daily averages during the winter months, while the opposite holds in summer, i.e. the correlations of monthly averages drop below the values of daily data. From this it follows that both monthly and daily data should be used for data homogenization; daily data are more sensitive to inhomogeneity detection, especially during the summer months.

Air temperature, daily

Air pressure, daily

1.00

1.00 0.95

0700

1400

0.95

2100

0.90

0.90

0.85

0.85

0.80

0.80

0.75

0.75

0.70

0.70 J

F

M

A

M

J

J

A

S

O

N

D

Air temperature, monthly

1.00

J

F

M

0.95

0.90

0.90

0.85

0.85

0.80

0.80

0.75

0.75

0.70

M

J

J

A

S

O

N

D

O

N

D

Air pressure, monthly

1.00

0.95

A

0.70 J

F

M

A

M

J

J

A

S

O

N

D

J

F

M

A

M

J

J

A

S

Fig. 3. Medians of correlation coefficients for all pairs of stations, for daily and monthly air temperature (50 values) and air pressure (6 values – without Prague-Klementinum at 1400 hours)

4. HOMOGENIZATION

Homogenization includes the following steps: detection, verification and possible correction of outliers (extreme values), creation of reference series, homogeneity testing (various homogeneity tests), determination of inhomogeneities in the light of test results and 1.000 metadata, adjustment of inhomogeneities and filling in missing values (Fig. 4). Air pressure, monthly 0.950

0.900

P_07

P_14

P_21

0.850 I

II

III

IV

V

VI

VII

VIII

IX

X

XI

XII

Fig. 4. Plan of the homogenization process

4.1 Outlier identification Data quality control was carried out in two ways in this study: (i) by applying limits derived from interquartile ranges (either to individual series, i.e. absolutely or, better, to difference series between candidate and reference series, i.e. relatively), (ii) by comparing candidate station values to values from neighbour stations. In comparisons with neighbour stations, the five best correlated series were selected (correlations calculated from series of first differences – see e.g. Peterson, 1998), the values of correlation coefficients being at least 0.50; no limit for distance or altitude difference has been applied. Only series with the same observation hours were selected. For the evaluation of outliers, various characteristics were considered. A count of statistically significant different neighbours (compared to candidate station) exceeding the confidence limit (0.95) was evaluated by means of difference series (neighbour minus candidate station), for each month individually. Cases in which more than 75% of neighbours differed significantly from the base station values were checked visually. To help in establishing the nature of the outliers, the values of neighbours were standardized with respect to candidate station average and standard deviation and a new (theoretical) value for the candidate station was also calculated – as a weighted average from the standardized values of the neighbours. Further, the coefficient of interquartile ranges (q75–q25) above q75 (or below q25) were evaluated (calculated from the standardized neighbour values), and applied to candidate station value. The reason for this was to assess similarity of neighbour values used with regard to test value: the more values of neighbours are similar, the higher is the value of the coefficient. The final decision on removing outliers was based on the percentage of the count of significantly different neighbours, difference from “expected value”, coefficient of

interquartile range, and finally by visual (subjective) comparison of the standardized values of neighbours with the candidate station values. Fig. 5 shows an example of the output for decision-making about outliers.

Fig. 5. Example of output with auxiliary characteristics for quality control evaluation

In some cases, in which at least two neighbours were not available, interquartile ranges for each individual month of the candidate series were applied (i.e. absolutely) and the errors emerging were checked. This method has considerably inferior results in comparison with the relative method, but no other possibility existed for cases in the distant past. 4.2 Homogeneity test As well as monthly, seasonal and annual averages, series of daily data were also tested. In this case we used all days of a particular month and further an aggregation of “seasons and year” calculated from the first days of all months, the second days, etc. (see Fig. 6). Although such “aggregate” series cannot be used for common time series analysis because the time is “cracked”, it can be very useful for the purposes of finding discontinuity (seasonal to annual resolution), while original daily values, even when used only within particular months, can suffer from annual course (this is the case for air temperature rather than air pressure, mainly in winter) and normality is sometimes on the border of the 0.05 significance level. Using the aggregates over seasons and year leads to series for which normality is fulfilled without problems, and thanks to lower signal-to-noise ratio this approach is better for detecting real inhomogeneities in the series. Significant autocorrelations within a number of first lags (days) appear to present a larger problem and have to be further investigated. Series are more persistent in winter with stronger circulation effects, rather than in summer with its prevailing radiation factors.

Fig. 6. An example of using daily data for homogeneity testing

Several relative homogeneity tests (significance level 0.05) were used: the Alexandersson Standard Normal Homogeneity Test SNHT (Alexandersson, 1986, 1995), the Maronna and Yohai bivariate test (Potter, 1981), the Pettit test (Pettit, 1979), the t-test (Mitchell et al., 1966) and the Easterling and Peterson test (Easterling, Peterson, 1995). Tests were applied to

40-year sections of the series tested for monthly averages and 30-years series of daily data because the alternative hypothesis of the Alexandersson and bivariate tests assumes the presence of only one inhomogeneity in a series (we applied SNHT for a single shift). Series longer than 40 years were divided into several parts with an overlap of ten years (or five years for daily data). This is important in the light of tendencies to overestimation of detected inhomogeneities near the ends of series (see Alexandersson, 1995). Reference series were created separately with respect to each 40-year (30-year) parts of a candidate series (this means with its own selection of neighbours in each part). For daily data, 185 sections of series (of 49 original elements-terms-stations) were created and tested. The use of series with durations of 40 and/or 30 years seems to be reasonable for homogeneity testing. Shorter series would not be so suitable from a statistical point of view, while, on the other hand, longer series usually contain more than one inhomogeneity (the typical duration of a period with one inhomogeneity does not usually exceed 30–40 years – see e.g. Auer et al., 2001). To ensure that only one inhomogeneity detected by the Alexandersson or bivariate tests was present in a series, a further modification was introduced into the AnClim software. The series was divided at the position of a detected inhomogeneity and sections before and after it were tested separately. If no other inhomogeneity was found, we can rely on the results of the given test for the whole length of the series (especially the significance of a test statistic). 4.3 Reference series creation Reference series were created in two ways: (i) an average from the best correlated stations, (ii) an average from nearest stations. Correlation coefficients used for station selection were calculated from the series of first differences, when inhomogeneities are manifested in the only value (see e.g. Alexandersson, Moberg, 1996; Peterson, 1998). Various types of reference series with analysis of their advantages and drawbacks have been discussed, for example, by Štěpánek (2005). The values of correlation coefficients were not allowed to drop below 0.60 between neighbour stations (selection by means of correlation) and no distance or altitude limits were applied as additional conditions for air pressure and temperature. Weighted averages were calculated using correlations and/or reciprocal values of station distances as weights. Values of selected neighbour stations were standardized to candidate station average and standard deviation to avoid problems with biased reference series. This can often happen in the event of missing data in one of the neighbour series. The standardization was done for each particular month individually (also for daily data). No transformation of values has been applied to the data. In the first stage, a list of proposed neighbour stations was obtained, which was subsequently checked and its approved version was then finally used for the reference series calculation. 4.4 Assessment of detected inhomogeneities The main criterion for determining a year of inhomogeneity was the probability of the given inhomogeneity, i.e. the ratio between the count of detections for a given year from all tests for the given station (using all types of reference series, tests, daily, monthly, seasonal and annual series) and the count of all theoretically possible detections. The count of detections for groups of years was also taken into account (some inhomogeneities started in

the course of the year and thus were manifested in at least two years). If metadata did not confirm the detected shift (in most cases), the percentage limit of all possible detections was taken higher and some other information (e.g. distribution of the given year within individual months or seasons, graphs of differences with reference series and some other characteristics) was required to decide whether the undocumented inhomogeneity could be regarded as “indubitably” proven and consequently corrected. For assessment of the inhomogeneities detected, the Real Precision Index (RPI – see Petrovic, 2004) may also be applied to find sections of series that exhibit change in the quality of measurements. 4.5 Adjustment of inhomogeneities Adjustment of inhomogeneities detected was addressed by means of the reference series calculated from the average over the five stations with the highest correlation coefficients with the series being adjusted (correlations were calculated again from the series of first differences). The adjustment value was estimated as the difference between averages calculated from difference series between the tested and the reference series. The start of inhomogeneity was allocated to a particular month (where this was possible). When dealing with daily data, there are several approaches to adjusting data for inhomogeneities detected. We may use either monthly adjustments which can be distributed into individual days (e.g., Mekis and Vincent, 2004) or we can calculate adjustments for daily data directly. Using monthly data in this paper, the estimated individual monthly adjustments were smoothed by low-pass filter (weights applied to adjacent months were approximately 1, 2, 1) to suppress the influence of random errors in the series (the effect of smoothing results in a more realistic annual course for the adjustments, in line with what is better physically justified). The monthly adjustments obtained were then distributed (interpolated) among individual days and the final daily adjustments (again possibly smoothed to eliminate the edges of lines occurring each month) were then applied to data. In the second case, the daily adjustments difference series (reference and tested) for each day of the year were used, taking 20 years before and after the change. Final daily adjustments were then smoothed using a low-pass filter for 60 days. Various characteristics were analyzed before applying the adjustments: increment of correlation coefficients between candidate and reference series after adjustments, change of standard deviation in differences before and after the change, presence of linear trend, etc. In the event of any doubts, the adjustments were not applied. 4.6 Further considerations The above-mentioned steps were performed in several iterations. At each iteration, more precise results were obtained. Missing values were filled in only after homogenization and adjustment of inhomogeneities in the series. The reason for this was that the new values were estimated from data not influenced by possible shifts in the series. Moreover, when missing data are filled in before homogenization, they may influence inhomogeneity detection in a negative way. The gaps were filled by means of linear regression between filled value series (dependent variable) and a reference series (independent variable), separately for each month. For assessing the quality of the process, various statistics were monitored, e.g. differences of averages and standard deviations in periods before and after the gap.

5. HOMOGENIZATION RESULTS

As has been shown above, the values of correlation coefficients for daily data (using each month individually) are comparable with values gained from monthly averages. The same holds true of correlations between tested and reference series. The medians of correlation coefficients for monthly air temperature range from 0.87 in the summer months to 0.98 in the winter months for individual observation hours; again the results at 1400 hours correlate the best. For daily data, the correlations for individual months range from 0.87 to 0.95. For air pressure, daily data correlates between 0.97 in summer and 0.99 in winter, monthly data between 0.94 and 0.99. From these results, it follows that it is worth working with daily data in the course of homogenization, even if it is more demanding compared to “simple” monthly averages. By employing daily data we have longer series (28 to 31 times, depending on number of days in a particular month) and we can then also better detect shifts near the end of the series (not resolvable for monthly averages with breaks of less then five years to the end of series). Fig. 7 gives count of inhomogeneities detected for daily and monthly series by the Alexandersson test with reference series created by means of correlations. 100

100

Air pressure, daily

Air temperature, daily

80

80

%

0700

1400

2100

60

60

40

40

20

20

0

0 J

F

M

A

M

J

J

A

S

O

N

D

J

100

F

M

A

M

J

J

A

S

O

N

D

J

J

A

S

O

N

D

100 Air pressure, monthly

Air temperature, monthly 80

60

60

40

40

20

20

%

80

0

0 J

F

M

A

M

J

J

A

S

O

N

D

J

F

M

A

M

Fig. 7. Percentage of inhomogenities in air pressure and temperature series for daily data and monthly averages, detected by the Alexandersson test, related to the total number of series used

The annual course of numbers of inhomogeneities is evident from the figure, as are the differences between air pressure and temperature readings, as well as observation hours. The large difference between monthly- and daily-based detections is, among other things, due to the fact that in the Alexandersson test the series is divided into sections in the position of each detected break. Since the series contain more members, we are able to detect relatively more inhomogeneities (mainly in the shorter sections). In this sense, the numbers between daily and monthly series are not comparable. But the aim was to show that during homogenization we

Temperature (°C)

should try to use information that is as dense as possible, using daily data, individual observation hours, etc. The advantages of using daily data mentioned above are apparent from the example in Fig. 8. In the event of missing values and breaks near the ends of series it is more difficult to detect inhomogeneities in the series if one works with only monthly data.

Fig. 8. Differences between tested and reference series for daily (left) and monthly (right) data for Brno, air temperature at 0700 (0600) hours, July 1873–1902

6. HOMOGENIZED PRESSURE AND TEMPERATURE SERIES OF BRNO

The creation of homogenized air pressure and temperature series for Brno covering the period 1848–2005 consists of several steps. First, the individual series for the different Brno stations (Brno stations before 1890 & Brno-Pisárky, Brno-Května, Brno-Tuřany – see Chapter 2) were homogenized according to the methodology described in Chapter 4. In the second step, a common compiled Brno series was developed by adjusting the individual parts. Starting from the recent observing station at Brno-Tuřany (reference station, 1958–2005), Brno-Květná data were adjusted to its measurements to obtain a series for the period 1923– 2005. In the next step, the Brno-Pisárky station was adjusted to the combined Květná-Tuřany reference series to obtain a series for the period 1890–2005 (Fig. 9). This approach was applied separately for each observation time (0700, 1400 and 2100 hours).

Air temperature 06,14,22

07,14,21

term

06,13,21

Brno (various places)

period 07, 14, 21

Brno – Pisárky

07, 14, 21

Brno – Květná

07, 14, 21

Brno – Tuřany

Air pressure 06,14,22

07,14,21

term

06,13,21

Brno (various places) Brno – Pisárky Brno – Květná Brno – Tuřany

period 07, 14, 21 07, 14, 21 07, 14, 21

Fig. 9. Scheme of creation for the series compiled by combining measurements from several locations in Brno

The principle of combination used for the individual Brno stations is identical to that employed for adjustments of inhomogeneities applied to data in the course of homogenization. Two approaches may be selected: one that uses monthly averages or one that works directly from daily data. The only difference is that final offsets are not computed by comparing periods before and after the change; in this case we use the whole period in common (shortened to 20 years if it is longer). The overlap periods vary from 5 years (air pressure) or 13 years (air temperature) in the first round to 15 years (air pressure) or 20 years (air temperature) in the second round. Fig. 10 gives an example of when final adjustment is obtained either from monthly averages or through direct use of daily data. It seems appropriate to calculate adjustments from daily values using a low-pass filter for 60 days, or, leading to the same results, using a low-pass filter for two months and subsequently distributing the smoothed monthly adjustments into daily values. 1.0

1.0

a)

b) 0.5

0.0

0.0 °C

°C

0.5

-0.5

-0.5

UnSmoothed 1 3 ADJ_C_INC ADJ_ORIG 2

01-Dec

01-Nov

01-Oct

01-Sep

01-Aug

01-Jul

01-Jun

01-May

01-Apr

01-Mar

01-Feb

-1.5

01-Dec

01-Nov

01-Oct

01-Sep

01-Aug

01-Jul

01-Jun

01-May

01-Apr

01-Mar

01-Feb

-1.5 01-Jan

ADJ_ORIG 4 ADJ_SMOOTH30 5 ADJ_SMOOTH60 6 ADJ_SMOOTH90 7

-1.0

01-Jan

-1.0

Fig. 10. Annual variations of adjustments applied to air temperature series at 1400 hours for Brno-Květná to the reference station Brno–Tuřany: a) monthly-based approach (1 – raw adjustments, 2 – smoothed adjustments, 3 – smoothed adjustments distributed into individual days), b) daily-based approach (4 – individual calendar day adjustments, 5 – daily adjustments smoothed by low-pass filter for 30 days, 6 – for 60 days, 7 – for 90 days)

The values measured at different observation hours exhibited quite different annual variations of adjustment, making it useful to work with them directly, and not just with calculated daily averages. For example, depending on the formula used for the calculation of daily averages, real inhomogeneities may be masked there. A fully compiled series for the period 1848–2005 was again tested for homogeneity as a whole. Finally, homogenous Brno pressure and temperature series for 1848–2005 at 0700, 1400 and 2100 hours were obtained, from which corresponding daily and monthly averages were calculated. Fig. 11 shows fluctuations in annual averages both for the compiled homogeneous Brno series and the original series from the various places, in which it was derived.

16.0 15.0

Temperature (°C)

14.0 13.0 12.0 11.0 10.0 9.0 1

2

3

4

5

6

7

8 1998

1988

1978

1968

1958

1948

1938

1928

1918

1908

1898

1888

1878

1868

1858

1848

8.0

Fig. 10. Fluctuations of annual averages of air temperature series at 1400 hours (1 – compiled Brno series, 2 – Brno–various places and Brno-Pisárky, 3 – Brno-Květná, 4 – original Brno–Tuřany, 5, 6, 7, 8 –series smoothed by Gaussian low pass filter for 10 years)

7. CONCLUSIONS

This work was carried out in quest of a proper methodology for daily data homogenization and made an attempt to apply it subsequently to daily air pressure and temperature series for Brno in the period 1848–2005. Different methods for the homogenization of daily values were sought, and finally applied to find possible inhomogeneities and to obtain adjusted, homogeneous series. Although further investigation in this matter is required, progress so far may be summarized as: (i) Two basic approaches, based on the homogenization of monthly series and projection of estimated monthly adjustments into a smoothed annual course of daily adjustments, or homogenization of daily values of individual months, estimating proper adjustments for each calendar day with smoothing adjustments, can be used. (ii) The same final adjustments may be obtained from either monthly averages or through direct use of daily data. For the daily-values-based approach, it seems reasonable to smooth them with a low-pass filter for 60 days. The same results may be derived using a low-pass filter for two months (weights approximately 1:2:1) and subsequently distributing the smoothed monthly adjustments into daily values. (iii) The values of the correlation coefficients between the candidate and reference series for daily data (working with each month individually) are comparable with values gained from monthly averages, although daily data are better in some months, monthly data in others. For this reason, a combination of both approaches in (i) is useful. (iv) It is profitable to analyze series of individual observation hours because inhomogeneities manifest in different ways within their series – this is the case for the number of inhomogeneities detected, the value of change, the correlations between reference and tested series (and thus detectability of inhomogeneities) and other characteristics. Series of daily averages can serve as complementary information in the course of homogeneity test evaluation. For inhomogeneity assessment, we recommend the use of as much information as possible.

(v) The data processing in this work has been done by means of LoadData software (application for downloading data from central database, e.g. Oracle), ProClimDB software for processing whole datasets (finding outliers, combining series, creating reference series, preparing data for homogeneity testing, etc.) and AnClim software for homogeneity testing (http://www.klimahom.com/software). Further development of the software, e.g. connection with R software, is to be assumed.

Acknowledgements The authors would like to acknowledge the financial support of the Grant Agency of the Czech Republic for project no. 205/05/0858. Sincere thanks should be extended to Ms Zuzana Pilařová, Institute of Geography, Masaryk University, Brno for digitizing and checking a great part of data used for analysis. Mr. Tony Long, Auchrae, Scotland, undertook the English editing.

References Alexandersson, A. (1986): A homogeneity test applied to precipitation data. Journal of Climatology, 6, 661–675. Alexandersson, A. (1995): Homogenity testing, multiple breaks and trends. In: Proc. 6th Int. Meeting on Stat. Climatology, Galway, Ireland, 439–441. Alexandersson, H., Moberg, A. (1996): Homogenization of Swedish temperature data. Part I: A homogeneity test for linear trends. In: Moberg, A: Temperature Variations in Sweden Since the 18th Century. The Department of Physical Geography, Stockholm University. Dissertation Series No. 5, 98 pp. Auer, I., Böhm, R., Schöner, W. (2001): Austrian long-term climate 1767–2000, Multiple instrumental climate time series form Central Europe. Österreichische Beiträge zu Meteorologie und Geophysik, Heft 25, Wien, 155 pp. Brandsma, T. (2000): Weather-type dependent homogenization of daily Zwanenburg/De Bilt temperature series. http://www.met.hu/omsz.php?almenu_id=omsz&pid=seminars&pri=6&mpx=1&sm0=0&tfi=brandsma Brázdil, R. (1979): Historie měření srážek v Brně (History of precipitation measurements in Brno). Scripta Fac. Sci. Nat. UJEP Brunensis, 9, Geographia 2, 55–74. Brázdil, R., Valášek, H., Macková, J. (2005): Meteorologická pozorování v Brně v první polovině 19. století. Historie počasí a hydrometeorologických extrémů (Meteorological Observations in Brno in the First Half of the Nineteenth Century. History of Weather and Hydrometeorological Extremes). Archiv města Brna, Brno, 452 pp. Easterling, D. R., Peterson, T. C. (1995): A new method for detecting undocumented discontinuities in climatological time series. International Journal of Climatology, 15, 369–377. Mekis, É., Vincent, L. (2004): New developments in the homogenization of precipitation and temperature series in Canada. In: Fourth seminar for homogenization and quality control in climatological databases (Budapest, Hungary, 6-10 October 2003), WCDMP-No. 56. WMO, Geneva, 39–62. Mitchell, J. M., ed. (1966): Climatic Change. WMO, T.N. 79, Geneva, 80 pp. Peterson, T. C. (1998): Homogeneity adjustments of in situ atmospheric climate data: a review. International Journal of Climatology, 18, 1493–1517. Petrovic, P. (2004): Detecting of inhomogeneities in time series using Real Precision Metod. In: Fourth seminar for homogenization and quality control in climatological databases (Budapest, Hungary, 6-10 October 2003), WCDMP-No. 56. WMO, Geneva, 79–88. Pettit, A. N. (1979): A non-parametric approach to the change-point problem. App. Statist., 28, 2, 126–135. Štěpánek, P., 2005: Variabilita teploty vzduchu na území České republiky v období přístrojových měření (Air Temperature Fluctuations in the Czech Republic in the Period of Instrumental Measurements). Ph.D. Thesis. Institute of Geography, Masaryk University, Brno, 136 pp. Štěpánek P. (2006a): AnClim - software for time series analysis. Institute of Geography, Masaryk University, Brno, 1.47 MB. http://www.klimahom.com/software/AnClim.html Štěpánek, P. (2006b): ProClimDB – software for processing climatological datasets. CHMI, regional office Brno. http://www.klimahom.com/software/ProcData.html Wijngaard, J. B. , Klein Tank, A. M. G. , Können, G. P. (2003): Homogenity of 20th century European daily temperature and precipitation series. Int. J. Climatol., 23, 679–692

Příloha č. 3

ČESKÁ GEOGRAFIE V EVROPSKÉM PROSTORU

XXI. sjezd České geografické společnosti České Budějovice 30. 8. – 2. 9. 2006

Homogenizace řad denních hodnot tlaku a teploty vzduchu v Brně v období 1848-2005 Homogenization of Daily Air Pressure and Temperature Series in Brno for the Period 1848-2005 Ladislava Řezníčková1), Petr Štěpánek2), Rudolf Brázdil1) 1) 2)

Masarykova univerzita v Brně, Přírodovědecká fakulta, Geografický ústav, Kotlářská 2, 611 37, Brno Český hydrometeorologický ústav – pobočka Brno, Kroftova 43, 616 67, Brno e-mail: [email protected], [email protected], [email protected]

Klíčová slova: homogenizace denních řad, teplota vzduchu, tlak vzduchu, Brno Key words: homogenization of daily series, air tepmerature, air pressure, Brno

Úvod V posledních letech byla věnována zvýšená pozornost homogenizaci a analýze řad denních hodnot meteorologických prvků. Do současné doby však nebyl nalezen obecný postup homogenizace, který by se dal aplikovat na různé meteorologické prvky, odlišný charakter klimatu určité lokality atd. V současnosti existuje jen několik zahraničních studií, zabývajících se homogenizací denních řad (viz např. Brandsma 2000; Wijngaard a kol. 2003; Mekis, Vincent 2004; Brandsma, Können 2006). Předložená studie je příspěvkem k metodice zpracování a homogenizace denních údajů, která byla následně aplikována na řady denních hodnot tlaku a teploty vzduchu v Brně v období 1848-2005 s cílem získat homogenní řady denních hodnot pro oba meteorologické prvky. Protože měření v Brně probíhala v uvedeném období na několika lokalitách, bylo třeba pro následnou homogenizaci sestavit odpovídající souvislou reprezentativní brněnskou řadu. Takovéto řady obou meteorologických prvků z jednotlivých míst Brna byly nejprve testovány na relativní homogenitu a poté byly podle souběžně probíhajících měření kompilovány odpovídající řady pro oba meteorologické prvky a celé zpracované období. Všechny výpočty byly prováděny v programech AnClim a ProClimDB (Štěpánek 2006a, 2006b).

Stručná historie meteorologických pozorování v Brně Dosud nejstarší denní meteorologické záznamy z Brna pocházejí od penzionovaného setníka Ferdinanda Knittelmayera z let 1799-1812. Pro následující období 1813-1819 nezůstaly meteorologické záznamy zachovány v podobě denních měření. Denní záznamy tlaku a teploty vzduchu se znovu objevují od roku 1820, přičemž až do roku 1847 byly pravidelně publikovány v novinách Brünner Zeitung. Měsíční řady obou analyzovaných meteorologických prvků pro období 1800-1850 byly již homogenizovány a zpracovány v publikaci Brázdila a kol. (2005). Rokem 1848 začíná v Brně oficiální řada pravidelných denních meteorologických pozorování, která vykonával dr. Pavel Olexík, lékař ve všeobecné nemocnici u sv. Anny na Pekařské ulici č. 53 (Obrázek 1). Pozoroval třikrát denně v termínech 06:00, 14:00 a 22:00 hodin. Dne 3. prosince 1853 přemístil pozorovací přístroje z nemocnice do domu na Pekařskou ulici č. 100, kde sám bydlel. Toto místo bylo od původní stanice vzdáleno přibližně 250 m a výškový rozdíl činil asi 7-8 vídeňských sáhů (tj. 13,3-15,2 m) (Liznar 1886). Na novém místě pozoroval Olexík

1258



až do 30. června 1878. V době jeho nepřítomnosti či nemoci mu s meteorologickými pozorováními pomáhal opat augustiniánského kláštera na Starém Brně Gregor Johann Mendel, který od 1. července 1878 pozorování po vážně nemocném Olexíkovi přebral. Konal je třikrát denně v prelátské zahradě kláštera ve standardních klimatologických termínech 07:00, 14:00 a 21:00 hodin. Od července 1883 vykonával pozorovací činnost místo nemocného Mendla řádový kněz Leo Ledwina. Na meteorologická pozorování ze starobrněnského kláštera navázal od 1. ledna 1884 profesor Alfred Lorenz v prostorách c. k. vysoké školy technické v Brně. Měřil až do své smrti v červnu roku 1890, kdy byla pozorování teploty a tlaku vzduchu ukončena. V témž měsíci však již byla uvedena do provozu stanice Brno-Pisárky, umístěná v areálu městské vodárny. Stanice působila jako klimatologická do roku 1937, přičemž měření teploty vzduchu zde pokračovala až do roku 1962, i když stanice fungovala již jen jako srážkoměrná (Brázdil 1979). Další klimatologické stanice vznikly v Brně na počátku 20. století a v jeho dalším průběhu. V této práci byla dále použita stanice na ulici Květná, která byla v činnosti od 1. července 1922 do 31. prosince 1970. Další zpracovávaná stanice je lokalizována jihovýchodně od města na letišti v Tuřanech. Pozorování zde byla zahájena dne 14. dubna 1958 a pokračují do současnosti. Proto byla homogenizace řad denních hodnot teploty a tlaku vzduchu provedena s ohledem na tuto stanici.

Obrázek 1 Poloha meteorologických stanic v Brně

Stanice Vodní tok Zastavěná plocha Vodní plocha Vegetace

Zdroj: Štěpánek a kol. (2006) Vysvětlivky: 1 – nemocnice u sv. Anny, 2 – Pekařská ulice č. 100, 3 – augustiniánský klášter, 4 – c. k. vysoká škola technická, 5 – Pisárky (vodárna), 6 – Květná, 7 – Tuřany (letiště).

Použité údaje K nalezení odlehlých hodnot v měřeních tlaku a teploty vzduchu a pro testování relativní homogenity brněnských řad byly použity některé další referenční stanice s dlouhodobými měřeními (Tabulka 1). Při analýze obou meteorologických prvků se kromě řad denních průměrů vycházelo i z údajů z jednotlivých pozorovacích termínů. Přestože se pozorování před rokem 1890 nekonala ve standardních klimatologických termínech (tj. 07:00, 14:00 a 21:00 hodin), bylo rozhodnuto, že se pro tyto termíny

1259



nebude provádět korekce naměřených hodnot a že případná nehomogenita bude odstraněna při testování.

Tabulka 1 Základní informace o stanicích použitých při homogenizaci brněnských řad Název stanice Brno (různá místa) Brno-Pisárky Brno-Květná Brno-Tuřany Bystřice pod Hostýnem Český Těšín

Zeměpisná šířka (s.š.)

Zeměpisná Nadmořská Začátek délka (v.d.) výška (m) řady Teplota vzduchu

Konec řady

Termíny pozorování 07 (06), 14 (13), 21 (22) 07, 14, 21 07, 14 , 21 07, 14, 21 07 (06), 14, 21 (22) 07, 14, 21 07, 14, 21 (22) 07 (08), 14, 21 (22) 07 (08), 14, 21 (20)

49°12´

16°37´

225

01.01.1848

31.12.1889

49°12´ 49°12´ 49°09´

16°34´ 16°34´ 16°42´

203 223 241

01.06.1890 01.07.1922 14.04.1958

31.12.1962 31.12.1970 31.12.2005

49°24´

17°40´

315

01.09.1865

31.12.2005

49°44´

18°37´

280

01.01.1885

31.10.1938

Holešov

49°19´

17°34´

224

01.07.1895

31.12.2005

Jihlava

49°23´

15°32´

560

27.07.1873

31.12.1934

Olomouc

49°36´

17°15´

215

01.01.1876

31.12.1960

50°05´

14°25´

191

01.01.1775

31.12.2005

07, 14, 21

49°25´

17°24´

203

01.04.1874

31.12.1979

07, 14, 21

48°13´

16°21´

199

01.01.1872

31.12.2005

07, 14, 19

PrahaKlementinum Přerov Vídeň-Hohe Warte

Tlak vzduchu Brno (různá místa) Brno-Pisárky Brno-Květná Brno-Tuřany Holešov PrahaKlementinum Vídeň-Hohe Warte

49°12´

16°37´

225

01.01.1848

31.12.1889

49°12´ 49°12´ 49°09´ 49°19´

16°34´ 16°34´ 16°42´ 17°34´

203 223 241 224

01.06.1890 01.07.1922 14.04.1958 01.01.1961

31.12.1962 31.12.1970 31.12.2005 31.12.2005

07 (06), 14 (13), 21 (22) 07, 14, 21 07, 14 , 21 07, 14, 21 07, 14, 21

50°05´

14°25´

191

01.08.1787

31.01.2001

14

48°13´

16°21´

199

01.01.1872

31.12.2005

07, 14, 19

Homogenizace denních řad tlaku a teploty vzduchu v Brně Řady tlaku a teploty vzduchu byly zpracovány pro denní, měsíční, sezónní a roční hodnoty. Před vlastní homogenizací řad byly nejprve zjištěny a ověřeny vychýlené denní hodnoty (odlehlá pozorování), které mohly být chybné. Testování řad na homogenitu bylo provedeno pomocí softwaru AnClim (Štěpánek 2006a) s použitím Alexanderssonova testu pro jednoduchý zlom (Alexandersson 1986, 1995), bivariačního testu Maronny a Yohaie (Maronna, Yohai 1978), testu Easterlinga a Petersona (Easterling, Peterson 1995) atd., a to v několika iteracích, při nichž byly postupně zpřesňovány výsledků testů. Podle těchto výsledků pro různé referenční řady a testy a s ohledem na metadata byly pro jednotlivé brněnské stanice určeny statisticky významné nehomogenity, které se následně opravily. Ke stanovení příslušných oprav byly použity dva přístupy: a) z měsíčních oprav shlazených nízkofrekvenčním filtrem byly pomocí lineární regrese určeny příslušné opravy pro jednotlivé dny, b) opravy pro jednotlivé dny v průběhu roku byly shlazeny nízkofrekvenčním filtrem a aplikovány vždy na odpovídající den v roce. Oba přístupy vedly ke stejným výsledkům. Chybějící

1260



hodnoty v řadách byly doplněny až po homogenizaci (podrobný popis metodiky viz Štěpánek 2005; Štěpánek a kol. 2006). Vytvoření kompilovaných homogenních řad teploty a tlaku vzduchu v Brně tedy zahrnovalo několik kroků. Nejdříve byly jednotlivé brněnské řady samostatně homogenizovány podle výše popsané metodiky. Z nich byly vytvořeny kompilované řady pro oba meteorologické prvky za celé období 1848-2005 podle souběžně probíhajících měření na více stanicích prodloužením řad stanice Brno-Tuřany do minulosti, a to nejprve podle údajů stanice Brno-Květná (do roku 1923) a následně podle měření v Brně-Pisárkách (do roku 1890). Poté byly řady tlaku a teploty vzduchu bez opravy prodlouženy o naměřené hodnoty až do roku 1848 (Obrázek 2). Následně byla provedena homogenizace takto kompilované řady pro období 1848-2005. Tento přístup byl aplikován pro jednotlivé pozorovací termíny.

Obrázek 2 Schéma sestavení kompilované řady tlaku a teploty vzduchu z měření na několika stanicích v Brně

Teplota vzduchu

Termín pozor.

06,14,22 07,14,21 06,13,21

Staré Brno*

Období pozor. 07, 14, 21

Pisárky 07, 14, 21

Květná 07, 14, 21

Tuřany

Tlak vzduchu 06,14,22

Termín pozor.

07,14,21 06,13,21

Staré Brno*

Období pozor. 07, 14, 21

Pisárky 07, 14, 21

Květná 07, 14, 21

Tuřany

Poznámka: * Staré Brno – nemocnice u sv. Anny, Pekařská ulice, augustiniánský klášter, vysoká škola technická.

Výsledkem provedeného zpracování jsou řady denních hodnot tlaku a teploty vzduchu v období 1848-2005 v termínech 7:00, 14:00 a 21:00 hodin homogenizované na měření stanice Brno-Tuřany. Ty pak byly použity pro výpočet denních a měsíčních průměrů. Obrázek 3 ukazuje porovnání výsledné kompilované homogenní řady Brna s jednotlivými původními řadami od roku 1848 do současnosti.

1261



Obrázek 3 Kolísání ročních průměrů teploty vzduchu v termínu 14:00 hodin shlazených Gaussovým nízkofrekvenčním filtrem pro 10 let 15

Staré Brno

Teplota vzduchu (°C)

14

Brno - Květná Brno - Tuřany

13

Brno - Pisárky

12

11

10 3

4

1

2 1998

1988

1978

1968

1958

1948

1938

1928

1918

1908

1898

1888

1878

1868

1858

1848

9

Vysvětlivky: 1 – Brno (různá místa), 2 – kompilovaná řada Brna po homogenizaci.

Závěr Předložená práce se zabývala nalezením vhodné metodiky pro homogenizaci řad denních údajů meteorologických prvků. Ta byla aplikována na řady denních hodnot tlaku a teploty vzduchu v období 1848–2005 v Brně s cílem zjistit možné nehomogenity v odpovídajících řadách. Po jejich odstranění byly získány homogenní řady denních hodnot obou meteorologických prvků, které budou základem pro další analýzy. Při zpracování byly ověřeny nové postupy, především pokud jde o homogenizaci denních řad.

Literatura ALEXANDERSSON, A. (1986): A homogeneity test applied to precipitation data. International Journal of Climatology, 6 (6): 661-675. ALEXANDERSSON, A. (1995): Homogeneity testing, multiple breaks and trends. In: KOLEKTIV (eds.): „Proceedings of the 6th International Meeting on Statistical Climatology“. Galway, Ireland, s. 439-441. BRANDSMA, T. (2000): Weather-type dependent homogenization of the daily Zwanenburg/De Bilt temperature series. Dostupné na http://www.met.hu/omsz.php?almenu_id=omsz&pid= seminars&pri=7&mpx=1&sm0=0&tfi=brandsma. BRANDSMA, T., KÖNNEN, G. P. (2006): Application of nearest-neighbor resampling for homogenizing temperature records on a daily to sub-daily level. International Journal of Climatology, 26 (1): 75-89. BRÁZDIL, R. (1979): Historie měření srážek v Brně. Scripta Facultatis Scientiarum Naturalium Universitatis Purkynianae Brunensis, Geographia, 9 (2): 55-74. BRÁZDIL, R., VALÁŠEK, H., MACKOVÁ, J. (2005): Meteorologická pozorování v Brně v první polovině 19. století. Historie počasí a hydrometeorologických extrémů. Archiv města Brna, Brno, 452 s. EASTERLING, D. R., PETERSON, T. C. (1995): A new method for detecting undocumented discontinuities in climatological time series. International Journal of Climatology, 15 (4): 369377. LIZNAR, J. (1886): Ueber das Klima von Brünn. Sonder-Abdruck aus dem 24. Bande der Verhandlungen des naturforschenden Vereines in Brünn, Brünn, 70 s.

1262



MARONNA, T., YOHAI, V. J. (1978): A bivariate test for the detection of a systematic change in mean. Journal of American Statistical Association, 73: 640-645. MEKIS, É., VINCENT, L. (2004): New developments in the homogenization of precipitation and temperature series in Canada. In: KOLEKTIV (eds.): “Fourth Seminar for Homogenization and Quality Control in Climatological Databases”. World Meteorological Organization, Geneva, s. 39-62. ŠTĚPÁNEK, P. (2005): Variabilita teploty vzduchu na území České republiky v období přístrojových měření. Disertační práce. Masarykova univerzita v Brně, Přírodovědecká fakulta, Geografický ústav, 136 s. ŠTĚPÁNEK P. (2006a): AnClim – software pro analýzu časových řad. Masarykova univerzita v Brně, Přírodovědecká fakulta, Geografický ústav. Dostupné na http://www.klimahom. com/software/AnClim.html. ŠTĚPÁNEK, P. (2006b): ProClimDB – software pro zpracování klimatologických databází. Český hydrometeorologický ústav, pobočka Brno. Dostupné na http://www.klimahom.com/ software/ProcData.html. ŠTĚPÁNEK, P., ŘEZNÍČKOVÁ, L., BRÁZDIL, R. (2006, v tisku): Homogenization of daily air pressure and temperature series for Brno (Czech Republic) in the period 1848-2005. World Climate Data and Monitoring Programme, World Meteorological Organization, Geneva. WIJNGAARD, J. B., KLEIN TANK, A. M. G., KÖNNEN, G. P. (2003): Homogeneity of 20th century European daily temperature and precipitation series. International Journal of Climatology, 23 (6): 679-692.

Grantové agentuře České republiky patří poděkování za finanční podporu pro řešení grantu č. 205/05/0858, v rámci něhož tento příspěvek vznikl.

Summary Homogenization of daily meteorological series is a difficult task. Several kinds of problems have to be taken into consideration in the course of homogenization: selection of a proper homogenization method with regard to the data used, creation of reference series, completion of missing values, annual course of adjustments, and others. This paper presents an attempt to create a homogeneous series of daily air pressure and temperature readings in the city of Brno in the period 1848-2005. Two basic approaches were adopted: (i) homogenization of monthly series and projection of estimated smoothed monthly adjustments in annual variation of daily adjustments and (ii) homogenization of daily values in individual months and direct estimation of daily adjustments, again smoothed by low-pass filter.

1263

Příloha č. 4

Theor. Appl. Climatol. 88, 179–192 (2007) DOI 10.1007/s00704-006-0253-5 Printed in The Netherlands

1 2

Institute of Geography, Masaryk University, Brno, Czech Republic Czech Hydrometeorological Institute, Prague, Czech Republic

Meteorological singularities in the Czech Republic in the period 1961–2002 ˇ eznıćˇkova´1 , R. Bra´zdil1 , and R. Tolasz2 L. R With 6 Figures Received December 2, 2005; revised May 12, 2006; accepted June 1, 2006 Published online September 6, 2006 # Springer-Verlag 2006

Summary Based on the mean daily values of air temperature, air pressure, and precipitation totals at 13 climatological stations within the territory of the Czech Republic in the period 1961–2002, a statistical analysis of ‘‘meteorological singularities’’ (i.e., calendar-dependent deviations from the mean annual variation for selected meteorological elements) was performed. At the 13 stations analysed, a total of 45 meteorological singularities (37 singularities in air temperature, 35 in air pressure, and 30 in precipitation) were found. The singularities detected correlate well with cases traditionally recognised in the Czech Republic as well as with the results of analyses performed for Germany. Despite the considerable variability of singularities in time and space, most of them are found across the entire territory of the Czech Republic and can be observed for the most part in all three elements processed. The majority of the singularities detected may be explained on the basis of circulation mechanisms, by relating them to a significantly higher occurrence of certain groups of synoptic situations characterised by anomalous temperature or precipitation effects. Cases of ‘‘competition’’ between singularities, when different singularities may occur on the same calendar day, were found.

1. Introduction The mean annual variation in the daily values of individual meteorological elements occurs as a result of changes in annual insolation resulting astronomically from the orbital motion of the earth

around the sun. At different times of the year, however, larger or smaller deviations from the mean annual variation may occur, usually related to a higher occurrence of typical synoptic situations; such deviations are commonly known as ‘‘meteorological singularities’’. This term, which denotes those deviations in the mean daily values of meteorological elements from corresponding smoothed mean annual variation, showing some relation to calendar date, was first used by A. Schmauss in 1928 as an analogy to singularities of mathematical functions (Schmauss, 1928). The study of meteorological singularities has become a relatively frequent topic of study, particularly in German climatology (e.g., Schmauss, 1938; Flohn and Hess, 1949; Flohn, 1954; Pleiss, 1961; Gerstengarbe and Werner, 1987), but also in the Czech Republic (e.g., Zikmunda, 1954; Bayer, 1955, 1960; Nosek, 1957; Stuchlı´k, 1960). Some work has dealt directly with particular singularities, such as the ‘‘Christmas thaw’’ (Bayer, 1956a) or the ‘‘European summer monsoon’’ (Bayer, 1956b; Bra´zdil, 1982). Since the late 1980s, new statistical approaches have come to be applied to the study of singularities, as can be seen primarily in papers by Bissolli and Sch€onwiese (1987, 1990) and Bissolli (1991). A statistical approach has also been used in

180

L. Rˇeznıćˇkova´ et al.

the Czech Republic to analyse singularities in several meteorological elements at Milesˇovka Mt. (Bra´zdil et al., 1999). More recently, using German methodology, meteorological singularities in the Czech Republic were reviewed in a diploma thesis by Rˇeznıćˇkova´ (2004). The typical and recurrent effects of some meteorological singularities have also given rise to a body of popular weather lore (Munzar, 1985) expressing long-standing experience accumulated in observing weather variations. Those close to the land have thus established meteorological rules with which they not only described the individual phases of weather and nature in their annual cycles, but also tried to predict future developments in the meteorological situation based on weather occurring on ‘‘critical days’’ (e.g., St. Medard’s Day). In Sect. 2, the meteorological singularities usually mentioned in the Czech Republic are briefly characterised. After a description of the basic meteorological data and methodology in Sects. 3 and 4, the following section presents the results of singularity analysis for the entire Czech Republic. The final part of the paper summarises and discusses the results obtained.

v)

vi)

vii)

2. Meteorological singularities in the Czech Republic The following meteorological singularities have been described in the Czech Republic (Souborna´ studie, 1969; Bra´zdil et al., 1999): i) deep winter: This is the period of the lowest winter temperatures, around mid-January until the beginning of the third decade of January, characterised by a more frequent occurrence of anticyclones or cooler eastern airflow. ii) early February warming with subsequent cooling: Return of cold weather at the end of the first and the beginning of the second decade of February after a previous warming. iii) mid-March cooling: This singularity was listed in Souborna´ studie (1969) without detailed comments. iv) May cooling: This corresponds with a return of cold weather due to the outbreak of Arctic air into the warming European

viii) ix)

continent during the first half of May. Night frosts often result in damage to vegetation. This singularity is traditionally associated with the days known in Czech as ‘‘ledovı´ muzi’’, the ‘‘Icemen’’ (12–14 May). European summer monsoon: This phenomenon expresses itself in a temperature drop, higher cloud amount, and several precipitation waves with rises in precipitation totals and frequencies of occurrence. Its onset is usually related to the turn of the first and second decades of June (often associated with 8 June, St. Medard’s Day), though several ‘‘monsoon waves’’ may develop later in the summer as a result of westerly and north-westerly airflow with which individual frontal systems advance into Central Europe. high summer: Marked by the highest air temperatures of the year, accompanied by a decrease in precipitation totals and precipitation activity during the second half of July. This period is generally related to the stabilisation of a sunny, anticyclonic weather regime. Indian summer: The dominant anticyclonic character of the weather in September or October results in dry, fairly still, sunny, and – during the daytime – warm weather. Nights are already rather cold with frequent radiation fogs. This type of weather is caused by an extensive anticyclone over Central and South-Eastern Europe. Although known literally in Czech as ‘‘babı´ leto – Old women’s summer’’, it is comparable with ‘‘Indian summer’’ defined for North America and used in Great Britain for a very similar phenomenon. late November temperature rise: This singularity was listed in Souborna´ studie (1969) without detailed comment. Christmas thaw: A period of rather warm and wet weather beginning around Christmas in Central Europe and lasting until the end of the year. It is related to the flow of warmer oceanic air from the south-west–west after a previous period of frosts. At lower and medium elevations, it is accompanied by rains and thaws, while at higher elevations it can take the form of heavier snowfalls.

Meteorological singularities in the Czech Republic in the period 1961–2002

3. Data used Mean daily surface air temperatures and pressures, as well as precipitation totals from 13 climatological stations of the Czech Hydrometeorological Institute (CHMI) within the territory of the Czech Republic for the period 1961–2002 were used for statistical analysis of meteorological singularities. The selected stations (Fig. 1, Table 1) represent various geographical areas and various elevations, from mountain sites (e.g., Lysa´ hora Mt. and Milesˇovka Mt.) to medium-elevations (e.g.,

181

Prˇibyslav) and lowland sites (e.g., Brno-Turˇany, Olomouc). However, in several cases the meteorological data used was not complete and the missing values had to be filled in. This was done as follows: Neighbouring stations with correlation coefficients of at least 0.95 were used to calculate the absent values for air temperature and pressure for a given station. Using these stations, an average (reference) series was then calculated from which the missing values were determined by linear

Fig. 1. Geographical distribution of climatological stations used in the Czech Republic Table 1. Stations used for the analysis of meteorological singularities Abbreviation

BTUR CBUD CHEB HRAD LIBC LYSA MILO MOSN OLOM PKAR PLZD PRIB SUMP

Station name

Brno-Turˇany Cˇeske Budeˇjovice Cheb Hradec Kra´love Liberec Lysa´ hora Mt. Milesˇovka Mt. Ostrava-Mosˇnov Olomouc Prague-Karlov Plze n-Doudlevce Prˇibyslav Sˇumperk

Geographical coordinates Latitude

Longitude

49 090 N 48 580 50 050 50 110 50 460 49 320 50 330 49 410 49 340 50 040 49 430 49 350 49 580

16 420 E 14 280 12 240 15 500 15 010 18 270 13 560 18 070 17 140 14 260 13 240 15 460 16 580

Altitude (m a.s.l.) 241 388 474 278 398 1324 833 251 225 232 311 530 311


182

Table 2. Groups of temperature-significant synoptic situations (after Bra´zdil et al., 1999) Abbr. of group

Synoptic situations

Characteristics of air temperature

A1 A2 A3 A4 A5 A6 A7

C, B, Bp, Nc NWa, NWc SWa, Sa, Wal SWc1, SWc2, SWc3 Wc, Wcs, Wa A Ec, Ea, SEc, SEa

below normal for the whole year below normal in summer, above normal above normal for the whole year above normal for the whole year below normal in summer, above normal above normal in summer, below normal above normal in summer, below normal

in winter

in winter in winter in winter

Table 3. Groups of precipitation-significant synoptic situations (after Bra´zdil et al., 1999) Abbr. of group

Synoptic situations

Characteristics of the group

B1 B2 B3 B4 B5

C, B, Bp NEc, Ec Wc, Wcs NWc, Nc SWc2, SWc3

precipitation in central lows and troughs of low air pressure precipitation related to retrograde lows and eastern airflow precipitation related to upper frontal zones precipitation in cold air during surface north-western airflow precipitation during upper south-western airflow on undulated cold fronts, and=or in cold ridges of high pressure beyond fronts, typical of summer

regression. In this manner, the data were completed for the following stations: Olomouc (31 October 2002), Plzen-Doudlevce (1 August–31 October 1967; 1 June–31 December 1968; 1–30 April 1969; 1 June–31 August 1969; 1–30 November 1969), and Sˇumperk (1 January 1971–28 February 1974; 14–31 March 1974; 1 October 1989–31 March 1990). Air temperature for the PlzenDoudlevce station from 1 June 1997 to 31 December 2002 was calculated differently, using Plzen-meˇsto (city) as the reference station. From the 7.5-year overlap of observations at both stations, the mean temperature difference for each day was determined and used to correct the series for the Plzen-city station. When filling in precipitation totals, figures from the nearest stations that most resembled the station processed in their physical-geographical character were always taken. The missing precipitation totals for the Plze n-Doudlevce station were thus adopted from the Plze n-Radcˇice station (1 August–31 October 1967; 1 June–31 December 1968; 1–30 April 1969; 1 June–31 August 1969; 1–30 November 1969) and the Plze n-city station (from 1 June 1997). The data for Sˇumperk (1 October 1989–31 March 1990) was replaced by precipitation totals from the Velke Losiny station.

Because a CHMI classification of synoptic situations (Katalog, 1967, 1972) has been available since 1946, the relation of the singularities detected in meteorological elements to the occurrences of selected synoptic situations in the period 1961– 2002 was also analysed. Several groups of synoptic situations that show significant deviations in temperature (Table 2) and precipitation (Table 3) were established. Circulation patterns for the selected synoptic situations are shown in Fig. 2. 4. Methods of analysis The analysis of meteorological singularities is based on the methodological approach of Bissolli and Sch€onwiese (1987, 1990) and Bissolli (1991). As regards air temperature and pressure, singularities were determined using the mean, median, skewness, and relative frequency of the occurrence. Due to the asymmetric distribution of the data for precipitation totals, all these characteristics could not be applied and thus singularities were determined only from means and frequencies of occurrence. The occurrence of synoptic situations and the relation between the individual meteorological elements and=or between the meteorological elements and the synoptic situations were investigated using frequency analysis.


183

Fig. 2. Scheme of synoptic situations of the CHMI classification (modified after Krˇivancova´ and Vavrusˇka, 1997). The abbreviation for the situation is in the upper right corner, appertaining to the group of situations significant for temperature and precipitation, respectively, in bottom right corner (see Tables 2–3). Key: H high, L low, 1 mean position of frontal zone, 2 border between cyclonic and anticyclonic fields, 3 border of region with alternation of highs and lows, 4 pathway of highs, 5 pathway of lows

184


Fig. 2 (continued)

4.1 Data filtration To remove annual variation, temperature and pressure series were filtered using a sixty-day Gaussian high-pass filter. In addition to removing annual

variation, this filtering also allows the removal of possible trends in a time series. For both elements, the filtering effect is illustrated using the example of the Prague-Karlov station for the year 1982 (Fig. 3). While the sixty-day low-pass filter

Fig. 3. Mean daily air temperature (left) and pressure (right) for the Prague-Karlov station in 1982, smoothed by 60-day Gaussian low-pass filter (a) and high-pass filter (b)


eliminates periods T <60 days and emphasises rather well the annual variation resulting from astronomical factors, the high-pass filter removes the annual variation and preserves only fluctuations with shorter periods (T < 60 days), i.e., deviations at; j from long-term variability are obtained (t ¼ 1; 2; . . . ; 365 are the individual days of the years j ¼ 1; 2; . . . ; N). 4.2 Method of the mean To remove large fluctuations, the values at; j were smoothed before further analysis using calendarrelated three-day running averages (values mt ) according to PN j¼1 ðat1; j þ at; j þ atþ1; j Þ : ð1Þ mt ¼ 3N Thanks to this data smoothing, it is also easier to compare the values of meteorological elements with the duration of one synoptic situation, which is usually at least three days. If there is no singularity in a time series, then the theoretical expectation value should be mt ¼ 0. Thus the null hypothesis H0 : mt ¼ 0 is tested against the alter-

185

native hypothesis H1 : mt 6¼ 0. The relevant significance levels are given by s ð2Þ m0 z pffiffiffiffi ; ns where m0 is the mean of the sample, i.e., the mean of all values at; j ; z is the abscissa value of the standard normal (Gaussian) distribution; s is the standard deviation of the sample, i.e. for mean values according to relation (1) for all years; and ns is the sample size ðns ¼ 3NÞ. It is assumed in formula (2) that the data set has a normal distribution. If the values of mt exceed a significance level of ¼ 80%, they are statistically significant, and the occurrence of a ‘‘singularity in the mean’’ on the respective day can be stated (Fig. 4a). The values mt characterise the intensity of singularity. The number of successive days with a singularity determines its mean duration, but says nothing about the dispersion of values during a given day. 4.3 Method of the median To detect singularities, the median Med can also be used, which divides a sequence of values (arranged in ascending order) into two equally large parts

Fig. 4. Mean air temperature deviations (a), the median of pressure deviations (b), coefficient of skewness of pressure deviations (c) and relative frequency of the incidence of ‘‘singular events’’ (d) in the period 1961–2002 for the PragueKarlov station. Significance levels of 80% (thin line) and 95% (thick line) determine ‘‘singularities in the mean’’, ‘‘median’’, ‘‘skewness’’ and singularly ‘‘warm’’ days

186


according to the number of elements. It holds that aj ak for j0; if it is skewed to the right, Sf <0. Skewness is also calculated for calendar-related three-day running averages, as in Sect. 4.2. Similarly, the null hypothesis H0 : Sft ¼ 0 is tested against the alternative hypothesis H1 : Sft 6¼ 0, where t ¼ 1; 2; . . . ; 365. The corresponding significance levels are calculated from rffiffiffiffi 6 ; ð6Þ Sf0 z ns where Sf0 is the skewness of the basic sample, i.e., the skewness based on all values at; j . The sig-

nificance level of ¼ 80% was again chosen as the threshold for determining ‘‘singularities in the skewness’’ (Fig. 4c). Using the skewness coefficient, information on the time of occurrence and duration of the singularity may be obtained. 4.5 Method of frequencies of occurrence The last characteristic that enables the detection of the occurrence of singularities is the incidence of ‘‘singular events’’. Daily values of ai ; i ¼ 1; 2; . . . ; n, are used as basic data. A value is considered ‘‘singularly’’ high if it exceeds the 80th percentile of the data and ‘‘singularly’’ low if it falls below the 20th percentile. Specifically, this terminology means ‘‘singularly warm’’ and ‘‘singularly cold’’ regarding air temperature, ‘‘singularly high’’ and ‘‘singularly low’’ regarding air pressure, and ‘‘singularly wet’’ and ‘‘singularly dry’’ regarding precipitation. In contrast to the previous methods, this approach can also be used to evaluate frequencies of synoptic situations, significant for temperature and precipitation, which are determined nominally. For some meteorological elements, the frequencies of ‘‘singular events’’ may be influenced by a large dispersion of the values and by extremes during a certain part of the year. This is most evident for air pressure during winter and for precipitation totals in summer. To remove this effect on the results obtained, values entering the frequency analysis were first normalised by using standard deviation. For every event A from a given time series, the relative frequency for the filtered values smoothed using calendar-related three-day running averages was calculated. The null hypothesis is H0 ðAÞ ¼ Ht ðAÞ, where t ¼ 1; 2; . . . ; 365. Numerical data (temperature and pressure, precipitation totals) has the advantage of the fact that H0 ðAÞ ¼ p0 ¼ 0:2;

ð7Þ

where p0 is the expected probability of the occurrence. The appropriate significance levels are given by rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 H0 ðAÞ H0 ðAÞ z H0 ðAÞ ; ð8Þ N where N is the number of years ðN ¼ 1; 2; . . . ; 42Þ. If the significance level ¼ 80% is exceeded, it indicates the occurrence of a ‘‘singularity in the frequency of the occurrence’’. The topic of


interest is only the upper significance level calculated from (8). Figure 4d shows the relative frequency of ‘‘singularly warm’’ days. In the same way as for skewness, singularities calculated according to the frequency of ‘‘singular events’’ yield information on their time of the occurrence and mean duration, but not on their intensity.

187

(e.g., air temperature) and A2 (e.g., precipitation) – which gives an event A (e.g., temperature= precipitation). In line with the previous section, the relative frequencies related to a calendar date can be expressed as HðAÞ ¼ HðA1 ; A2 Þ

ð9Þ

and it holds that 4.6 Two-dimensional frequency of the occurrence

H0 ðAÞ ¼ H0 ðA1 ; A2 Þ ¼ H0 ðA1 ÞH0 ðA2 Þ;

The frequencies of ‘‘singular events’’ from Sect. 4.5 can serve to ‘‘join’’ two types of events – A1

where H0 ðA1 Þ and H0 ðA2 Þ are calculated for the same period as in Sect. 4.5, as well as the

ð10Þ

Fig. 5. Relative frequency of temperature-pressure and temperature-precipitation singularities with significance levels of 80% (thin line) and of 95% (thick line) for the Prague-Karlov station in the period 1961–2002: a) warm – low pressure (TN), cold – high pressure (SV); b) warm – wet (TW), cold – dry (SD)

Fig. 6. Relative frequencies of synoptic situations of the group A7 with respect to ‘‘cold’’ days (a) and of the group B1 with respect to ‘‘wet’’ days with significance levels of 80% (thin line) and of 95% (thick line) for the Prague-Karlov station in the period 1961–2002

188


significance test according to (8). H0 ðAÞ corresponds to the frequency of the joint event, i.e., the relation between the two characteristics. This approach was used to find relations between the following characteristics: air temperature-air pressure (Fig. 5a), air temperatureprecipitation (Fig. 5b), air temperature-synoptic situation (Fig. 6a), and precipitation-synoptic situation (Fig. 6b). 5. Meteorological singularities in the Czech Republic in the period 1961–2002 For establishing a singularity in a series of air temperature, air pressure, and precipitation, at least two of the three characteristics m, Med, and Sf had to be statistically significant. Table 4 gives an overview of the temporal and spatial occurrence of meteorological singularities within the territory of the Czech Republic in the period 1961–2002. Air temperature was a decisive meteorological element for the identification of singularities. However, if a singularity appeared without any link to temperature in the pressure or precipitation series, the singularity was also listed in Table 4 and a possible causal explanation was sought with respect to the occurrence of synoptic situations. Applying the methods described to temperature, pressure and precipitation series from 13 climatological stations led to the disclosure of a total of 45 meteorological singularities, of which 37 were significant through temperature, 35 through pressure, and 30 through precipitation. The mean duration of temperature singularities is 5.9 days and that of pressure singularities 5.1 days; the singularity ‘‘third phase of high summer’’ has the longest duration: 12 days for both elements. The average duration of significant precipitation singularities is 6.7 days. The longest period of increased precipitation activity results in a 10-day singularity in late January. In contrast, a significantly dry 15-day period occurs in late September and in the first half of October during the ‘‘Indian summer’’. Although most singularities are expressed through several meteorological elements, these need not coincide in terms of days. Of the 45 detected meteorological singularities, 21 (46.7%) were recorded at all 13 stations analysed, but in 12 cases (26.7%) they occurred at fewer than half the stations. In some cases, the spatial scope of a

singularity is clearly limited, something perhaps related to the territorial extent of certain synoptic situations. For example, the ‘‘mid-March cooling’’ and the ‘‘Christmas thaw’’ are limited only to the stations in the eastern half of the Czech Republic while the ‘‘late November warming’’ was detected largely at stations located more towards the north of the country. In contrast, the analysis did not establish any unambiguous dependence of singularities on the elevation of the stations. The spatial distribution of the singularities detected also reflects the peculiarities of the areal structure of the individual meteorological elements. While air temperature and pressure fields typically exhibit spatial homogeneity and high correlation over increasing distance (effects of the macro-circulation processes), in the case of precipitation one must take into account its higher spatial variability, related to processes on synoptic and smaller scales, i.e., a more rapid decline in correlation with increasing distance. With respect to temperature and pressure, these properties were visible not only in the fact that the processing methods (analysis of deviations and frequencies) usually led to similar results, but also in the fact that singularities of both elements at the individual stations occurred approximately simultaneously. For precipitation singularities, in contrast, there was less agreement between individual stations. Moreover, for precipitation, the two methods of analysis did not yield results as consistent those for air temperature and pressure. It is obvious from these facts that certain singularities are better revealed by precipitation totals than by the frequencies of significant precipitation days and vice versa. The meteorological singularities detected generally confirm singularities usually cited for the Czech Republic (Souborna´ studie, 1969; Bra´zdil et al., 1999) as well as for Germany in the period 1946–1986 (Bissolli, 1991), for which the occurrence of a coincident singularity was established in 25 cases (55.6%) (Table 4). However, the occurrence of the May-cooling singularity (the ‘‘Icemen’’) is an exception to this, as already pointed out by Bissolli (1991) in his analysis of German stations and by Bra´zdil et al. (1999) in their analysis of singularities at Milesˇovka Mt. in north-western Bohemia. In fact, the period around mid-May actually shows a warm singularity in the second half of the 20th century.

1st high summer 2nd high summer 3rd monsoon wave

2nd monsoon wave

– 12–18 May 21–25 May 29 May–1 June 4–8 June – 15–18 June 25–28 June 2–5 July 11–14 July 17–22 July

–V– T–D S–W S– – TV– –NW S– – T– – TVD TVD SNW

1st monsoon wave

– 31 Mar.–7 Apr. (31 Mar.–7 Apr.) 12–19 Apr. (12–19 Apr.) 28 Apr.–2 May

–NW TND TNW SN– SVD TN–

18–21 Mar.

mid-March cooling

SNW

14–18 Feb. 20–25 Feb.

5–13 Feb.

– 5–11 Mar.

beginning February warming February cooling

–V– TVD

SVD S– –

TNW

29 Jan.–2 Feb.

TVW

12–14 Jan. (12–14 Jan.) 16–21 Jan. 23–27 Jan.

1st deep winter 2nd deep winter 3rd deep winter

SVD SVD SVD

Temperature

5–9 May – – – 1–4 June 8–12 June – – 29 June–3 July 10–13 July 17–19 July

26–29 Mar. 1–6 Apr. (1–6 Apr.) 11–13 Apr. 13–17 Apr. 30 Apr.–4 May

17–20 Mar.

25–27 Feb. 5–11 Mar.

18–23 Feb. –

9–15 Feb.

31 Jan.–4 Feb.

24–28 Jan.

8–10 Jan. 11–19 Jan. 20–22 Jan.

Pressure

Period of singularity occurrence

TNW

Type of singularity

Sing. type

– 13–18 May 18–23 May – – 5–11 June – – 29 June–5 July 9–13 July 14–20 July

25–28 Mar. 30 Mar.–4 Apr. 5–9 Apr. – 18–23 Apr. –

17–21 Mar.

– – A1s, A7s, B1 A1s – B1, B2 – A3t, A4t A3 – A1s, B1

A2s, A5s, B3, B4 B2, B3, B5 A7t B1, B5 A1s – A4t

– A7t

A7s, A6s B4

15–19 Feb. – – 4–11 Mar.

A5t, B3, B1

A4t, A5t, A2t, B3, B4, B5 –

A7s, A6s – –

7–14 Feb.

(23 Jan.–1 Feb.)

23 Jan.–1 Feb.

6–14 Jan. 15–22 Jan. (15–22 Jan.)

Precipitation

Group of synop. situat.

Table 4. Meteorological singularities on the territory of the Czech Republic in the period 1961–2002

all stations CHEB, MILO, MOSN, PKAR, PLZD, SUMP LYSA, PLZD, PRIB BTUR, CHEB, MILO, MOSN, OLOM, SUMP BTUR, HRAD, LYSA, MOSN, OLOM, PRIB all stations all stations all stations all stations except LYSA, MILO, MOSN, PRIB BTUR, HRAD, LYSA, MOSN, PKAR, PLZD, PRIB, SUMP except PLZD, CHEB all stations all stations MOSN, PLZD, SUMP all stations except LYSA all stations all stations all stations LIBC, LYSA, MILO, MOSN, PKAR all stations

BTUR, LIBC, MILO, MOSN, OLOM, PLZD, PRIB all stations

BTUR, HRAD, LIBC, PKAR CBUD, HRAD, MILO BTUR, CBUD, MOSN, OLOM, PKAR, PLZD, PRIB, SUMP all stations

Regional peculiarities

(continued)

– Fs M1 – Sf M2 – – Sh1 Sh1 M4

– Fv2 – VM1 – –

–

Ws2 –

Ws1 –

T5

–

T4

Wh1 Wh2 Wh3

Sing. for Germany

Meteorological singularities in the Czech Republic in the period 1961–2002 189

–

31 Dec.–3 Jan.

– 22–24 Dec. 27–29 Dec.

6–10 Dec.

– 13–20 Nov. 22–25 Nov. 1–4 Dec.

25–30 Sep. – 16–20 Oct. 22–29 Oct.

12–16 Sep.

–

26 Jul.–6 Aug. 14–17 Aug.

Pressure

–

17–20 Dec. – –

6–11 Dec.

1–5 Nov. 13–20 Nov. (13–20 Nov.) 28 Nov.–4 Dec.

28 Sep.–12 Oct. (28 Sep.–12 Oct.) – 24–31 Oct.

11–17 Sep.

–

26 July–1 Aug. 13–17 Aug.

Precipitation

–

B3 – A5t

A2s

A5t A1t, B1, B2 A1 –

– A6t, A3t A6s, A7s –

–

A1s, A4s

A3t, A4t A3t, A4t

Group of synop. situat.

LIBC, LYSA, MILO, MOSN, PKAR, PLZD, PRIB, SUMP all stations all stations BTUR, HRAD, LYSA, MOSN, OLOM, SUMP CHEB, MOSN, PKAR, PLZD, PRIB

all stations except HRAD, LIBC, MOSN, OLOM, PKAR except BTUR, CBUD, LYSA, PKAR, PRIB BTUR, CBUD, LYSA, PKAR, PLZD, PRIB except CBUD, MILO all stations all stations HRAD, LIBC, MOSN, OLOM, PRIB, SUMP except CHEB, MILO, PKAR all stations all stations HRAD, LIBC, LYSA, SUMP

Regional peculiarities

–

– Wf T3

–

– – Hs3 –

– Hm – Hs1

M6

M5

Sh2 Sh3

Sing. for Germany

For each singularity type, its temperature (S cold, T warm), pressure (V high, N low) and precipitation (D dry, W wet) characteristics are given progressively; for the groups of synoptic situations A1–A7, their temperature characteristics (s cold, t warm). Occurrence of the singularity over all stations is expressed in bold type. The period of the singularity occurrence for the given meteorological element, conditioning several types of singularity, is shown in brackets. In case of agreement, a corresponding singularity for Germany in the period 1946–1986 (Bissolli, 1991) is mentioned: Wh deep winter (‘‘Hochwinter’’), T period of thaw (‘‘Tauwetterperiode’’), Ws late winter (‘‘Sp€atwinter’’), Fr early spring (‘‘Vorfr€ uhling’’), VM pre-monsoon wave (‘‘Vormonsunwelle’’), Fs late spring (‘‘Sp€atfr€ uhling’’), M monsoon wave (‘‘Monsunwelle’’), Sf early summer (‘‘Fr€ uhsommer’’), Sh high summer (‘‘Hochsommer’’), Hm mid-autumn (‘‘Mittherbst’’), Hs late autumn (‘‘Sp€atherbst’’), Wf early winter (‘‘Fr€ uhwinter’’)

–N–

– – 26–29 Dec.

– –W –V– TN–

Christmas thaw

6–12 Dec.

late November warming

1–10 Nov. 17–24 Nov. (17–24 Nov.) 26 Nov.–4 Dec.

TVD

T–D SNW SV– TVD

21–28 Sep. 6–13 Oct. 19–22 Oct. 24–28 Oct.

SVD T–D SV– SVD

1st Indian summer 2nd Indian summer

–

–NW

29 July–9 Aug. 14–17 Aug. 22–31 Aug.

3rd high summer 4th high summer

TVD TVD

Temperature

Period of singularity occurrence

S– –

Type of singularity

Sing. type

Table 4 (continued)

190 L. Rˇeznıćˇkova´ et al.


The clear time variability in the onset of some singularities was also established by comparing the results from the period 1961–2002 with a similar analysis for Milesˇovka Mt. in the periods 1905–1994 and 1946–1995 (Bra´zdil et al., 1999). The variability results in time shifts (e.g., the cooling in mid-April and the May warming almost a week earlier, warming in the first half of August one week later), the complete lack of some singularities (e.g., cooling around the February=March turn, warming in early December), or the occurrence of new singularities (e.g., warming in the third decade of January and in the first half of February, cooling in mid-April and at the second=third decade turn in October). Some singularities occur in several phases in the Czech Republic. This is the case, for example, with the ‘‘deep winter’’, which appears in three phases in January. Bissolli (1991) also came to the same conclusion for the German territory, finding that the singularity had the same number of phases but that these occurred one day earlier. Only three of five ‘‘monsoon waves’’ occurring within German territory during the summer months have been detected in the Czech Republic. This is because the monsoon wave from the June=July turn, denoted as M3 by Bissolli (1991) for the German territory, does not occur at all, and the subsequent M5 wave from the last decade of August is not expressed in precipitation in the Czech Republic, but only by a temperature drop. The ‘‘first phase of high summer’’ in early July lasts 12 days in German territory. In the Czech Republic, this period is not continuous and is divided into two phases. In contrast to Germany, two phases of ‘‘Indian summer’’ (there is a single phase in Germany) and only three warming phases (‘‘thaws’’) during the winter half-year were determined in the Czech Republic. However, Bissolli (1991) found a total of five such phases (‘‘Tauwetterperiode’’). In general, the frequency analysis of synoptic situations significant for temperature and precipitation, giving rise to singularities of a certain meteorological element, yielded the results expected. Warm singularities in winter are thus related to the dominant westerly and north-westerly airflow occurring during synoptic situations belonging to groups A2 and A5 (e.g., warming in the last decade of January, ‘‘Christmas thaw’’).

191

Precipitation activity is most often caused by airflow with a westerly component occurring in synoptic situations belonging to groups B3, B4, and B5. Cold singularities in winter are mainly related to easterly and south-easterly airflow occurring in situations belonging to group A7, and=or by the anticyclonic character of the weather (e.g., ‘‘deep winter’’, ‘‘February cooling’’). Temperature increases during the summer halfyear are mostly caused by westerly, south-westerly to southerly winds occurring in situations belonging to groups A3 and A4 (e.g., the warming in the last pentad of June, the ‘‘high summer’’), while the ‘‘Indian summer’’ is related to anticyclonic weather. In contrast, cold singularities in summer are largely caused by frontal activity in cyclones or troughs of low pressure over Central Europe that occur in synoptic situations of the A1 group (the ‘‘monsoon waves’’, cooling in the last decade of August). Such airflow over Central Europe is also associated with precipitation activity occurring in synoptic situations from the group B1. Several types of singularities related to a certain meteorological situation can occur on the same day (‘‘competition’’ of singularities – see Bissolli and Sch€onwiese, 1987, 1990; Bissolli, 1991). For example, in addition to the ‘‘midFebruary cooling’’ singularity, which features dry weather resulting from situations of the groups A6 and A7, effects of significant precipitation situations during prevailing situations of the group B4 have also been found. The above singularity corresponds with a significant decrease in ten-day pressure averages during the second decade of February in Croatia, resulting from a high frequency of low pressure situations accompanied by cloudy and rainy (snowy) weather (Loncˇar et al., 1996). Nevertheless, for some singularities, it is difficult to prove their dependence on the higher incidence of a certain distinct synoptic situation or group of situations. 6. Conclusions In the study of meteorological singularities, a stereotyped approach, based on the premise that we know in advance the singularities that should be found in a given time series (including their origin of circulation), has prevailed for a long time. Whenever a deviation did not fall on the anticipated calendar day, deviations were sought

192

L. Rˇeznıćˇkova´ et al.: Meteorological singularities in the Czech Republic in the period 1961–2002

with the nearest occurrence date in order to complete the picture. One example is a statement by Stuchlı´k (1960) that ‘‘most singularities remain intact despite some minor shifts’’. However, if such a deviation does not differ significantly from the smoothed annual variation of the given meteorological element, it is necessary to view it rather as part of the random fluctuations within the given time series that can be seen in any part of the year. Therefore the analysis of meteorological singularities requires the assessment of statistical significance (see, e.g., Bissolli and Sch€ onwiese, 1987, 1990; Bissolli, 1991; Bra´zdil et al., 1999). Although the present analysis has confirmed the existence of meteorological singularities in the annual weather pattern in the Czech Republic, their occurrence has to be approached critically and with considerable caution. It is clear from the present survey and its comparison with other works that meteorological singularities in Central Europe are not generally associated with an exact calendar date and that their occurrence may vary strongly both in time and space. Acknowledgements The authors would like to thank: the Grant Agency of the Czech Republic for financial support to the Grant No. 205=05=0858; also Dr. P. Bissolli (German Meteorological Service, Offenbach am Main) and Dr. M. Budı´kova´ (Masaryk University, Brno) for their advice during the preparation of the present study; Tony Long (Svinosˇice) for English style corrections. References Bayer K (1955) Singularity teploty na Milesˇovce v obdobı´ 1910–1939. Meteorol Zpr 8: 35–42 Bayer K (1956a) Vańocˇnı´ oblevy v obdobı´ 1905–1954. Meteorol Zpr 9: 8–15 Bayer K (1956b) Pravidelna´ ochlazenı´ na konci kveˇtna a v cˇervnu v obdobı´ 1906–1955. Meteorol Zpr 9: 97–105 Bayer K (1960) Witterungssingularit€aten und allgemeine Zirkulation der Erdatmosph€are. Geofyz Sbor 7: 521–634 Bissolli P (1991) Eintrittswahrscheinlichkeit und statistische Charakteristika der Witterungsregelf€alle in der Bundesrepublik Deutschland und West-Berlin. Ber Inst Met Geoph Univ Frankfurt=Main Frankfurt am Main, 88, 566 pp Bissolli P, Sch€onwiese C-D (1987) Singularit€aten in der Bundesrepublik Deutschland 1946–1986. Vorl€aufige Ergebnisse einer statistische Analyse anhand ausgew€ahlter Stationen. Meteorol Rdsch 40: 147–155

Bissolli P, Sch€ onwiese C-D (1990) Spatial and temporal variations of singularities in the Federal Republic of Germany 1881–1986. Meteorol Rdsch 42: 33–42 Bra´zdil R (1982) Precipitation singularities in the variation of diurnal sums of precipitation in the summer season on the territory of the Czechoslovak Socialist Republic (CSSR). Scripta Fac Sci Nat Univ Purk Brun 12 (Geographia): 169–202 Bra´zdil R, Sˇtekl J, Budı´kova´ M (1999) Poveˇtrnostnı´ singularity. In: Bra´zdil R, Sˇtekl J et al (eds) Klimaticke pomeˇry Milesˇovky. Praha: Academia, 237–257 Flohn H (1954) Witterung und Klima in Mitteleuropa. S. Hirzel Verlag, Stuttgart, 214 pp Flohn H, Hess P (1949) Grosswetter-Singularit€aten im j€ahrlichen Witterungsverlauf Mitteleuropas. Meteorol Rdsch 2: 258–263 Gerstengarbe F-W, Werner PC (1987) Ist der Baur’sche Kalender der Witterungsregelf€alle heute noch g€ ultig? Z Meteorol 37: 263–272 Katalog poveˇtrnostnıćh situacı´ pro uzemı´ CˇSSR. Praha , 94 pp 1967: HMU Katalog poveˇtrnostnıćh situacı´ pro uzemı´ CˇSSR. Praha , 38 pp 1972: HMU Krˇivancova´ S, Vavrusˇka F (1997) Za´kladnı´ meteorologicke prvky v jednotliv ych poveˇtrnostnıćh situacıćh na uzemı´ Cˇeske republiky v obdobı´ 1961–1990. Na´rodnı´ klimatick y program Cˇeska´ republika 27, Praha, 114 pp ˇ ˇ Loncar E, Sinik N, Zaninovic K (1996) Pressure singularity in climate dynamic studies. Zes Nauk Uniw Jagiell – Prace Geogr 102: 207–211 Munzar J (1985) Medardova ka´peˇ aneb pranostiky ocˇima meteorologa. Horizont, Praha, 240 pp Nosek M (1957) Sra´zkove singularity na Moraveˇ a ve Slezsku. Sbor Cˇeskosl Spol Zemeˇp 62: 18–27 Pleiss H (1961) Wetter und Klima des Fichtelberges. Abh Meteorol Hydrol Dienstes DDR 62 (Band VIII). Akademie-Verlag, Berlin, 323 pp Rˇeznıćˇkova´ L (2004) Poveˇtrnostnı´ singularity na uzemı´ Cˇeske republiky. Diplomova´ praće, Geografick y ustav PrˇF MU, Brno, 74 pp Schmauss A (1928) Singularit€aten im j€ahrlichen Witterungsverlauf in M€ unchen. Dtsch Met Jahrb Bayern, Anhang B Schmauss A (1938) Synoptische Singularit€aten. Meteorol Z 55: 385–403 Souborna´ studie. Podnebı´ Cˇeskoslovenske socialisticke republiky. Praha 1969: Hydrometeorologick y ustav, 357 pp Stuchlı´k F (1960) Struktura rocˇnı´ho chodu sra´zek podle stodvacetilete rˇady Praha-Klementinum. Meteorol Zpr 13: 137–140 Zikmunda O (1954) Singularity sra´zek na Milesˇovce. Meteorol Zpr 7: 86–87 Author’s address: Prof. Rudolf Bra´zdil (e-mail: [email protected]), Institute of Geography, Faculty of Science, Masaryk University Brno, Kotlarˇska´ 2, CZ-61137 Brno, Czech Republic.

Příloha č. 5

Rudolf Brázdil – Rostislav Krušinský – Ladislava Řezníčková

Zprávy o počasí z let 1655-1656 v deníku Jana Františka Bruntálského z Vrbna1 Úvod Osobní deníky často obsahují kromě řady nejrůznějších informací také stručné popisy počasí či poznámky ve vztahu k němu. Z tohoto pohledu je lze proto považovat za jeden z významných historicko-klimatologických dokumentárních pramenů, důležitých pro poznání klimatu v období před začátkem systematických meteorologických pozorování.2 Platí to i o českých zemích, kde několik takových deníků bylo hodnoceno právě se zřetelem na meteorologické informace v nich obsažené.3 Jejich řadu rozšiřuje také deník Jana Františka Bruntálského z Vrbna,4 vedený od 1. ledna 1655 do 4. března 1656, jehož meteorologická analýza je předmětem tohoto příspěvku. Jan František Bruntálský a jeho deníkové záznamy Jan František Bruntálský z Vrbna se narodil dne 3. listopadu 1634. Svůj deník začal psát na prahu zletilosti od 1. ledna 1655 jako budoucí fulnecký pán, neboť byl jediným synem Václava Bruntálského, zesnulého v roce 1649, a Alžběty Polyxeny Dembinské. Kromě Fulneku patřily k jeho panství také Paskov a Studénka a v Čechách Mrač. Jako většina příslušníků vyšší šlechty té doby budoval svou kariéru službou v zemské a dvorské správě a 1

Studie byla vypracována s finanční podporou Grantové agentury České republiky pro řešení grantu 205/05/0858 „Časová a prostorová variabilita klimatu České republiky podle denních vizuálních záznamů počasí a přístrojových měření“. Za cenné připomínky k textu příspěvku stejně jako za poskytnutí informací z pamětní knihy Velvar patří poděkování PhDr. Oldřichu Kotyzovi z Oblastního muzea v Litoměřicích. Zemskému archivu Opava a Archivu města Brna jsme povinováni díkem za možnost publikovaní obr. 1-3 z jejich archivních pramenů. 2 Brázdil, Rudolf, Pfister, Christian, Wanner, Heinz, von Storch, Hans, Luterbacher, Jürg: Historical climatology in Europe – the state of the art. Climatic Change 70, 2005, s. 363-430. 3 Viz např. Pejml, Karel, Munzar, Jan: Matyáš Borbonius z Borbenheimu a jeho meteorologická pozorování z let 1596-1598, 1622. Meteorologické zprávy 21, 1968, s. 93-95. Brázdil, Rudolf, Kotyza, Oldřich: Denní záznamy počasí v Praze v letech 1649-1650. Meteorologické zprávy 48, 1995, s. 109-111. Titíž: Daily meteorological observations of Charles Senior of Žerotín in the years 1588–1591. Scripta Facultatis Scientarum Naturalium Universitatis Masarykianae Brunensis 25 – Geography, 1995, s. 7-39. Munzar, Jan: Meteorologická pozorování Karla ze Žerotína z let 1588-1589 a 1591. Meteorologické zprávy 49, 1996, s. 58-61. Brázdil, Rudolf, Kotyza, Oldřich: History of Weather and Climate in the Czech Lands II. The earliest daily observations of the weather in the Czech Lands. Brno 1996, 177 s. Titíž: History of Weather and Climate in the Czech Lands III. The earliest daily observations of the weather in the Czech Lands II. Brno 1999, 228 s. Titíž: Meteorologické záznamy děkana Bartoloměje Michala Zelenky z Čech z let 1680-1682, 1691-1694 a 1698-1704. Meteorologické zprávy 54, 2001, s. 145-155. Brázdil, Rudolf, Valášek, Hubert, Macková, Jarmila: Climate in the Czech Lands during the 1780s in light of the daily weather records of parson Karel Bernard Hein of Hodonice (southwestern Moravia): Comparison of documentary and instrumental data. Climatic Change 60, 2003, s. 297-327. 4 K jeho podrobnému zhodnocení viz Krušinský, Rostislav: Deník Jana Františka Bruntálského z Vrbna. Diplomová práce FPF SU. Opava 2004, 258 s.

stal se osobností s důležitým společenským postavením.5 Postupně zastával významné úřady: dvorský sudí (1685-1688), zemský sudí (1688-1689), nejvyšší hofmistr (1689-1700), nejvyšší kancléř (1700-1705). Zároveň byl také císařským radou a komořím, stejně jako rytířem řádu zlatého rouna.6 Ze dvou manželství s Eliškou z Martinic a posléze s její sestrou Terezií Františkou zanechal po sobě Jan František Bruntálský 4 syny: Václava Bernarta, Jana Antonína, Josefa Františka a Norberta Václava.7 Zemřel dne 22. srpna 1705. Bruntálského deník, uložený v Zemském archivu v Opavě,8 je sepsán ve stostránkovém svázaném sešitu o rozměrech 15 x 20 cm, opatřeném pergamenovou vazbou. Na jeho titulní stránce (obr. 1) následuje za nadpisem jeho přiřazení vlastníkovi: „Deník roku 1655 L.P., který jsem začal psát já Jan František z Vrbna a Bruntálu atd., hrabě svaté říše římské, prvního ledna roku 1655.“9 Poté se objevuje stručný úvod hovořící o tom, že deník začal Jan František Bruntálský psát po návratu z kavalírské cesty.10 Text je doplněn rodovým erbem Bruntálského. Vlastní deníkové záznamy začínající na následující stránce jsou psány drobným písmem v té době běžným duběnkovým inkoustem, přičemž jejich občasná nečitelnost je způsobena vodou či snahou autora vměstnat záznam na již plnou stránku. Jazykem deníku je latina, velmi zřídka doplněná češtinou, němčinou či francouzštinou, a to jen v případě názvů knih, měst, nemocí či her. Ze struktury textu vyplývá, že záznamy byly vedeny pravidelně, a to pokud ne každý den, pak jen s minimálním denním intervalem. První zápis byl proveden 1. ledna 1655 a 4. březen následujícího roku je datem posledního uceleného záznamu. Po něm následuje kratičký zápis, že v onom měsíci se již nic zvláštního nestalo a že autor šťastně dorazil dne 8. března 1656 do Vídně. Poslední řádky tvoří nadpis začínajícího měsíce „Mensis Aprilis“ a datum prvního dne v měsíci. Tím také celý deník končí. Možnost případných pokračování deníku v dalších knihách, jak tomu někdy bývá u jiných pramenů tohoto typu, lze zřejmě vyloučit vzhledem ke způsobu ukončení a k několika volným listům na konci díla. 5

Turek, Adolf: Fulnecko. Brno 1940, s. 96n. Kolektiv autorů: Biografický slovník Slezska a Severní Moravy. Opava 1998, sešit 10, s. 45. 6 Ke kariéře aristokratů blíže Maťa, Petr: Svět české aristokracie (1500-1700). Praha 2004, od s. 328, k Bruntálskému podrobněji s. 815. 7 Turek, A.: Fulnecko, s. 97. 8 Zemský archiv Opava. Velkostatek Fulnek, inv. č. 1. 9 „Diarium Anni 1655 A.D. A.ME. Joanne Francisco S.R.I. Comite de Wirbna, Et Fraidenthal etc; Prima die Januarij Anni 1655 Conscribi Ceptum.“ 10 „Cum praeteriti Anni, 12 die Mensis Octobris, ex Externis Provintijs DEO O: M: Sint Laudes, feliciter Ollomutium ad Illustrissimam Dominam Matrem Redijssem, una cum Domino Daniele Crasny de Löwenfeld, qui meus Prefectus in Externis Provintijs fuit; Et tertia die Novembris Anni 1654 Vigesimum vitatis Annum implluissem.“ (Když jsem se 12. dne měsíce října minulého roku vrátil z ciziny, chvála Bohu nejlepšímu a největšímu, společně s Panem Danielem Crasnym z Löwenfeldu, který byl v cizině mým velitelem, šťastně do Olomouce k nejjasnější paní matce; a třetího listopadu roku 1654 jsem zcela dokonal 20. rok života.)

Zápisy Jana Františka Bruntálského jsou na první pohled formou i obsahem velice podobné. Začínají uvedením dne v měsíci, pokračují astronomicko-astrologickou značkou pro daný den (pondělí – Β – dies lunae, úterý – Φ – dies martiis, středa – ∆ – dies mercurii, čtvrtek – Γ – dies jovis, pátek – Ε – dies veneris, sobota – Η – dies saturnae, neděle – Α – dies solis), zápisem o účasti na mši a obvykle končí údajem o povětrnostní situaci toho kterého dne (obr. 2). Kromě těchto pravidelných údajů autor popisuje především události společenského života. Více než 200 osob z těch, které Bruntálský potkal, je zmíněno v jeho deníku. Jsou uváděny v souvislosti s nejrůznějšími společenskými událostmi jakými byly lov, tanec nebo hry. Zatímco první dvě aktivity byly poměrně sporadické, aktérem nejrůznějších her se Jan František stával prakticky každý druhý den. Udává takové hry jako mlýnek, vrhcáby neboli backgammon, kuželky nebo pálkovanou, což byla předchůdkyně dnešního tenisu.11 Zdaleka nejoblíbenější ovšem byly karty v nejrůznějších podobách – v deníku je zmíněno asi 11 druhů her s nimi. Četnost zápisů týkajících se nejrůznějších karetních her jen dokládá oblíbenost této po staletí marně zakazované zábavy. V souvislosti s hrou se Jan František často rozepsal o finančních záležitostech a takřka pravidelně zaznamenával své zisky a ztráty.12 Pokud jde o finance, vyskytují se i poznámky o koupi knih a koní.13 Ze soukromějších záležitostí se v deníku zmiňuje, kromě popisu účasti na bohoslužbách, především četba. Autor uvádí více než 15 děl, mezi nimiž nechybějí klasici antické literatury jako Ovidius či Vergilius, ale ani moderní autoři 16. a 17. století jako Torquato Tasso nebo Martin Opitz.14 Vedle toho věnoval mladý Jan František pozornost také odbornější literatuře, ať již jde o učebnici šermu od Salvatora Fabrise nebo o knihu Octavia Strady zabývající se emblematikou. Z četby je zřejmá Bruntálského znalost latiny, češtiny, němčiny a francouzštiny. Pisatel v deníku rovněž zmiňuje svoji překladatelskou činnost, zřejmě šlo však spíše o aktivitu v rámci studia cizího jazyka než o cílevědomou práci na překladu celého díla. V zápiscích se objevují také zmínky o nemocech Jana Františka i o

11

Deník k 17. září 1655, s. 59: „Sumpsit apud nos prandium Dominus Christophorus Skrbenski, cum quo post prandium pyramidibus, et deinde Zik Zak lusi.“ (Obědval u nás pan Kryštof Skrbenský, se kterým jsem po obědě hrál pyramidy a potom Zik-zak.) 12 Deník k 22. říjnu 1655, s. 66: „ … lusimus vsque post horam quintam vespertinam Landsknecht, perdidiq[ue] viginti quatuor Ducatos.“ (…hráli jsme až do páté hodiny odpolední landsknecht a prohrál jsem 24 dukátů.); viz také Deník k 25. říjnu 1655, s. 66: „…cum quo Landtsknecht lusi, lucratusq(ue) sum quinquaginta sex ducatos.“ (... se kterým jsem hrál landsknecht a vyhrál jsem 56 dukátů.) 13 Deník k 28. dubnu 1655, s. 30: „Emi a Capitaneo Staudinge Paulo Schetzl dicto, Equum coloris rubei, qui Germanice fuchsfarb dicitur pro viginti octo Imperialibus.“ (Koupil jsem od hejtmana Studénky řečeného Paulus Schetzl koně načervenalé barvy, která se německy nazývá fuchsfarb, za 28 císařských.) 14 Deník k 17. březnu 1655, s. 19: „Legi mane in Torquato Tasso.“ (Četl jsem si ráno v Torquatu Tassovi.)

nemocech jeho blízkých, stejně jako o jejich léčení soudobými metodami jako bylo např. u lékařů oblíbené pouštění žilou nebo užívání projímadel.15 V deníkových záznamech se často odráží i soudobé dění. Nejinak je tomu v případě Jana Františka, který si nechával posílat novinky hlavně z Itálie a Francie. V deníku se tak zmiňuje i o vojenských akcích Francie, Španělska a italských vévodství. Velký prostor věnuje úmrtí papeže Inocence X.,16 následnému konkláve a volbě Alexandra VII., stejně jako úmrtí Eleonory, císařovny vdovy. Z pohledu meteorologické interpretace Bruntálského záznamů je méně příznivá skutečnost, že se jeho zápisy vztahují k různým místům, kde momentálně dlel. Tak začátkem roku 1655 pobýval v Paskově, pak tři dny ve Studénce a od 17. ledna do 13. dubna ve Fulneku (obr. 3). Zbytek dubna 1655 pak strávil opět v Paskově a ve Fulneku, aby odtud odcestoval přes Olomouc a Vyškov do Vídně, kam dorazil dne 9. května. Z Vídně směřovaly jeho kroky na několik dnů nejprve do Bratislavy a pak i do Brna, ale od 8. června až do 5. července pobýval v Badenu. Odtud se vrátil opět do Vídně, kde strávil čas až do 14. prosince, aby opět cestoval na Moravu. Mezi místy jeho pobytu se potom objevuje zejména Brno (16.24. prosince, 29. prosince 1655-19. ledna 1656, mezitím Letovice), odkud se pak přes Letovice (20.-25. ledna) a Pardubice vydal do Prahy, kde pobýval od 28. ledna do 4. března 1656, kdy odjel opět do Vídně. Průběh počasí od ledna 1655 do února 1656 podle Bruntálského záznamů a jejich meteorologické interpretace Terminologie použitá Janem Františkem Bruntálským pro popis počasí v jeho deníku se vztahuje k různým meteorologickým prvkům a jevům. Pro popis teploty vzduchu použil následující termíny: frigus (mráz) – magnum (velký), horendum (hrozný), multum (silný), extraordinarius (mimořádný), gelavit (mrzlo); degelauit (tálo); dies (den) – calida (teplý), calidissima (velmi teplý, horký), temperata (mírný), hyemalis (zimní), ut initio veris esse solet (jaký bývá na počátku jara), aestiua (letní), frigida (studený); Sreckhitz (strašné horko). Srážkové poměry charakterizuje Bruntálský pojmy: nix (sníh) – ninxit (sněžilo), pluit (pršelo). V řadě případů uvedl pisatel pocitovou a percepční charakteristiku celého dne: dies (den) – lepida (pěkný), pulcherima (krásný), amoena (příjemný), melancolica (melancholický, 15

Deník k 22. lednu 1655, s. 6: „Purgarent me sat bene Pillulae, quas heri hora sexta post meridiem sumpsi.“ (Velmi dobře mne vyčistily pilulky, které jsem si vzal včera v šest hodin odpoledne.) 16 Deník k 1. únoru 1655, s. 8: „Accepi noua ex Italia, que mihi omni septimana mitti solent, que Papam Innocentium X ex familia Pamphiliana septima die Praeteriti Mensis, Exspirasse docent.“ (Dostal jsem novinky z Itálie, které jsou mi posílány každý týden, a které zpravují o úmrtí papeže Inocence X. z rodiny Pampfili, sedmý den minulého měsíce [7. ledna 1655].)

zádumčivý), turpis, turpissima (škaredý), tristis (pošmourný, nevlídný), inconstans (nestálý); tempus (počasí) – obscurum (temné), fastidiosum (odporné). Z ostatních projevů počasí věnoval pozornost zejména větrným poměrům: ventus (vítr) – magnus (velký), maximus (hrozný, obrovský), terribilis (strašný), extraordinarius (mimořádný), frigidissimus (ledový), frigidus (studený, mrazivý), lepidus (příjemný); ventosa dies (větrný den); nullus ventus (bezvětří). Ve vztahu k oblačnosti lze interpretovat pro jednotlivé dny pojmy serena (jasný) a obscura (zamračený). Vedle toho uvedl termíny charakterizující dohlednost jako nebulosa dies (zamlžený den) a nebula (mlha) – magna (velká), densissima (hustá). Leden 1655 patřil podle Bruntálského záznamů (obr. 4) k výrazněji chladnějším měsícům. Mrazivé a studené počasí panovalo do 19. ledna, kdy nastoupila obleva (20. leden označil jako krásný den, jaký bývá na počátku jara17), ale 24. ledna uhodil již opět hrozný mráz. Dne 14. ledna zaznamenal strašlivý vítr, přičemž 10. a 31. ledna celý den sněžilo. Další měsíc únor měl vcelku charakter teplotně průměrného měsíce, kdy mrazivé počasí v jeho první části bylo vystřídáno oteplením od poloviny měsíce. Tak po celodenním sněžení a silném větru dne 7. února a silných mrazech ve dnech 8.-13. února následující den trochu pršelo a přišlo náhlé tání, kdy během jediné noci ze 14. na 15. února povolil led na řekách. Druhá polovina měsíce byla spíše již mírnější. V souladu

s Bruntálského

zápisy

jsou

dobové

zprávy

o

počasí

a

hydrometeorologických extrémech z českých zemí, z nichž vynikají zejména informace o katastrofální povodni z poloviny února.18 Předpokladem pro její vznik byla předchozí tuhá zima s množstvím sněhu, kterou zmiňují např. prameny pro Čáslav,19 Žatec20 či Velvary.21 Kněz řádu křížovníků s červenou hvězdou Jan Beckovský uvádí tuhé mrazy zejména mezi 9. a 13. únorem, kdy zmrzlo na 20 osob.22 V Praze na Vltavě dosahoval led dokonce tloušťky jednoho lokte.23 Náhlé tání po silném teplém větru dne 14. února vyvolalo již následující den 17

Deník, k 20. lednu 1655, s. 5: „Fuit pulcherrima, et ut initio veris esse solet, Dies.“ Brázdil, Rudolf, Dobrovolný, Petr, Elleder, Libor, Kakos, Vilibald, Kotyza, Oldřich, Květoň, Vít, Macková, Jarmila, Müller, Miloslav, Štekl, Josef, Tolasz, Radim, Valášek, Hubert: Historické a současné povodně v České republice. Brno, Praha 2005, s. 213-214. Např. v Praze na Vltavě se jednalo o nejvýznamnější povodeň 17. století, která způsobila velké škody. 19 Archiv města Ústí nad Labem, Sbírka rukopisů, bez signatury, Letopisecké záznamy Jana Čeledínka z Čáslavi připsané k Veleslavínovu Kalendáři historickému z r. 1590, s. 84. 20 Katzerowsky, Wenzel: Meteorologische Aufzeichnungen aus Saaz. Mitteilungen des Vereins für Geschichte der Deutschen in Böhmen 21, 1883, s. 346. 21 Ve Velvarech k množství sněhu z ledna připadlo mnoho dalšího dne 12. února, ale následující den již hojně pršelo (Státní okresní archiv Kladno. AM Velvary, inv. č. 5, Knihy pamětní věcí hodnejch, obzvláštních i obecních, f. 88r-v). 22 Rezek, Antonín: Poselkyně starých příběhův českých. Sepsal Jan Beckovský, kněz řádu Křížovníků s červenou hvězdou. Díl druhý (Od roku 1526-1715). Sv. třetí (L. 1625-1715 i s dodatky). Praha 1880, s. 430. 23 1 loket český = 59,27 cm (viz Hofmann, Gustav: Metrologická příručka pro Čechy, Moravu a Slezsko do zavedení metrické soustavy. Plzeň, Sušice 1984, s. 71). 18

(15. února) a v dalších dnech odchod ledu a rozvodnění mnoha řek v celých Čechách.24 Na Moravě je zmínka o uvedené povodni v análech brněnského zábrdovického kláštera, kdy rozvodněná Svitava zaplavila kostel a další stavby, odnesla most, utopila na klášterních statcích 23 kusů dobytka a způsobila mnoho škod na dalším majetku.25 Ve východních Krkonoších (Rýchory – Sklenařovice) uvolnilo náhlé oteplení lavinu, která si vyžádala celkem osm obětí.26 Po deštích začátkem března se ochladilo a sněžilo ve dnech 7. a 9. března, ale 13. března celý den tálo. Nato podle Bruntálského bylo po tři dny opět chladno a sněžilo, ale od 18. března již nastoupilo příznivější počasí, které přes několik větrných a chladnějších dnů pokračovalo i v dubnu.27 V květnu pisatel uvádí převážně příjemné a teplé dny, přerušené deštivým počasím mezi 25. a 28. květnem. Dne 23. května pozoroval ve Vídni patrně polární záři: „Večer padala záře, která vydržela po celou noc.“28 První červnová dekáda byla podle Bruntálského zápisků spíše zamračená a zčásti i s dešti. V dalším průběhu měsíce se střídaly zamračené dny29 s příznivějším a teplejším počasím, které nastoupilo od 22. června a trvalo do 5. července.30 Poté následovala opět perioda zamračeného a deštivějšího počasí,31 přičemž přes několik dalších zamračených dnů, a to i s deštěm, označil Bruntálský většinu dnů za příjemné.32 Rovněž v srpnu měly

24

Podrobné zhodnocení této povodně v Čechách s uvedením literárních a archivních pramenů viz pozn. 18. Moravský zemský archiv Brno. G 10 Sbírka rukopisů Zemského archivu 1200-1999, inv. č. 296/1, Anály zábrdovického premonstrátského kláštera 1210-1673, f. 267v, 268r, 269v. 26 Vicena, Ivo, Pařez, Jan, Konôpka, Jozef: Ochrana lesa proti polomům. Praha 1979, s. 193. 27 Dne 8. dubna 1655 je Bruntálského rukopis na místě informace o počasí nečitelný, takže v obr. 4 je pro tento den uveden chybějící záznam. 28 Deník, k 23. květnu 1655, s. 35: „Cecidit versus noctem Iuber, qui pernocte durauit.“ V katalogu polárních září pozorovaných jižně od 55. stupně severní zeměpisné šířky je však pro rok 1655 uveden pouze případ z 10. července (Křivský, Ladislav, Pejml, Karel: Solar activity, aurorae and climate in Middle Europe in the last 1000 years. Travaux Géophysiques 33, 1985, s. 116). 29 Dne 15. června při pobytu v Badenu charakterizoval Bruntálský počasí jako nestálé. Jan Čeledínek si k témuž dni zapsal v Čáslavi bouřku, liják a povodeň (Archiv města Ústí nad Labem. Sbírka rukopisů, bez signatury, Letopisecké záznamy Jana Čeledínka, s. 325). Pamětní kniha Velvar cituje pro týž den velkou vodu z dešťů (Státní okresní archiv Kladno. AM Velvary, inv. č. 5, Knihy pamětní, f. 89r). 30 Dne 29. června, který Bruntálský charakterizoval v Badenu jako velmi teplý, způsobila bouřka doprovázená krupobitím četné škody v jižních Čechách (Mareš, František (ed.): Kronika Budějovská. Věstník Královské české spol. nauk za r. 1920, Praha 1922, s. 77), v Čáslavi (Archiv města Ústí nad Labem. Sbírka rukopisů, bez signatury, Letopisecké záznamy Jana Čeledínka, s. 358) a na panstvích Křešice, Ploskovice, Třebušín a Stvolínky (Anonym [Ankert, Heinrich]: Brände und Wetterschäden. Unsere Heimat (Beilage zur Leitmeritzer Zeitung) 14, 1933, nestr.). 31 Zatímco Bruntálský zaznamenal při svém pobytu v Badenu deště pro dny 5.–6. července a ve Vídni 9.–11. července, Jan Čeledínek ve svých záznamech z Čáslavi cituje k 11. červenci bouřku, místy silné lijáky a povodeň se škodami na Sázavě a Doubravě (Archiv města Ústí nad Labem. Sbírka rukopisů, bez signatury, Letopisecké záznamy Jana Čeledínka, s. 379). K témuž dni si zapsal povodeň se škodami také kronikář v Českých Budějovicích (Mareš, F. (ed.): Kronika Budějovská, s. 77-78). 32 Den 28. července charakterizoval Bruntálský ve Vídni jako zamračený a studený, kdy po několik hodin pršelo. Na Litoměřicku nadělal téhož dne škody silný liják (Anonym [Ankert, H.]: Brände und Wetterschäden, nestr.). 25

převažovat teplé a příjemné dny nad dny se zamračeným a deštivým počasím. Přesto k 6. srpnu uvedl Bruntálský dokonce „zamračený a přímo zimní den“.33 V září převládalo celkově příjemné teplejší počasí, ale od 25. září se ochladilo. Říjen začal zamračeným, studeným a deštivým počasím, ale již k 5. říjnu si Bruntálský poznamenal „příjemný den“ (Fuit amoena dies). V dalším průběhu měsíce se střídaly příjemnější teplejší dny se zamračenými, chladnějšími, mlhavými a deštivějšími. V listopadu již převažovaly zamračené a studenější dny. Stejné počasí pokračovalo i začátkem prosince, kdy navíc i sněžilo, popř. padal déšť se sněhem. Mrazy trvající od 8. do 19. prosince byly vystřídány krátkou oblevou, ale ve dnech 22.-23. prosince v noci znovu mrzlo. Nato nastoupilo opět zamračené a mírné počasí. Tento ráz počasí ukazuje na typickou povětrnostní singularitu vyskytující se ve střední Evropě a známou pod názvem „vánoční obleva“.34 Velké mrazy opět uhodily od 29. prosince, ale pokračovaly jen do 2. ledna 1656. Při pobytu v Brně si Jan František Bruntálský k 31. prosinci dokonce poznamenal „hrozný mráz, jaký tento celý rok nebyl“ a k následujícímu dni 1. ledna „hrozný a naprosto mimořádný mráz“.35 O mrazivém počasí je opětně zmínka dne 8. ledna a poté až od 18. ledna. V dalším měsíci mrzlo podle Bruntálského záznamů patrně před 12. únorem, neboť pisatel konstatuje, že bylo zamračeno a „téměř celý den tálo“.36 Mrazy ale výslovně zmiňuje jen pro 16.-17. únor a sněžení pro dny 14. a 18. února. Využití záznamů Jana Františka Bruntálského pro interpretaci teplotního a srážkového charakteru daného měsíce tak, jak je to u denních záznamů počasí obvyklé,37 je ale poměrně obtížné. Záznamy se totiž v řadě případů omezují jen na obecnou charakteristiku počasí, z níž není zcela jasný teplotní charakter daného dne. Ještě horší je situace v případě srážek, kdy Bruntálským uváděné počty srážkových dnů jsou velmi nízké (pouze v rozmezí 2 až 10 dnů za měsíc) nebo chybějí úplně (listopad 1655). Zejména menší srážkové úhrny a noční srážky tak zůstávaly nepochybně stranou jeho pozornosti. Porovnání interpretovaných teplotních záznamů s Německem38 a se Švýcarskem39 plyne z tab. 1. Shoda v interpretaci zpravidla

33

Deník, k 6. srpnu 1655, s. 51: „Fuit obscura, et plane hyemalis dies.“ Řezníčková, Ladislava, Brázdil, Rudolf, Tolasz, Radim: Meteorological singularities in the Czech Republic in the period 1961-2002. Theoretical and Applied Climatology 88, 2007, s. 179-192. 35 Deník, k 31. prosinci 1655, s. 80: „Fuit frigus maximum, quale toto hac Anno non fuit.“ Deník, k 1. lednu 1656, s. 81: „Fuit serena, sed terribile, et plane extraordinaria frigus.“ 36 Deník, k 12. únoru 1656, s. 89: „Fuit obscura dies, degelauitq[ue] quasi tota die.“ 37 Viz např. Brázdil, R., Pfister, C., Wanner, H., von Storch, H., Luterbacher, J.: Historical climatology, s. 363430. Kvalitativní popisy počasí jsou interpretovány v podobě vážených měsíčních teplotních indexů: 3 – extrémně teplý, 2 – velmi teplý, 1 – teplý, 0 – normální, -1 – studený, -2 – velmi studený, -3 – extrémně studený měsíc. 38 Glaser, Rüdiger: Klimageschichte Mitteleuropas. 1000 Jahre Wetter, Klima, Katastrophen. Darmstadt 2001, 227 s. 34

nastává buď s jednou nebo s druhou řadou indexů, přičemž zjevný nesoulad je patrný v interpretaci dubna 1655 a ledna 1656, kde Bruntálského záznamy nevypovídají o chladnějším rázu těchto měsíců. Závěr Záznamy o počasí nepochybně nebyly pro Jana Františka Bruntálského hlavním objektem jeho deníku, ale spíše dokreslovaly celkový kontext ostatních zpráv v jednotlivých dnech. Přestože v řadě dnů je jejich vypovídací meteorologická hodnota poněkud omezená, neztrácejí svůj význam jako historicko-klimatologický zdroj informací z let, která jsou z hlediska zpráv o počasí v ostatních dokumentárních pramenech v českých zemích také poměrně chudá. Bruntálského deník je tak nejen významným kulturně-historickým pramenem pro poznání dějin každodennosti raně novověkého aristokrata, ale i pro rekonstrukci počasí v oblasti střední Evropy. Tab. 1. Interpretace měsíčních teplotních indexů pro leden 1655 – únor 1656 podle deníkových záznamů Jana Františka Bruntálského (JFB), pro Německo (Ně)40 a Švýcarsko (Šv)41 I

II

III

IV

V

VI

VII

VIII IX

X

XI

XII

I

II

JFB -1

0

0

0

1

0

0

1

1

0

-1

-1

0

0

Ně

-1

0

0

-1

1

-1

0

1

0

1

0

-1

-1

-1

Šv

-

1

0

-2

0

0

0

0

0

0

-1

0

-3

0

Abstrakt Jan František Bruntálský z Vrbna (1634-1705), pán na Fulneku a majitel Paskova, Studénky a Mrače, patřil mezi významné představitele vyšší šlechty, zastávající řadu funkcí v zemské a dvorské správě. Jeho deník, psaný od 1. ledna 1655 do 4. března 1656, obsahuje vedle informací o jeho každodenních činnostech také denní charakteristiku počasí. Z ní lze získat cenné meteorologické informace ve vztahu k teplotě vzduchu, srážkám, větru, oblačnosti a dohlednosti. Podle těchto zápisků lze popsat průběh počasí ve výše uvedeném období a konfrontovat ho s dalšími prameny o počasí a příbuzných jevech z českých zemí. Jistou 39

Pfister, Christian: Klimageschichte der Schweiz 1525-1860. Das Klima der Schweiz von 1525-1860 und seine Bedeutung in der Geschichte von Bevölkerung und Landwirtschaft. Bern, Stuttgart 1988, 184 a 163 s. 40 Glaser, R.: Klimageschichte Mitteleuropas. 41 Pfister, C.: Klimageschichte der Schweiz, Tabelle 1/30.

nevýhodou je, že Bruntálského pozorování se nevztahují pouze k jednomu místu. Nejvýznamnějším extrémem analyzovaného období byla ničivá povodeň v polovině února 1655, o níž se však pisatel deníku nezmiňuje. Záznamy Jana Františka Bruntálského umožňují sestavit řadu vážených teplotních indexů, které se poměrně dobře shodují s indexy odvozenými dříve pro území Německa nebo Švýcarska. Analyzovaný deník je významným zdrojem informací pro rozšíření poznatků o průběhu počasí z let 1655-1656 v českých zemích a ve střední Evropě a příspěvkem do mozaiky počasí v období před začátkem systematických meteorologických pozorování. Klíčová slova: Jan František Bruntálský – osobní deník – záznamy o počasí – teplotní indexy

Obr. 1. Titulní list deníku Jana Františka Bruntálského z Vrbna

Obr. 2. Ukázka zápisů z přelomu měsíců dubna a května roku 1655 v deníku Jana Františka Bruntálského z Vrbna

Obr. 3. Veduta města Fulneku od Gottfrieda Herberta z roku 1720.42 Fulnek patřil k rodovémuvlastnictví Bruntálských z Vrbna v letech 1622-1788

42

Archiv města Brna. U 21 Knihovna Mitrovského – sbírka rukopisů, sg. 10/29, Veduty Josefa Ignáce Dismase Hofera z první poloviny 18. století.

Obr. 4. Meteorologická interpretace denních záznamů počasí v deníku Jana Františka Bruntálského z Vrbna z období leden 1655 – březen 1656: 1 – jasno, 2 – polojasno, 3 – zataženo, 4 – déšť, 5 – dešťová přeháňka, 6 – sněžení, 7 – vítr, 8 – studený vítr, 9 – mlha, 10 – mlhavo, vlhko, b – bezvětří, H – horko, W – teplo, C – chladno, F – mráz, M – mírně, X – údaj chybí. Podtržený symbol značí silnou intenzitu jevu a lomítko charakterizuje příslušnou část dne s výskytem jevu (ráno a dopoledne, resp. odpoledne, večer či v noci)

Obr. 4. Pokračování

Příloha č. 6

Weather information in the diaries of the Premonstratensian Abbey at Hradisko (nowadays Olomouc), in the Czech Republic, 1693–1783• 1

O O PR D TE

EC

Figure 1. Location of the Hradisko Abbey (1), Svatý Kopeček (2) and Stará Voda (3) in the Czech Republic.

R

source of weather information as they cover the immediately preceding period 1693–1783, although with some gaps. Their meteorological analysis is the object of the present article.

The diaries of the Hradisko Premonstratensian Abbey

U

N

C

O R

Various documentary sources can supplement natural proxy data by gathering information about the weather and climate before the beginning of systematic instrumental meteorological observations (Brázdil et al., 2005b). Particularly important among these sources are daily visual weather records, which are available in the Czech Lands from the first half of the sixteenth century onwards (Brázdil and Kotyza, 1996). These consist largely of regular qualitative records, taken daily for a number of reasons by professionals of several kinds, such as astronomers, physicians, priests, farmers and others. Many such records overlap with regular instrumental measurements (Brázdil et al., 2003). However, since they are associated with a certain period in the life of any given author, they usually cover periods measured in terms of only a few years to decades. In the Czech Lands, the first preserved instrumental records date from 21 December 1719 to 31 March 1720 and come from Zákupy (north Bohemia). They were published in volumes of the so-called ‘Breslau Network’ (Brázdil and Valášek, 2002). Further known observations come from the Prague Jesuit college of St Clement (shortly called Klementinum) for 1752 and from Telč (south-west Moravia) for the period 1771–1775 (Brázdil et al., 2002b). However, systematic and continuous meteorological observations began only as late as 1775 (air temperature at Prague-Klementinum – Pejml, 1975) and 1803 (precipitation in Brno – Brázdil et al., 2005c). Hence information about the climate of the Czech Lands before these years is very limited. From these reasons, the weather records contained in the diaries of the Premonstratensian Abbey at Hradisko (nowadays Olomouc) and the residence at Svatý Kopeček (Holy Hill – Figure 1) in central Moravia, are an especially valuable

FS

Institute of Geography, Masaryk University of Brno 2 Moravian Land Archives, Brno

Weather – Month 9999, Vol. 99, No. 99

Rudolf Brázdil,1 Tomáš Černušák2 and Ladislava Řezníčková1

AQ1

The Hradisko Abbey was founded in 1077 AD by the Olomouc prince, Oto Sličný (Otto the Fair) on a small hill not far from the River Morava, and populated by Benedictine monks (Foltýn, 2005). In the mid-12th century the Benedictine order was replaced by Premonstratensians at the instigation of Jindřich Zdík, Bishop of Olomouc. The Abbey’s economy and its property prospered until the Hussite wars (1420–1434), when many of its estates were plundered and the Abbey itself was looted. In the sixteenth century, the Abbey struggled for independence from the City of Olomouc at the same time as suffering from poor management by its abbots and the consequences of the spread of Protestantism. In 1642, during the Thirty Years’ War (1618–1648),

it was occupied and devastated by Swedish soldiers. In the latter half of the seventeenth and eighteenth centuries the Abbey recovered and flourished once more. The ostentatious construction of the Abbey and its church was completed (Figure 2) and a number of parish complexes, churches and agricultural buildings sprang up in the neighbourhood. The abbots assembled a rich library and a picture collection including work by important artists. On Svatý Kopeček, a nearby hill, the Abbey turned a small chapel into a pilgrimage church dedicated to the Virgin Mary and it became one of the most important pilgrimage sites in Moravia. The Premonstratensian residence there assumed the status of a priory in 1710 (Smejkal and Hyhlík, 1994) and served for some years as the site for weather observations. Despite being the beneficiary of support for the enlightened reforms of Empress Maria Theresa and her son Joseph II in the latter half of the eighteenth century, the Hradisko Abbey was dissolved in 1784. Its buildings served briefly as a seminary, then found use as a military hospital, which operates there to this day.

1

FS

PR

O

O

Weather information in diaries of Czech Abbey Weather – Month 9999, Vol. 99, No. 99

Figure 2. The Hradisko Abbey from the north-east, circa 1740, in a vista by Friedrich Bernhard Werner (Fiala, 2003).

The diaries of the Hradisko Abbey for 1693–1783 have largely survived the vicissitudes of this history (Oppeltová, 1999). The lack of preface or introduction for the earliest part, covering 1693, seems to imply that previous entries remain to be discovered, or have not survived. All of the available records are largely concerned with the everyday events of abbey life. They consist of 39 abbey diaries covering a total of 50 years. Reliable collateral evidence points to the existence of a further eight diaries, and it is probable that they constituted a continuous series. The character of the manuscripts indicates contributions from 51 writers, although some worked only as copyists; 22 of them can be identified by name. The Abbye’s diaries are partly supplemented by a series from the residence, later the priory, at Svatý Kopeček for the years 1734–1741 and 1750–1751 (Oppeltová, 1999). The actual daily records of the individual chroniclers (Figure 3) usually include the day expressed as an ordinal numeral and its Antic name, followed by the characteristics of the weather. The order and type of the appropriate divine services follow. Further information covers arrivals and departures of abbey functionaries and guests, amusements, the life of the abbey community and events beyond the abbey walls (Oppeltová, 1999).

U

N

C

O R

R

EC

TE

D

Weather reports in the diaries

2

Figure 3. Title page of the entry in the Hradisko Premonstratensian abbey diary for January 1700 (Moravian Land Archives Brno, E55 Premonstratensians Hradisko, catalogue No. II-9).

The records, written in Latin, give brief characteristics for the weather for each day. More detailed reports occur in the description of extreme hydrometeorological phenomena involving possible damage. There are occasional overviews of the weather for a whole month. January 1723 provides an example: ‘This January enormous and almost unbearable frosts set in and the snow was ample and spongy.’ From entries it is possible to judge the temperature characteristics of the day, such as mild weather (aura tepida), frost (frigus: intensum – heavy, siccum – dry, tolerabilis – tolerable), warm (calida) or hot (aestus, calor: intensissimus – great, intollerabilis – intolerable). Precipitation is characterized by expressions for rain (pluit: modicum – mild, aliquantulum – subtle, densa – intense), snowfall (nives, ninxit: densissimae – very intense), downpour (imber: maximus – heavy), etc. For wind the writers mentioned, for example, strong wind (ventus validus, magnus), cold or warm wind (ventus frigidus/ calidus), fresh wind (ventus asper) or gale (turbo). Among the more extreme hydrometeorological phenomena, the authors pay particular attention to thunderstorms (tempestas), hailstorms (grando) and River Morava floods (exundatio). The state of the sky is described in terms of clear (serenum) and overcast (nubilum). Fog (nebula) is also mentioned.

Weather information in diaries of Czech Abbey

PR

O

O

FS

(a)

D

Weather information occurs in the entries for a total of 52 years, of which 42 stem from the Hradisko Abbey and 10 from Svatý Kopeček. The overall density of weather records may be expressed as the number of days containing such information in individual years (Figure 4). Records for all days of the given year appear only in 1731, with more than 90% of all days also being described in the years 1695, 1696, 1724– 1728, 1734–1741, 1743, 1744, 1747–1754, 1757, 1759, 1771 and 1774. On the other hand, in 1694, 1698, 1700, 1701, 1781 and 1783 the daily weather records covered less than a half of the days of the year (the least being in 1694 with entries for 31 days only and in 1783, when the records end abruptly on 31 March).

U

N

C

O R

R

EC

TE

Meteorological evaluation of the weather reports The weather terminology employed by those who compiled the diaries may be transformed to current usage and expressed as graphs. Figure 5 presents such an interpretation of the daily weather records for 1726. To stress the great intensity of frosts, the writer used unusually florid terms: Frigus plane Tartaricum (‘a real Tartar frost’) on 16 January and Frigus ... plane Moscoviticus (‘a real Moscow frost’) on 20 January. For 6 February he alluded to a frost so hard that it almost exceeded that of the winter of 1708/1709, considered one of the hardest in Europe (Lenke, 1964; Pfister, 1999; Luterbacher et al., 2004). The writer also mentions, in the entry for 25 January, an unusual depth of snow in Bohemia and Moravia, more than had been seen for many years; his entry for 4 March records that the depth of snow stopped carts moving. Rain and the melting of this snow at the end of March caused the River Morava to flood, and the inundation is specified in the entries for 2 and 9 April. A severe summer drought is explicitly mentioned in the entry for 19 July, recording cracked earth and the River Morava running very low. The effects of the drought were exacerbated by an exceed-


Figure 4. Density of daily weather records in the diaries of the Hradisko Premonstratensian abbey (1) and the residence at Svatý Kopeček (2) for the years 1693–1783.

Figure 5. Interpretation of daily weather observations in the Hradisko abbey diary for 1726. Key: H – heat, W – warm, M – mild, C – cold, F – frost (Fã – intensification of frosts, Fç – alleviation of frost), x – missing record, 1 – clear, 2 – half-covered sky, 3 – overcast, 4 – rain, 5 – snow, 6 – shower, 7 – thunderstorm, 8 – wind, 9 – cold wind, 10 – warm wind, 11 – fog, 12 – humid, 13 – flood. Underlined symbols express stronger intensity of the phenomena. A slash (/) distinguishes night and morning from afternoon and evening occurrences of the phenomena. 3


ture and precipitation series in Olomouc for 1961–2000. Thresholds for the conversion of temperatures to indices were derived from monthly mean x and standard deviation s calculated for the period 1961–1990 (x + 0.75s and x – 0.75s for index 0, x + 0.75s to x + 1.50s for index 1, x – 0.75s and x – 1.50s for index –1, etc.). Thresholds for precipitation indices were calculated for the same period from corresponding gamma distribution with probability levels 8.3% (index –3), between 8.3% and 25.0% (index –2), 25.0%–41.7% (index –1) etc. More pronounced deviations from the distribution of indices between the two periods can be attributed either to the natural variability of the climate or the quality of the index interpretation from daily qualitative weather records. In the case of temperature indices, there appears to be a trend towards conspicuous overestimation of cold months as against the warm ones, particularly in October, November and March, when apparent indicators of warm weather are lacking from the daily records. In the case of precipitation indices, rather drier months prevail in the interpretation (most marked in April and again in December), which may be attributed to inconsistent registration of precipitation in the daily entries, in which shorter or less abundant precipitation in particular, remained unnoticed by the writer or he did not feel obliged to record it. Only in July was the number of interpreted wet months higher than that of the drier ones. A significantly lower frequency of normal months (index 0) in the instrumental period compared to the interpretation from documentary data may be attributed to an excessively narrow interval of precipitation derived from gamma distribution (thresholds corresponding to 42.8% to 58.4%). A further problem associated with the diaries is the fact that writers changed over the extended period covered by the records, devoting different degrees of attention to the weather. Furthermore, parts of the records offer information about the weather hardly interpretable in terms of temperature or precipitation (such as ‘clear’ or ‘overcast’). Further precision may be lent to the indices obtained by drawing on other reports from the existing historical-climatological database of the Czech Lands and comparison with them.

EC

TE

D

PR

O

O

FS


(b)

information about the number of precipitation days, and the duration, character and intensity of precipitation. In some cases it was possible to complete or correct the interpretation with reports from the diaries of a nearby Piarist College in Stará Voda (Diarium Stará Voda I and II). However, these records are not daily entries and concern only extreme weather phenomena. Thus, for February 1747 the writer noted there that Aura erat non hyemalis, sed potius vernalis, nives nullae, sed ... pluviae (i.e. the weather was not of a wintry character, but was rather spring-like, no snow, but rains). The interpreted temperature and precipitation indices are presented in Table 1. The reconstruction of the years 1693– 1783 may be evaluated with respect to the distribution of interpreted temperature and precipitation indices (Figure 6) by making comparisons with the frequencies of indices derived from homogenized tempera-

U

N

C

O R

ingly long period of hot weather which, with only a few short interruptions, extended from 8 May to the end of September. A strong wind on 17 January 1726, which broke some glass window-panes, was also deemed worthy of mention. The daily weather reports in the diaries make it possible to quantify temperature and precipitation characteristics for individual months in the form of weighted indices on a –3 to 3 scale (–3 extremely cold, –2 very cold, –1 cold, 0 normal, 1 warm, 2 very warm, 3 extremely warm; –3 extremely dry, –2 very dry, –1 dry, 0 normal, 1 wet, 2 very wet, 3 extremely wet), as is usual in such work with documentary sources (Brázdil et al., 2005b). In the interpretation of temperature indices, information taken into consideration concerned the occurrence of frosts and cold and warm days in the given month, including the nature of precipitation in winter (liquid or solid). The precipitation indices involved

R

Fig. 5 (Continued)

4

Hydrometeorological extremes in the diaries A number of items relating to the occurrence of extreme hydrometeorological phenomena (floods, gales, hailstorms, thunderstorms, late frosts, droughts) and their impacts may be derived from the diaries. These either extend already existing reports from other documentary sources, or add information quite new to the Czech Lands. This holds particularly for information about floods on the River Morava, which complete, or render markedly more exact,

Weighted monthly temperature (before the slash symbol) and precipitation (after the slash symbol) indices interpreted according to the records of weather and related phenomena at the Hradisko Abbey and the Svatý Kopeček residence in the years 1693–1783: x denotes a missing quantitative interpretation arising from the incompleteness of the records; – marks missing daily records.

C

D –1/–1 x/x –1/–2 –2/0 –1/–2 0/x x/x x/x –/– 0/–2 1/–1 0/–1 1/2 –1/–1 0/–1 0/–1 1/0 –1/–1 0/0 –/– –/– 1/–2 1/–2 0/0 0/0 –1/–2 0/–2 0/2 0/–3 0/–1 0/0 1/–2 2/2 0/2 0/0 –1/0 0/0 1/1 0/–2 –1/x –1/2 0/–2 –1/–2 –1/0 0/x 0/–1 2/2 3/0 –1/–3 –/– –1/0 –/–

FS

N –1/1 –/– 0/–3 0/–1 –1/0 0/x 0/0 –1/1 –/– –2/–1 0/0 0/–2 0/2 –2/0 0/0 0/–1 0/–2 0/0 0/–1 –/– –/– –2/0 0/–1 –2/0 –2/1 –2/–3 –3/–2 –1/0 0/–2 2/–2 0/1 0/0 0/1 0/–1 –1/3 –3/–2 0/0 –1/–1 1/–2 –2/0 –1/0 0/–1 –2/1 –1/3 0/1 1/0 –3/0 0/–1 –2/1 –/– –2/1 –/–

O

O –1/–2 –/– 0/–2 –1/–1 0/0 1/x –2/–3 –1/2 –/– 0/–1 –1/1 0/–2 1/–1 –1/–2 0/–2 0/–1 1/–1 0/–1 0/–3 x/x –/– –1/1 –1/1 0/1 0/2 0/0 –2/1 –3/1 1/–2 0/1 0/2 0/0 0/0 –1/0 –1/–1 –1/–1 0/–3 0/0 2/–2 0/–1 –2/0 0/–2 0/0 –2/3 1/–2 –2/1 0/–3 1/–1 0/–2 –/– –1/0 –/–

O

S –1/2 –/– –2/1 0/–2 0/–2 0/x 3/–2 x/x –/– 1/–2 1/0 0/2 –1/2 2/–2 –1/0 3/–3 3/–2 0/1 0/0 x/x x/x –1/1 1/0 1/1 0/0 0/–1 0/1 2/–2 –2/2 0/–2 1/–1 1/–2 0/0 0/–1 1/–3 –2/1 0/1 0/1 0/–2 0/0 –2/0 –1/1 –1/0 0/0 –1/0 0/1 0/1 0/–1 –2/2 x/x 0/–2 –/–

PR

A 2/–2 x/1 –1/2 0/–1 –2/1 0/x 1/x x/x 2/x 1/0 2/0 0/–1 0/0 2/–2 –2/1 2/–1 2/–2 1/–2 –1/0 x/x 0/1 –2/3 –1/–1 2/0 –2/0 –1/0 –1/0 –2/2 –1/1 0/–1 0/0 1/–3 1/1 0/0 0/0 0/0 –1/2 –1/1 1/–2 –1/2 1/0 2/–2 0/0 0/0 0/x –1/2 –1/2 0/0 0/–1 1/–1 0/–1 –/–

D

J 0/0 x/x –1/1 0/0 –1/2 0/2 1/x x/x 1/x –1/2 0/1 0/1 –2/1 1/–1 0/–2 3/–2 0/–2 2/–2 0/–1 0/x 1/0 –1/1 0/1 –1/3 1/0 0/0 1/1 –2/0 –1/2 –2/1 0/0 1/–2 0/–1 1/–2 0/1 –1/0 0/1 0/2 –1/–1 x/x 2/2 –1/1 1/0 0/–1 0/0 0/2 0/3 –1/2 0/–2 0/–1 1/–2 –/–

TE

J 2/x x/x –2/–1 –3/2 –1/–1 x/x x/x x/1 x/0 0/0 2/–1 0/0 0/1 2/–2 1/–1 3/–1 1/0 2/–2 –1/–2 0/–1 –2/0 –2/1 1/2 –1/3 0/–1 –1/1 –1/0 –1/0 –1/1 0/2 0/–2 2/–2 1/0 –2/1 0/1 –2/–2 1/–2 1/–2 0/0 2/0 1/0 1/–1 0/0 2/–1 –1/–1 –1/1 –1/3 1/–2 0/0 0/0 1/–2 –/–

EC

M –2/2 x/–2 –2/1 1/–2 0/0 –2/1 –1/0 x/x –2/0 –1/0 1/–1 1/0 0/–3 0/0 0/2 3/–3 3/–2 1/0 0/–1 1/–1 –1/1 –1/2 –1/0 –1/1 2/–2 0/0 0/–1 –3/1 –2/–2 0/–2 0/0 0/0 0/0 1/1 –2/1 –2/0 –2/3 0/–1 0/–1 –1/0 0/0 1/x –2/0 0/–2 0/–2 –1/0 2/–2 1/0 –1/0 x/x 0/1 –/–

R

A x/x –1/x 0/–1 –1/–1 0/–2 0/x x/x x/x –2/0 –1/1 2/–1 0/0 1/0 0/–1 2/–2 0/–1 0/–1 0/0 –1/–1 1/–3 1/–3 0/2 0/–1 1/–1 –1/–2 0/–1 –2/–1 –3/–1 –2/1 –2/0 2/–1 0/–2 1/1 0/–1 –1/0 –1/0 0/–2 1/–2 –1/0 0/0 1/0 0/0 1/–1 0/–2 0/0 0/–1 –1/0 0/2 0/0 –/– 0/0 –/–

O R

M –/– –/– –3/0 –2/0 0/0 –1/x –2/0 –3/x –1/–2 x/x –1/0 0/0 2/0 0/1 0/–1 –2/2 0/–2 1/–1 0/0 0/0 0/–2 0/0 1/0 1/–1 0/0 –1/–1 0/0 –2/0 –1/0 0/–1 –1/2 –3/–1 –2/–2 0/0 1/–1 1/2 1/–2 1/–1 –2/–1 0/x –1/–2 0/0 0/0 x/x –3/–2 0/1 –2/1 0/–2 1/–2 x/x –1/0 –1/1

N

F 1/x x/x –2/0 2/0 –1/–2 –1/x –2/x –1/1 0/x –/– –1/x 0/0 0/–2 3/1 –1/–3 –2/1 1/0 –1/–2 –1/–2 0/x x/x 1/0 0/–2 1/1 –1/–1 0/–2 2/–1 –3/2 1/–1 1/–2 –1/–1 2/1 –1/–2 0/0 2/0 –1/2 2/0 1/–2 0/0 1/–1 0/1 0/0 2/–1 1/0 –2/0 x/x –1/–2 0/–2 0/0 1/–2 –2/–1 2/1

U

J 0/–1 –2/x –3/–1 1/–1 –2/–2 –2/1 2/–1 –1/0 0/–1 –/– –1/x 0/–3 –3/1 3/1 0/x –3/1 2/1 1/–1 –3/–3 –2/x –/– 0/0 2/0 0/–2 1/2 –2/–1 0/1 –2/–1 0/1 0/–1 –2/–2 –1/–1 –1/–2 2/2 –1/–2 0/–1 0/0 –2/–2 0/–2 1/1 –2/3 –3/0 –1/–1 –1/0 0/–2 1/0 0/2 1/0 0/0 0/–1 1/x 2/2


Year 1693 1694 1695 1696 1697 1698 1699 1700 1701 1713 1718 1722 1723 1724 1725 1726 1727 1728 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1743 1744 1747 1748 1749 1750 1751 1752 1753 1754 1756 1757 1758 1759 1761 1768 1769 1771 1773 1774 1781 1782 1783


Table 1.

5


PR

O

O

FS


(a)

U

N

C

O R

R

EC

TE

D

(b)

6

Figure 6. Histogram of the distribution of frequencies of monthly temperature and precipitation indices interpreted according to the records of the Hradisko Abbey and the Svatý Kopeček residence in the period 1693–1783 (1) in comparison with the period 1961–2000 (2) at the Olomouc meteorological station (according to thresholds calculated for 1961–1990).

Apart from dendroclimatic knowledge (Brázdil et al., 2002a), documentary evidence provides the only source of information for the reconstruction of the climate of the Czech Lands with high temporal accuracy in the pre-instrumental period. The temporal and spatial discontinuity of the documentary data dictates that a mosaic of knowledge

FS

O

O

D

TE

EC

References

R

Brázdil R, Dobrovolný P, Elleder L, Kakos V, Kotyza O, Květoň V, Macková J, Müller M, Štekl J, Tolasz R, Valášek H. 2005a. Historical and recent floods in the Czech Republic. Masaryk University, Czech Hydrometeorological Institute: Brno, Praha. Brázdil R, Dobrovolný P, Štekl J, Kotyza O, Valášek H, Jež J. 2004. History of weather and climate in the Czech Lands VI: strong winds. Masaryk University: Brno. Brázdil R, Kotyza O. 1996. The earliest daily weather records in the Czech Lands and their utilisation for the reconstruction of climate. Weather 51: 341–349. Brázdil R, Pfister C, Wanner H, von Storch H, Luterbacher J. 2005b. Historical climatology in Europe – the state of the art. Climatic Change 70: 363–430. Brázdil R, Štěpánková P, Kyncl T, Kyncl J. 2002a. Fir tree-ring reconstruction of March–July precipitation in southern Moravia (Czech Republic), 1376–1996. Climate Res. 20: 223–239. Brázdil R, Valášek H. 2002. Meteorologická měření a pozorování v Zákupech v letech 1718–1720. Geografie – Sborník České geografické společnosti 107: 1–22. Brázdil R, Valášek H, Macková J. 2003. Climate in the Czech Lands during the

O R

C

N

U

Conclusions

PR

Acknowledgements This paper was made possible by financial support from the Grant Agency of the Czech Republic, Grant No. 205/05/0858 and by EU project FP-6 No. 017008 European climate of the past millennium (MILLENNIUM). We thank Bořivoj Herzlík (Brno) for the English translation, Tony Long (Svinošice) and Dennis Wheeler (Sunderland) for the English style correction.

1780s in light of the daily weather records of parson Karel Bernard Hein of Hodonice (southwestern Moravia): Comparison of documentary and instrumental data. Climatic Change 60: 297–327. Brázdil R, Valášek H, Macková J. 2005c. Meteorologická pozorování v Brně v první polovině 19. století (Historie počasí a hydrometeorologických extrémů). Archiv města Brna: Brno. Brázdil R, Valášek H, Sviták Z, Macková J. 2002b. History of weather and climate in the Czech Lands V. Instrumental meteorological measurements in Moravia up to the end of the eighteenth century. Masaryk University: Brno. Fiala J. 2003. Olomoucké veduty z díla Friedricha Bernharda Wernera. Nakladatelství DANAL: Olomouc. Foltýn D (ed.). 2005. Encyklopedie moravských a slezských klášterů. Nakladatelství Libri: Praha, pp. 513–522, 525–529. Lenke W. 1964. Untersuchungen der ältesten Temperaturmessungen mit Hilfe des strengen Winters 1708–1709. Berichte des Deutschen Wetterdienstes 13, 92, 45 pp. Luterbacher J, Dietrich D, Xoplaki E, Grosjean M, Wanner H. 2004. European seasonal and annual temperature variability, trends and extremes since 1500. Science 303: 1499–1503. Oppeltová J. 1999. Narativní prameny vzniklé v prostředí premonstrátské kanonie Klášterní Hradisko u Olomouce v 17. a 18. století. In: Macková M, Oppeltová J. (eds): Slavme chvíli ... Sborník k 70. narozeninám Jana Bystřického. Státní okresní archiv Ústí nad Orlicí, Ústí nad Orlicí, pp. 83–121. Pejml K. 1975. 200 let meteorologické observatoře v pražském Klementinu. Hydrometeorologický ústav: Praha. Pfister C. 1999. Wetternachhersage. 500 Jahre Klimavariationen und Naturkatastrophen (1496–1995). Verlag Paul Haupt: Bern, Stuttgart, Wien. Smejkal B, Hyhlík V. 1994. Svatý Kopeček. Poutní chrám Navštívení Panny Marie. Edice Církevní památky č. 14. Historická společnost Starý Velehrad: Velehrad.


has to be built gradually from them, which, in turn, is of use in deriving series of temperature and precipitation indices and their subsequent quantitative reconstruction. The same applies to the derivation of series of hydrometeorological extremes. From this point of view, the diaries of the Hradisko Premonstratensian abbey and the residence at Svatý Kopeček represent an exceptionally valuable new source which can be compared with a further body of historical-climatological data accessible in the database of the Geographical Institute of Masaryk University, to which it forms an important complement. The qualitative daily weather records for the years 1693–1783 from the Hradisko Abbey were forerunners of the instrumental meteorological observations carried out at Olomouc in 1790–1794 by Josef Gaar, a professor at the Olomouc lyceum (Brázdil et al., 2002b). At the Klášterní Hradisko itself, such observations continued in the first half of the nineteenth century and thereafter continuously from 1 January 1876, but now within the network of meteorological stations in the Czech Lands.


the chronology from the publication of Brázdil et al. (2005a) for 27 cases out of a total of 31 floods recorded in the diaries, these being those for 5 April 1695, 14 March 1697, 10 January 1698, 26–29 March 1698, 16 July 1698, 26–31 January 1700, 1 December 1700, 20–22 March 1701, 1–2 August 1713, 28 December 1723–1 January 1724, 2 and 9 April 1726, 20 October 1734, July 1736, January 1737, December 1748, 18 January 1749, 14 July 1750, 18 December 1750, 12–15 March 1751, 22 May 1752, 18 February 1753, 26 December 1753, 23–27 February 1761, 24 April 1768, 23 December 1769, 18–21 March 1771, 21–22 April 1771, 6 June 1771, 27 February 1774, 8 March 1781 and 11 January 1783. In similar fashion, records of gales with associated damage to buildings and stands of timber provided additional information for five cases in the existing chronology of strong winds for the Czech Lands (Brázdil et al., 2004) and identified 13 wholly new AQ2 incidences: March 1696,• 13 August 1701, 5 March 1718, 21 July 1724, 17 January 1726, 11 March 1727, 6 October 1728 (wind-breakage), 30 March 1731, 17 November 1731 (wind-breakage), 8 December 1731, 12 July 1734, 20 and 22 March 1740, 1 August 1740, 11 September 1751, 17 August 1757, 21–24 September 1757, 21 November 1771 (windbreakage) and 17 July 1782. Reports of damage associated with heavy thunderstorms are also frequent in the diaries, particularly that due to hailstorms (in 18 years) and lightning (setting fire to buildings or killing people, in 10 years). Thus, in 1749, hail was recorded five times while June 1743 was exceptional for thunderstorms which occurred daily from 3 to 8 June; in total there were 11 thunderstorm days that month. In June–August 1769, some 22 days with thunderstorms were recorded. Great drought is explicitly mentioned in the diaries for the years 1694 (May), 1699 (September), 1718 (June), 1723 (May), 1726 (July), 1737 (May), 1741 (May), 1743 (May), 1748 (July), 1749 (July), 1750 (September – prayers for rain), 1752 (June), 1753 (June– July), 1758 (June), 1771 (May), 1773 (before 1 October) and 1782 (August–September). The enumeration of extremes finishes with damage to blossoming trees and field crops due to late frosts, which are mentioned in the diary records for 18 May 1697, 30 April 1698, 1 June 1734, 1 May 1739, 10 and 22 April 1750, 5 and 11 June 1751, and 10 May 1768.

Archival sources Diarium I: Diarium piaristické koleje Stará Voda z let 1690–1718. Zemský archiv Opava – pobočka Olomouc, Sbírka rukopisů Metropolitní kapituly Olomouc, CO 632. Diarium II: Diarium piaristické koleje Stará Voda z let 1719–1798. Zemský archiv Opava – pobočka Olomouc, Sbírka rukopisů Metropolitní kapituly Olomouc, CO 633. Diaria kanonie Klášterní Hradisko 1693– 1783. Moravský zemský archiv Brno, fond E55 Premonstráti Hradisko, sign. II-4 to II-41.

Correspondence to: Prof. Rudolf Brázdil, Institute of Geography, Masaryk University, Kotlářská 2, 611 37 Brno, Czech Republic Email: [email protected] ©Royal Meteorological Society, 2008 DOI: 10.1002/wea.264

7

Příloha č. 7

Příloha č. 8

Příloha č. 9

EARLY INSTRUMENTAL METEOROLOGICAL OBSERVATIONS IN THE CZECH LANDS I: FERDINAND KNITTELMAYER, BRNO, 1799–1812 RUDOLF BRÁZDIL1, LADISLAVA ŘEZNÍČKOVÁ1, HUBERT VALÁŠEK2 1 2

Institute of Geography, Masaryk University, Kotlářská 2, 611 37 Brno, Czech Republic Moravian Land Archives, Žerotínovo nám. 3–5, 656 01 Brno, Czech Republic

In the Czech Lands there exist a number of early meteorological measurements carried out before the introduction of regular meteorological observations. With the intention of obtaining homogeneous climatological series for the city of Brno, meteorological measurements before 1848, i.e. before the beginning of official publication with observations made by Dr. P. Olexik, are systematically investigated. Currently the oldest known records for Brno date from 1799–1812, compiled by Captain Ferdinand Knittelmayer (rtd.), which remain preserved in the form of selected summary data from observations taken five times a day. The motivation for the observations was Knittelmayer’s belief that the course of the weather could be forecasted on the basis of the nineteen-year lunar cycle. This paper presents the results of statistical analysis of air pressure and temperature, wind direction and wind force, cloud characteristics and the number of precipitation days in comparison with measurements taken at the Brno-Tuřany meteorological station in the period 1961–1990. V českých zemích existuje řada časných meteorologických měření, prováděných před začátkem pravidelných meteorologických pozorování. S cílem získat homogenní klimatologické řady Brna jsou systematicky zpracovávána meteorologická měření před rokem 1848, tj. před začátkem oficiálně publikovaných pozorování dr. P. Olexika. Zatím nejstarší známé záznamy z Brna z let 1799–1812 pocházejí od penzionovaného setníka Ferdinanda Knittelmayera, které zůstaly zachovány v podobě vybraných shrnujících údajů z pěti denních pozorování. Motivací pro pozorování byla pro Knittelmayera víra v možnost předvídat průběh počasí na základě 19-letého měsíčního cyklu. Práce prezentuje výsledky statistické analýzy tlaku a teploty vzduchu, směru a síly větru, charakteristik oblaků a počtu srážkových dnů, porovnávaných s měřeními na meteorologické stanici Brno-Tuřany v období 1961–1990. Key words: early instrumental measurements, pressure, temperature, wind, cloudiness, precipitation days, meteorological singularities, Ferdinand Knittelmayer, Brno

INTRODUCTION Early instrumental meteorological measurements are important sources of documentary data that can broaden our information base for weather and climate before the establishment of national networks of meteorological stations. Although the beginning of the official network of meteorological stations in the Czech Lands is associated exclusively with the establishment of the Central Institute for Meteorology and Earth Magnetism in Vienna in 1851, which started a long tradition of systematic publishing with the results of observations for the year 1848 (Brázdil et al., 2005), measurements had been made at a number of meteorological stations long before that. Excluding isolated quantitative air temperature data, the first systematic daily measurements in Bohemia come from the physician Johann Carl Rost, in Zákupy, for the period 21 December 1719–31 March 1720. These were published together with measurements from further European stations within what was known as the “Breslau network” organised by Johann Kanold, a physician in Wrocław (Breslau in German)

(Brázdil, Valášek, 2002). Measurements survive from 1752, taken by Josef Stepling, the first director of the observatory in Prague’s Klementinum, which continued in systematic form from 1 January 1775 thanks to the initiative of the third director, Antonín Strnad (Pejml, 1975). Strnad himself became a significant propagator of meteorological observations, and they spread into further parts of Bohemia, such as Žitenice, Teplá and elsewhere (Seydl, 1952). Learned and economic societies played an important role in the development of meteorological observations in Bohemia. Convinced that the properties of the atmosphere and weather had an important influence on agriculture, the I. R. (Imperial-Royal) Patriotic-Economic Society established in 1796, thanks to the activities of the member Antonín Strnad, made meteorological observations in the Bohemian regions and had their own tables purpose-printed for recording. The Society also had an appropriate number of barometers and thermometers made and supplied to them for their observers. A landmark event was the publication of the results for 1817–1821 by Alois David, the fourth director of the Klementinum observatory, (Nachricht I–II,

Meteorologický časopis, 9, 2006, 59-71

|

59

1825–1826), which were to be followed by new papers from the Society, including observations from Bohemia and their analysis for the years 1822–1847 (Neue Schriften I–X, 1828–1847; Verhandlungen, 1849–1850). As far as is known, the earliest daily meteorological measurements in Moravia originated with the Telč physician František Alois Mag of Magg, beginning in his second observation diary on 7 May 1771 and ending on 9 March 1775 (Valášek et al., 2001; Brázdil et al., 2002a, 2002b). According to data sent to Antonín Strnad, it is evident that Mag was actively observing in the winter of 1788/89, at the very least (Strnadt, 1791, 1793). In Olomouc, meteorological observations were carried out by Josef Gaar, a local lyceum professor, from 1790 onwards, part of his observations being presented in the oldest description of the climate of Moravia, by Kryštof Passy, another professor from Olomouc (Brázdil, Valášek, 2001). The establishment of the Meteorological Section of the I. R. Moravian-Silesian Economic Society in 1815 lent important and lasting impetus to the development of meteorological observations in Moravia and Silesia. As well as receiving detailed printed instructions as to the techniques of observations, participants could also buy a barometer and a thermometer with a Réaumur scale, made by Prof. Kassián Hallaschka, and they were asked to send their observations to the address of the Meteorological Section in Brno. Parts of the records remain preserved in the Society archives; furthermore, the daily observations for Brno in 1820–1847 were even published regularly in Brünner Zeitung, the local newspaper (for details, see Brázdil et al., 2005). Brno has a long tradition of meteorological observations. Although it is known that they started in 1797, only those made two years later and onwards survive, coming from Captain Ferdinand Knittelmayer (rtd.) (Brázdil et al., 2005). The present paper deals with a detailed analysis of those observations, from 1799 to 1812.

FERDINAND KNITTELMAYER Ferdinand Knittelmayer (also spelt Knittlmayer or Knitlmajer) was born on 20 January 1750 in Vienna. After completing grammar school studies at the age of 18, he joined the army as an imperial cadet. During his military service he took part in several campaigns: the Bavarian succession wars in 1778 and 1779 subsequent to the demise of the Bavarian branch of the Wittelsbach family, the war with the Turks in Moldavia in 1790, and the field campaigns in the Rhineland and Alsace, 1792–1794. He was then deemed unfit for further field service and transferred to the peacetime garrison in Brno, where he commanded a reserve division. In 1801, Knittelmayer retired from the army on medical grounds, with the rank of captain. During his time in Brno he had become a member of the I. R. Moravian-Silesian Economic Society, devoting himself to astronomical and meteorological observations. In the winter of 1812/13 he contracted a painful eye disease which eventually blinded him, a crushing blow to his service as an active observer. He died in Brno on 21 November 1814 (Buse, 1816).

60

|

Meteorologický časopis, 9, 2006

METEOROLOGICAL OBSERVATIONS MADE BY FERDINAND KNITTELMAYER Early instrumental meteorological measurements are, as a rule, partly published in contemporary materials, frequently accessible only with difficulty. The original records are often lost or remain preserved in the form of manuscripts in district, regional or other archives. Typically enough, Knittelmayer’s meteorological observations for Brno have unfortunately not been preserved in the form of original measurements from the individual daily terms. The author made only monthly, seasonal, half-yearly and annual overviews of his daily observations for all the meteorological elements and phenomena monitored for the period 1799–1812 (Table 1). Table 1. The beginning, interruption and end of Knittelmayer’s observations in Brno according to individual meteorological elements. Element

Beginning

End

Interruptions 21 May 1799 23 Apr–7 May 1804 28–30 June 1809

Pressure

1 May 1799

26 Sep 1812 (3 Oct 1812)

Temperature

1 May 1799

4(5) Oct 1812

28–30 June 1809

20 Mar 1812

19–22 Jan 1801 22–28 Dec 1811 27 Jan–3 Feb 1812 22–31 Dec 1811

Wind

1 May 1799

Clouds

1 May 1799

19 Sep 1812

Weather phenomena

1 May 1799

22 Dec 1812

-

The surviving extracts of Knittelmayer’s observations consist of 15 individual unbound files and one file of introductory text, deposited among the manuscripts of the Archives of the City of Brno (Meteorologische Beobachtungen). They are individual 40×25-cm sheets, folded and not bound; at first sight they give the impression of individual copybooks. In his seventeen pages of introductory text, Knittelmayer presents, in a neat cursive hand, basic information about his measurements. He gives explanations of the appended tables and outlines instructions for their later utilisation. He measured and observed air pressure and temperature, the force and direction of the wind, the clearness and cover of the sky, the movement and properties of clouds and other meteorological phenomena. The observer also mentions the location of the instruments in his flat, from which he had a pleasant and free view. The Réaumur thermometer was placed in free air in front of the window, facing east, but in such a fashion as to be permanently protected from the rising sun. The barometer, with a scale in Vienna inches was, in contrast, hung up within the room, on the inner side of the window. Knittelmayer determined the direction of the wind by the free-swinging weathercock on the spire of St. James’ church, also making use of the direction of the smoke from chimneys. Wind force was determined as he himself felt it. He could observe the clouds in the sky all around, as well as their movement and properties. Other characteristics were observed in the normal way.

Dřímal (1956) located Knittelmayer’s Brno observations in what was once a Dominican monastery (around 230 m a.s.l.). In its tower, Knittelmayer may have had a small observatory in which he carried out his astronomical and meteorological observations (Figure 1). The monastery was used by the military garrison in which Knittelmayer was a captain. It is probable that his flat, with the instruments described, was on the second floor of this building with windows facing east and a view of the city centre that included the spire of St. Jacob’s church. Knittelmayer took proper meteorological records five times a day, although he was aware of the fact that at astro-

nomical observatories such observations were made only three times a day. His first term, winter or summer, was at sunrise, “of which it is known to me from experience that at that moment the thermometer had day-by-day its lowest value, and subsequently the greatest cold times set in”. The morning term was at 9 o’clock in summer and at 10 o’clock in winter. The third daily observation took place at 14.00 in summer and soon after 12.00 in winter “when about this time the thermometer will stand highest, i.e. it shows the highest daily temperature”. The evening observations were bound to the time of sunset. The last daily term for Knittelmayer was at 23.00 and often at exactly midnight.

Figure 1. a) A view of the former Dominican monastery in Brno (see arrow), in the tower of which, from 1802 onwards, Ferdinand Knittelmayer may well have had a small observatory in which to perform his astronomical and meteorological observations (drawing by František Richter from 1827 – Brodesser, 2005),

1

b) locations of meteorological stations used in Brno: 1 – former Dominican monastery (Knittelmayer), 2 – the Brno-Tuřany meteorological station.

2


|

61

For the presentation of the results of his measurements, Knittelmayer used the astronomical division of the year with respect to the days of summer and autumn equinoxes, which he divided into 6 summer months (all of them of 31 days) and 6 winter months (of 30 days each, and the last one only 29 days): • summer months 21 March–20 April (“Frühmonat” or “Keimmonat”), 21 April–21 May (“Blüthenmonat”), 22 May–21 June (“Rosenmonat”), 22 June–22 July (“Wärmemonat”), 23 July–22 August (“Erndtemonat”), 23 August–22 September (“Obstmonat”), • winter months 23 September–22 October (“Weinmonat”), 23 October–21 November (“Nebelmonat”), 22 November–21 December (“Schneemonat”), 22 December–20 January (“Kältemonat”), 21 January–19 February (“Eismonat”), 20 February–20 March (“Thaumonat”). Further folded sheets within Knittelmayer’s set of papers contain the processed results of his observations,

always arranged as units. These consist of individual sheets and some hand-made forms with fixed columns into which mean measured values were entered, as well as all notes of importance. The heading of each sheet refers to the year and the kind of measurements taken, together with a brief note about the time-span within which the measurements were carried out. For each observed element there always exist three independent files. The first file includes the tables of the six summer months of the given astronomical year (i.e. from the spring to the autumn equinox of the same year) for the period 1799–1812 (Figure 2). The second file, also constituting an independent unit, contains the results for the winter months, i.e. from the autumn equinox of the given year to the spring equinox of the subsequent year, again for the above period. The third set gives an overview of the results obtained for all years and months of the whole period 1799–1812. A voluminous extract of Knittelmayer’s meteorological observations was bought by the I. R. Moravian-Silesian Economic Society, who donated it to the Franz Museum, an affiliated organisation. In 1900, Knittelmayer’s set of results was moved to the Moravian Land Archives, and in 1941 it was further transferred to the Brno City Archives.

Figure 2. A specimen of Knittelmayer’s summary records of air pressure.

62

|


WHY KNITTELMAYER MADE METEOROLOGICAL OBSERVATIONS Knittelmayer’s motivation for meteorological observations was his astronomical experience. He reasoned that, largely, the same course of weather recurs after fulfilling the nineteen-year lunar cycle, because the light of the moon on the corresponding days is the same as 19 years previously (Buse, 1816). Earlier, Antonín Strnad was also persuaded of the prognostic importance of the nineteen-year lunar cycle for the weather forecast. He published his prognostica for the years 1788–1790, given as parallels to the weather in the given year as observed at Prague-Klementinum in 1769–1771 (see Strnadt, 1788, 1789, 1790). Knittelmayer ascribed the formation of the weather to two unchangeable factors (the sun and moon) and three changeable and iregularly-operating ones (local patterns, chemical processes and the winds). This is why he paid attention to all phenomena that could possibly be connected with the above factors in his observations. At the end of each lunar quarter and the whole month, in the same way as at each quarter and half year, the daily observations were compared with one another and also with the course and state of the moon. The objective was, on the one hand, to determine the mean characteristics of the weather or similarities and deviations of the course of the weather over several months through all respective combinations. On the other hand, he sought to derive estimates from those observations for the probable course of the weather during the following change of the moon. The results of the meteorological observations thus made it possible for Knittelmayer to determine in advance the course of the weather for the following moon quarter or for the whole month. In essence, it was an attempt at a weather forecast by an analogue method. For a clearer comparison of his results, Knittelmayer made graphs of the course of air temperature and pressure, a specimen of which, with corresponding text and subsequent tabular expression, may be found in the daily Patriotisches Tageblatt (Knittelmayer, 1800a, 1800b, 1800c, 1801). His graphical expression of the course of meteorological elements between 21 March and 21 May 1800 in Brno are among the earliest approaches of this kind in the Czech Lands (Knittelmayer, 1800b, Annex). The following specimen also documents the style of his thinking: “Both in the first [Frühmonat] and in the second month [Blüthenmonat] the temperature rose to the New Moon, mainly with the passage of the Moon across the equator; towards the first quarter it dropped a little. Towards the Full Moon soon afterwards the temperature dropped perceptibly, particularly due to the fact that the southern deviation of the Moon increased. Towards the last quarter the temperature slowly rose again, as the Moon approached the equator.” (Knittelmayer, 1800b). He also gave an analogous explanation for the wind directions: “In consideration of the phases of the Moon it appeared that the first three months [Frühmonat–Rosenmonat] from New Moon to the first quarter there blew mainly south-easterly winds, which towards the third [Rosenmonat] changed to north-easterly. In the further three months [Wärmemonat–Obstmonat] the

north-easterly was the most frequent at the beginning, then the north-westerly, and at the end the southerly wind.” (Knittelmayer, 1800c).

THE CHARACTER OF THE WEATHER IN BRNO, 1799–1812 The technique of early instrumental meteorological measurements encounter a number of difficulties when compared with the analysis of modern measurements. As well as problems of accessibility, the place of observation itself within the given community is usually not specified. Detailed information about the instruments employed and their location is also lacking. Observation hours, which were often not stable even within the same set of measurements, as well as the methods of calculating the mean values, differ from current practice. The observations are not always complete, while what is complete coincides with only a limited number of contemporary stations that might be utilised as reference points. There are also problems with visual observations, the use of different terminology or definitions of meteorological phenomena. These facts make the process itself difficult, as well as hindering the statistical analysis of older observations, including their comparison with present-day standards (Brázdil et al., 2005). This holds true to a considerable extent for Knittelmayer’s observations from Brno, for which, in the publication by Brázdil et al. (2005), monthly means of air temperature and pressure were calculated for the period 1800–1812. They were further used for the compilation of Brno temperature and pressure series for 1800–1850. In the following part of the paper are the results of Knittelmayer’s meteorological observations arranged by individual meteorological elements and phenomena in the period 1799–1812. This, in comparison with the standard period 1961–1990, appeared conspicuously colder from December to March, partly in April and November as well (Figure 3), according to the data from the secular Prague-Klementinum and Vienna-Hohe Warte stations. On the other hand, higher temperatures were achieved above all in May and August, but also in September and July. In the case of air pressure, higher values as against the standard period were recorded mainly in June and August, but also in February–April, a conspicuously lower pressure occurring in December. Air pressure For air pressure, measured in Vienna inches and lines with an accuracy of one-eighth of a line (1 Vienna inch = 26.34 mm, 1 Vienna line = 2.195 mm), Knittelmayer recorded the lowest air pressure, corrected mean air pressure and the highest daily pressure change. The corrected daily mean pressures transferred to hPa were used in the further processing, although Knittelmayer’s description does make the nature of the correction clear. From those values the monthly means of air pressure were further calculated, then tested for relative homogeneity by the Alexandersson test (Alexandersson, 1986) with respect to the mean series of air pressure for the reference stations Prague-Klementinum


|

63

Figure 3. Differences in mean air temperature and pressure for the years 1799–1812 and the normal period 1961–1990 for secular Prague-Klementinum (only temperature) and Vienna-Hohe Warte stations.

Temperature [°C]

2.5 1.5 0.5 -0.5 Prague - Klementinum -1.5

Vienna - Hohe Warte

-2.5 J

F

M

A

M

J

J

A

S

O

N

D

N

D

2.5 1.5 Pressure [hPa]

and Vienna-Hohe Warte. Statistically significant inhomogeneities in 1802 were found in the months of December– February and in August, so these months were subsequently adjusted (both for the monthly and for the daily values). Higher values for air pressure before 1802 in comparison with the subsequent period follow logically from Dřímal’s account (1956), according to which Knittelmayer established a small observatory in 1802 in the tower of the former Dominican monastery. The annual variations of air pressure in Brno in the period 1799–1812 according to Knittelmayer’s observations are expressed by monthly means, the means of daily maxima and minima and absolute monthly extremes (Figure 4). In comparison with modern measurements from the Brno-Tuřany station (Figure 1b) in the period 1961– 1990, Knittelmayer recorded lower values for air pressure from October to February and higher values from April to June. The annual variations also differ from the point of view of the occurrence of maxima (in Knittelmayer’s series the maximum was in September, at Tuřany in October), whereas the minimum was recorded for both places in April.

0.5 -0.5 Vienna - Hohe Warte

-1.5 -2.5 J

Air temperature

64

|


M

A

M

J

J

A

S

O

Figure 4. A comparison of the annual variation of mean air pressure in Brno according to measurements made by Knittelmayer (1) in the period 1799–1812 (3 – mean of daily maxima/ minima, 4 – absolute daily maximum/minimum) and the BrnoTuřany meteorological station (2) in the period 1961–1990. 1020 1010 1000

Pressure [hPa]

For the air temperature, given in Réaumur degrees (1°R = 1.25°C), the mean from the highest and the lowest daily temperature, the average of all five daily observations, the mean over the course of the day from the morning to the evening and the mean during the night from the evening to the morning were always given. The object of our further analysis became the mean temperature calculated by the two methods mentioned above. Using the daily maximum and minimum has the disadvantage that the resulting average is based on only two extreme values. In the calculation from five daily readings the mean is loaded with the error following from unequal observation terms in different parts of the year (such as those generated by measurements at the variable times of sunrise and sunset). To judge the differences between the two methods, a box diagram was compiled of the annual variation of differences (Figure 5). Further, the significance of the differences was established by t-test for the paired values. It appears that the means of the five daily observations more frequently exceed the averages from two measurements, at the maximum by more than 3°C. In the opposite case, the maximum differences exceeded by as much as 2°C. The differences found between the means are, however, in all months statistically insignificant at the α = 0.05 level of significance. The daily means calculated from five daily readings were used for the calculation of monthly averages, which were further tested by using the Alexandersson test (Alexandersson, 1986) for relative homogeneity using reference temperature series from the Prague-Klementinum and Vienna-Hohe Warte stations, both for the individual months and for the seasons and the year. The test indicated a possible statistically significant inhomogeneity only in April 1801 as a consequence of a somewhat lower temperature in April 1800, but the correction of only one month was omitted.

F

990 980 970 960 950

1

2

3

4

940

J

F

M

A

M

J

J

A

S

O

N

D

Figure 5. Box plot of the annual variation of differences of daily means of air temperatures calculated from the daily maximum and minimum, and/or five daily readings, in Brno in the period 1799–1812. The graph includes values corresponding to the median, the first and the third quartiles, and absolute positive and negative differences.

Figure 6. Comparison of the annual variation of mean air temperature in Brno according to Knittelmayer’s measurements (1) in the period 1799–1812 (3 – mean of daily maxima/minima, 4 – absolute daily maximum/minimum) and the Brno-Tuřany meteorological station (2) in the period 1961–1990.

Figure 7. Annual variation of the mean number of days with a certain degree of cloudiness in Brno in the period 1799– 1812 according to Knittelmayer’s observations: 1 – clear or individual clouds, 2 – more or less overcast, 3 – completely overcast. Day

40

30 Number of days

Temperature [°C]

30 20 10 0

20

10

-10 -20

1

2

3

0

4

J

-30 J

F

M

A

M

J

J

A

S

O

N

F

M

A

M

J

J

A

S

O

N

D

D

Night

According to standard climatological observations of cloudiness, it is possible to classify individual days as clear (0–2/10 of cloud cover), as half-covered (2–8/10 of cloud cover) and overcast (8–10/10). This type of information does not quite fall into line with Knittelmayer’s observations of the state of the sky, which specify days as divided into those with a clear sky and individual clouds, more or less overcast, quite overcast and cloudy. He used the same system for the night. The annual variation of the number of days with certain degrees of cloudiness in Brno in the period 1799–1812 is shown in Figure 7. The category “clear and individual clouds” shows its lowest values in the course of the day from November to February, the highest number of these days falling to May and August–September. At night the number of these cases increases. A practically reverse annual variation can be recorded for “quite overcast and cloudy” skies, with the highest frequency of such days from November to February. On the other hand, cases that correspond to half-covered sky according to Knittelmayer are most frequent in June and July. The interannual variability of the number of days with a certain degree of cloudiness in Brno, with the differentiation into day and night, is evident from Figure 8.

10

J

F

M

A

M

J

J

1

A

2

S

O

N

D

3

Figure 8. The fluctuation of the annual number of days with a certain degree of cloudiness in Brno in the period 1799– 1812 according to Knittelmayer’s observations: 1 – clear or individual clouds, 2 – more or less overcast, 3 – completely overcast. Day 200

Number of days

Cloudiness

20

0

150 100 50 0 1799

1801

1803

1805

1807

1809

1811

Night 200

Number of days

The annual variation of air temperature in Brno in the period 1799–1812 according to Knittelmayer’s measurements is characterised on the one hand by the mean temperature, on the other by the mean of the daily maxima and minima and their absolute monthly extremes (Figure 6). In comparison with modern measurements at the Brno-Tuřany station in the period 1961–1990 it follows that in all months, with the exception of March, Knittelmayer recorded higher temperatures. The smallest differences (0.1–0.4°C) correspond to January–February and April, and in the remaining months they were from 0.6°C (June) to 2.0°C (August).

Number of days

30

150 100 50 0 1799

1801

1803

1805

1

2

1807

1809

1811

3


|

65

Wind For each day of the month, Knittelmayer recorded the prevailing wind direction, any second direction and the force of the wind, dividing it into strong, mild and weak. Relative frequencies of wind directions for the year and the season in the years 1799–1812 are shown in Figure 9, completed by his observations of the wind force. A dominance of airflow from the west to north quarter is evident (with northerly prevailing except in winter), as is a south-easterly airflow (in winter together with an east-south-east wind). Among the prevailing wind directions in the individual years, the most frequently represented was the south-easterly (1801, 1803, 1804, 1806, 1807, 1810) followed by the northerly (1805, 1808, 1809, 1812) and the north-westerly (1799, 1802, 1804). On one occasion each, the most frequent winds were the north-north-westerly (1800) and the east-south-easterly (1811). The second most frequented direction over three years were the northerly and the south-easterly, over two years the west-north-westerly and the north-north-westerly, and in one year the directions westerly, north-westerly, eastsouth-easterly and south-south-easterly. The highest percentage representation of strong winds is evident in spring. On the whole, this information correlates well with the observations of the Brno-Tuřany meteorological station for the period 1961–1990, when the windiest period occurred from February to May with the highest mean wind speed in April. With respect to its position, on the Tuřany terrace outside town, this station is less representative for wind directions in the city itself. In the above thirty years, the prevailing directions were north-westerly (April, June– September), east-north-easterly (February–March, October– December), both together in May and south-easterly in January (Brázdil et al., 2005).

the number of recorded episodes being higher, particularly in the case of weak or short-term precipitation, the daily total of which might not be measurable. The annual numbers of precipitation days recorded by the two authors correlate almost equally in terms of annual sums of precipitation measured (0.83 for Melzer’s values and 0.84 for Knittelmayer’s). Figure 9. Relative frequencies [%] of the prevailing wind direction and force for the seasons and the year according to Knittelmayer’s observations in Brno for the period 1799–1812. Summer

Winter N

N

NNW

NNW

NNE

NW

NE

WNW

ENE

WNW

E

W

W

5

WSW

SSW

|


5

Autumn N

NNW

NNW

NNE

NW WNW W

5

ENE

WNW

E

W

ENE E

5

SSW

SSE

ESE

10

SW

SE

15

NE

WSW

ESE

10

NNE

NW

NE

SSW

SSE S

N

SW

SE

15

SSW

Spring

WSW

ESE

10

SW

SSE

15 S

15 S

S

SE SSE

Year N NNE

NW

Precipitation days

66

ENE E

SE

NNW

Within his meteorological phenomena, Knittelmayer used special marks for every day to record the variation and properties of clouds, the rainy or snowy character of the day, rain or snowfall, fog and foggy circumstances, frost, thunderstorms and heat lightning. The number of precipitation days is worthy of attention. It may be divided into days with liquid, mixed or solid precipitation. A maximum in December and a minimum in September were typical of the annual variation in the mean number of precipitation days in the period 1799–1812 in Brno (Figure 10). Knittelmayer recorded comparable high frequencies of precipitation days in January–April, June– July and November. At Brno-Tuřany station the highest number of precipitation days in 1961–1990 occurred in November–February with the main maximum in December, while the secondary maximum culminated in June. On the other hand, the main minimum occurred in October. Measurements of precipitation in Brno are available from 1803 onwards, made by Zacharias Melzer (Brázdil et al., 2005), and Knittelmayer’s records may prove a suitable supplement to them. Knittelmayer’s annual numbers of precipitation days are, however, higher than the numbers stated by Melzer (Figure 11). This is evidently connected with the fact that five observations per day inevitably led to

NE

WSW

ESE

10

SW

NNE

NW

NE

WNW

ENE

W

E

slight 5

WSW

ESE

10

SW SSW

mild strong

SE

15

SSE S

Figure 10. Annual variation in the mean number of precipitation days in Brno in the period 1799–1812 according to Knittelmayer’s observations and the Brno-Tuřany meteorological station in the period 1961–1990: 1 – liquid precipitation, 2 – mixed precipitation, 3 – solid precipitation.

Figure 11. Fluctuation of the annual number of precipitation days in Brno for the period 1799–1812 according to Knittelmayer (1) and Melzer (2). Knittelmayer’s data for 1799 and 1812 are incomplete. 250

Precipitation days

1

2

200 150 100 50 0 1799

1801

1803

1805

1807

1809

1811

Meteorological singularities The fact that Knittelmayer’s legacy includes continuous daily records of meteorological elements also makes it possible to address the study of meteorological singularities, such as calendar-bound deviations of a given meteorological element from its mean long-term variation. Singularities were studied for air pressure, air temperature and numbers of precipitation days using methodology applied by Řezníčková et al. (2006), which makes it possible to evaluate statistically significant deviations in the annual variation of those elements. In Figures 12–14 appear annual variations of three processed elements, both smoothed by threeday running averages and sixty-day low-pass filter, and with values smoothed by three-day running averages with

Table 2. Meteorological singularities in air pressure, temperature and precipitation days in Brno for 1799–1812 according to Knittelmayer’s measurements. Key: MD – H – L – C – T – W – D –

mean deviation higher pressure lower pressure cold warm wet dry

marking intervals of reliability and statistically significant singularities. In the case of air pressure, in the period 1799–1812 in Brno a total of 32 singularities were found with a mean duration of 2.8 days, among which a significant deviation was recorded on only one day in six cases, with the duration of the longest singularities not exceeding six days (Table 2, Figure 12). The best-known singularity expressed in air pressure, the “Indian summer”, is indicated by two peaks of air pressure around mid-September (14–19 September) and the beginning of October (2–6 October). These, however, are overcome in terms of magnitude by several further cases from November to March. The most conspicuous negative singularities corresponding to low air pressure are bound to the period December–February. For air temperature in the period 1799–1812 in Brno, a total of 35 singularities were found, again with a mean duration of 2.8 days (Table 2, Figure 13). In 10 cases a statistically significant deviation from the average was recorded on only one day and in 16 further cases for only 2–3 days. With the exception of the longest singularity, lasting 7 days from 28 July to 3 August, corresponding to “high summer”, further singularities lasting 5–6 days were bound solely to the period from the last days of August to the end of the year, including both positive and negative deviations. Above all worthy of mention is the “Christmas thaw” (24–29 December), which exhibited the highest positive temperature deviation (1.7°C) and which was preceded by the most conspicuous negative singularity (–2.0°C) on 17–22 December.

Pressure No Type

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

H H L L H H L H L H H L H L H H H L L L L L H L H L H H L L H L

Period

2 Jan 4–8 Jan 12–14 Jan 20–22 Jan 2 Feb 9 Feb 12–16 Feb 19–21 Feb 26–27 Feb 12–13 Mar 23–27 Mar 29–30 Mar 4–6 Apr 11–12 Apr 18–19 May 8–10 June 25 June 2–5 July 3–4 Aug 9–12 Aug 20–21 Aug 1 Sep 14–19 Sep 23–24 Sep 2–6 Oct 8–11 Oct 14–15 Nov 19 Nov 2–6 Dec 8–9 Dec 19–21 Dec 26–29 Dec

Temperature MD [hPa]

1.7 3.2 –3.4 –3.9 1.7 1.5 –3.5 3.0 –1.9 2.2 2.8 –2.1 2.2 –2.4 1.0 1.4 1.2 –1.2 –0.8 –1.2 –1.7 –1.0 2.1 –1.8 2.5 –2.6 2.6 2.0 –2.6 –2.4 3.5 –3.2

No Type

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

Period

C 6 Jan C 12–13 Jan T 21–22 Jan C 27 Jan T 27 Feb T 3–4 Mar T 19–20 Mar C 23–26 Mar C 12 Apr T 16 Apr C 25 Apr T 2–4 May T 9–10 May T 22–23 May T 28–30 May C 2 June T 6–8 June C 18–20 June C 23–25 June T 9–10 July C 17–19 July T 24 July T 28 July–3 Aug C 6 Aug C 16–18 Aug T 28 Aug–1 Sep T 7 Sep T 6–9 Oct C 11–15 Oct T 24–29 Oct C 2–3 Nov C 13–16 Nov T 7–9 Dec C 17–22 Dec T 24–29 Dec

Precipitation MD [°C]

No Type

–1.3 –1.2 1.2 –1.2 0.7 0.8 1.4 –1.4 –0.8 1.0 –0.8 1.3 0.9 0.8 0.9 –0.8 1.0 –0.9 –0.9 1.1 –0.8 0.6 1.0 –0.7 –0.9 1.0 0.8 0.8 –1.0 1.3 –1.2 –1.2 1.1 –2.0 1.7

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

D D W W D W D W W D W D W W W D D W D W D W W D W D W D W D D D W D W

Period

MD of days

4–5 Jan 8 Jan 12–13 Jan 17 Jan 30 Jan–2 Feb 15–16 Feb 19–22 Feb 26–28 Feb 1 Mar 3–4 Mar 8–10 Mar 18–20 Mar 30 Mar 11 Apr 20–21 Apr 2–3 May 5–8 May 14–15 May 27–29 May 4–5 July 8–9 July 23–24 July 15–16 Aug 30 Aug 2 Sep 7–8 Sep 11–12 Sep 16 Sep 9–11 Oct 15 Oct 22 Oct 29 Nov 2–5 Dec 12–14 Dec 26–27 Dec

–2.2 –1.8 1.9 1.6 –2.3 1.6 –2.1 2.7 1.8 –1.5 2.1 –3.3 2.1 1.7 1.7 –1.7 –1.9 3.0 –2.4 2.1 –1.7 1.8 1.8 –2.1 2.3 –1.7 2.0 –1.4 1.9 –1.4 –1.9 –1.4 2.5 –2.3 1.7


|

67

Figure 12. Annual variation of mean daily air pressure in Brno for the period 1799–1812: a) three-day running means smoothed by 60-day low-pass filter, b) three-day running deviations with intervals of reliability (80 %) and marking of singularities with a duration of ≥ 2 days (for numbers, see Table 2).

Figure 13. Annual variation of the mean daily air temperature in Brno for the period 1799–1812: a) three-day running means smoothed by 60-day low-pass filter, b) three-day running deviations with intervals of reliability (80 %) and with marking of singularities with a duration of ≥ 2 days (for numbers, see Table 2).

68

|


Figure 14. Annual variation of the number of precipitation days in Brno for the period 1799–1812: a) three-day running means smoothed by 60-day low-pass filter, b) three-day running deviations with intervals of reliability and with marking of singularities with a duration of ≥ 2 days (for numbers, see Table 2).

In the case of the number of precipitation days in Brno in the period 1799–1812, a total of 35 singularities were recorded with a mean duration of 2.1 days (Table 2, Figure 14). In 11 cases the significant deviation referred to only one day and only four times did it reach 4 days, three times as a drier episode (30 January–2 February, 19–22 February, 5–8 May) and once as a wetter one (2–5 December). The best-known precipitation singularities, bound to the individual waves of the so-called “European summer monsoon” (e.g. Brázdil, 1982) thus remained limited to only three two-day episodes in July and August in the period studied. From the preceding analysis it follows that the number of singularities found for air pressure, air temperature and precipitation activity is more or less comparable; the mean duration of a singularity drops from 2.8 days for air pressure and temperature to 2.1 days for precipitation. In terms of the best-known weather singularities traditionally associated with the Czech Republic (see e.g., Souborná studie, 1969; Brázdil et al., 1999; Řezníčková et al., 2006), the situation for Brno in the period 1799–1812 was as follows: • “deep winter” – was expressed only on the days 12– 13 January, bound to a negative anomaly of air pressure, not to the prevalence of an anticyclonic regime of weather • “May cooling” – usually connected with the “Ice Men” did not appear at all; on the contrary, in four cases

•

•

•

•

there occurred a two- to-three-day positive temperature anomaly “European summer monsoon” – from precipitationexpressed waves (see e.g. Brázdil, 1982; Bissolli, 1991), higher precipitation activity in June was altogether absent, with three waves falling to the days 4– 5 July, 23–24 July and 15–16 August; a negative anomaly of air pressure on 2–5 July corresponded to the first wave in July “high summer” – this was a clearly perceptible singularity in the case of air temperature (28 July–3 August) which, however, did not find expression in a significant rise of air pressure or fall in precipitation activity “Indian summer” – was well expressed by two episodes of high air pressure on 14–19 September and 2–6 October, which were, however, not reflected in temperature singularities (positive deviation only for 6–9 October) and in a fall in precipitation activity “Christmas thaw” – a conspicuous positive temperature singularity (24–29 December) was supported by a negative anomaly of air pressure (26–29 December) and increase in precipitation activity (26–27 December); this singularity became conspicuous after the preceding significant cooling (17–22 December), connected with a prevailing anticyclonic weather regime (positive anomaly of pressure on 19–21 December).


|

69

CONCLUSIONS Knittelmayer’s meteorological records open a history of more than two hundred years of meteorological observations in Brno. They constitute exceptionally valuable data from a period in which meteorological observations were carried out in only a few places. In the Brno environment they were much appreciated, as is evident from an expression of approbation from Karel Josef Jurende, a wellknown Moravian pioneer of meteorology, in his description of the climate of Moravia from 1813 (“The weather and climate of Brno has been observed most accurately and completely for 13 years by the honourable Captain Knittelmayer … Only he, himself, is perhaps able to say anything premeditated and based upon experience about the climate and weather in Brno thanks to his arduous observations. May his precious observations never be lost! – May many follow his example!” – see Brázdil, Valášek, 2003). Later, Kassián Hallaschka, the professor at Prague university who wrote an expert opinion for the purchase of Knittelmayer’s observations for the I. R. Moravian-Silesian Economic Society said: “With what diligence, with what combination of ability, and finally with what erudition these data are collected, I am unable to describe adequately. With real pleasure, I have gone repeatedly through this paper and admired the investigations and the work of this man, who compiled it according to fixed principles and enlightened ideas in such a way that one cannot wish for more.” – see Brázdil et al., 2005). Particularly valuable is the fact that, unlike some other Brno observations (such as Kassián Hallaschka, Zacharias Melzer – see Brázdil et al., 2005) the records have remained in the form of daily observations. At the same time, it is necessary to evaluate its importance for compiling pressure and temperature series for Brno since 1799 up to the present. Setting aside the pure facts, the records themselves are of cultural and historical importance, illustrating as they do the contemporary role of meteorology, something that must not be neglected. Acknowledgements The work on this study was supported from the financial means of the Grant Agency of the Czech Republic, grant No. 205/05/0858. Thanks are also extended to the Archives of the City of Brno for making accessible the manuscript of the meteorological observations of Ferdinand Knittelmayer. Bořivoj Herzlík (Brno) was responsible for the English translation and Tony Long (Svinošice) for English style corrections.

REFERENCES Alexandersson, H., 1986, A homogeneity test applied to precipitation data. Journal of Climatology, 6, 6, 661–675. Bissolli, P., 1991, Eintrittswahrscheinlichkeit und statistische Charakteristika der Witterungsregelfälle in der Bundesrepublik Deutschland und West-Berlin. Berichte des Instituts für Meteorologie und Geophysik der Universität Frankfurt/Main 88, Frankfurt am Main, 566 pp.

70

|


Brázdil, R., 1982, Precipitation singularities in the variation of diurnal sums of precipitation in the summer season on the territory of the Czechoslovak Socialist Republic (CSSR). Scripta Fac. Sci. Nat. Univ. Purk. Brun., 12, 4– 5 (Geographia), 169–202. Brázdil, R.–Macková, J.–Sviták, Z.–Valášek, H.–Hradil, M., 2002a, Nejstarší moravská meteorologická měření v Telči od Františka Aloise Maga z Maggu z let 1771–1775. Meteorologické zprávy, 55, 2, 50–60. Brázdil, R.–Štekl, J.–Budíková, M., 1999, Povětrnostní singularity. In: Brázdil, R., Štekl, J. a kol.: Klimatické poměry Milešovky. Academia, Praha, 237–257. Brázdil, R.–Valášek, H., 2001, Popis klimatu Moravy od Kryštofa Passyho z roku 1797. Geografie – Sborník České geografické společnosti, 106, 4, 234–250. Brázdil, R.–Valášek, H., 2002, Meteorologická měření a pozorování v Zákupech v letech 1718–1720. Geografie – Sborník České geografické společnosti, 107, 1, 1–22. Brázdil, R.–Valášek, H., 2003, Karel Josef Jurende a jeho popis klimatu Moravy. Vlastivědný věstník moravský, 55, 2, 129–141. Brázdil, R.–Valášek, H.–Macková, J., 2005, Meteorologická pozorování v Brně v první polovině 19. století. Historie počasí a hydrometeorologických extrémů. Archiv města Brna, Brno, 452 pp. Brázdil, R.–Valášek, H.–Sviták, Z.–Macková, J., 2002b, History of Weather and Climate in the Czech Lands V. Instrumental meteorological measurements in Moravia up to the end of the eighteenth century. Masaryk University, Brno, 250 pp. Brodesser, S., 2005, Staletími podél řeky Svitavy. Moravské zemské muzeum, Brno, 112 pp. Buse, G. H., 1816, Nekrolog. Hauptmann Knitlmayer, mit einigen durch das erneuerte Andenken an den Verewigten veranlaßten Ideen. Hesperus, 8, 57–64; 12, 91–96. Dřímal, J., 1956, Archiv města Brna. Průvodce po fondech a sbírkách. Archivní správa ministerstva vnitra, Praha, 268 pp. Knittelmayer, F., 1800a, Einiges über den Gang und Wechsel der Witterung überhaupt, und der von Brünn insbesondere. Patriotisches Tageblatt, 17, 65–68. Knittelmayer, F., 1800b, Allgemeine Resultate meiner meteorologischen Beobachtungen vom 21. März an bis 21. Mai 1800. Patriotisches Tageblatt, 18, 71–72 + Annex. Knittelmayer, F., 1800c, Fortgesetzte Meteorologische Beobachtungen in den 4 Sommermonaten vom 22. Mai bis 22. September zu Brünn 1800. Patriotisches Tageblatt, 92–93, 409–412. Knittelmayer, F., 1801, Fortgesetzte Meteorologische Beobachtungen in den sechs Wintermonaten von 1800– 1801 zu Brünn. Patriotisches Tageblatt, 111, 594 + Annex. Meteorologische Beobachtungen in Brünn 1799–1812. Archiv města Brna, fond IV A 6. Meteorologická a přírodovědná pozorování, rkp. čís. 7263. Nachricht von den Witterungsbeobachtungen, welche die kais. koen. Patriotisch-Oekonomische Gesellschaft in den Kreisen Böhmens veranstaltet hat. Verfasst von Professor Aloys David. 1ste Lieferung vom Jahre 1817–1819. Prag, 1825. 2te Lieferung vom Jahre 1820 u. 1821. Prag, 1826. Neue Schriften der kaiserl. königl. patriotisch-ökonomischen Gesellschaft im Königreiche Böhmen. Band I–X. Gottlieb Haase Söhne, Prag 1828–1847. Pejml, K., 1975, 200 let meteorologické observatoře v pražském Klementinu. Hydrometeorologický ústav, Praha, 78 pp.

Řezníčková, L.–Brázdil, R.–Tolasz, R., 2006, Meteorological singularities in the Czech Republic in the period 1961– 2002. Theoretical and Applied Climatology, DOI 10. 1007/s00704-006-0253-5. Seydl, O., 1952, Meteorologie na pražské hvězdárně a v Čechách v minulosti (1752–1839). In: Hanzlíkův sborník. K sedmdesátým narozeninám. Publikace řady C, svazek VI. Nákladem Státního meteorologického ústavu, Praha, 13–51. Souborná studie. Podnebí Československé socialistické republiky. Hydrometeorologický ústav, Praha 1969, 357 pp. Strnadt, A., 1788, Physikalischer Witterungskalender. K. k. Normalschulbuchdruckerey, Prag, 152 pp. Strnadt, A., 1789, Physikalisches Taschenbuch auf das Jahr 1789. Für Freunde der Oekonomie und Witterungskunde. K. k. Normalschulbuchdruckerey, Prag, 183 pp. Strnadt, A., 1790, Chronologisches Verzeichniss der Naturbegebenheiten im Königreiche Böhmen vom Jahre Christi 633 bis 1700 mit einigen ökonomischen Aufsätzen samt der periodischen Witterung auf das Jahr 1790. Gerlische Buchhandlung, Prag, 259 pp.

Strnadt, A., 1791, Meteorologische Resultate der in Prag und einigen andern Orten in Böhmen gemachten Luftbeobachtungen und andern Erscheinungen. In: Neuere Abhandlungen der k. Böhmischen Gesellschaft der Wissenschaften. Erster Band. J. V. Degen, Wien und Prag, 235–256. Strnadt, A., 1793, Beyträge zu der Geschichte des Winters im Christmonate 1788. In: Mayer, J., ed.: Sammlung Physikalischer Aufsätze, besonders die Böhmische Naturgeschichte betreffend, von einer Gesellschaft Böhmischer Naturforscher. Dritter Band. In der Waltherischen Hofbuchhandlung, Dresden, 39–88. Valášek, H.–Brázdil, R.–Sviták, Z., 2001, František Alois Mag z Maggu a jeho nejstarší přístrojová meteorologická měření na Moravě. Časopis Matice moravské, 120, 1, 37–65. Verhandlungen und Mittheilungen der k. k. patriotisch-ökonomischen Gesellschaft im Königreiche Böhmen. Band I– II. Commission der J. G. Calveschen Buchhandlung, Prag 1849–1850.


|

71

Příloha č. 10

EARLY INSTRUMENTAL METEOROLOGICAL OBSERVATIONS IN THE CZECH LANDS II: ANDREAS STERLY, JIHLAVA, 1816–1840 (1844) RUDOLF BRÁZDIL1, LADISLAVA ŘEZNÍČKOVÁ1, HUBERT VALÁŠEK2 1 2

Institute of Geography, Masaryk University, Kotlářská 2, 611 37 Brno, Czech Republic Moravian Land Archives, Žerotínovo nám. 3–5, 656 01 Brno, Czech Republic

Systematic meteorological observations associated with the meteorological activities of the Imperial-Royal Moravian-Silesian Economic Society in Brno began in many parts of Moravia and Silesia in the first third of the 19th century. Outstanding amongst them were the activities of the magistrate councillor Andreas Sterly at Jihlava, whose daily observations are preserved (with gaps) for the years 1816–1826 and can be found in the archives of the Society. Monthly summaries for the period 1817–1840 (precipitation 1821–1844) appeared in Pokorny (1852). This current paper gives a detailed description of these measurements, including also the results of a statistical analysis of air pressure and temperature, precipitation, cloudiness, wind strength and direction, fog, hail and thunderstorms. V návaznosti na meteorologické aktivity c. k. Moravskoslezské hospodářské společnosti v Brně začala na mnoha místech na Moravě a ve Slezsku v první třetině 19. století systematická meteorologická pozorování. Mezi nimi vynikají především pozorování magistrátního rady v Jihlavě Andrease Sterlyho, dochovaná na bázi denních pozorování (s mezerami) z let 1816–1826 v archivních materiálech Společnosti a na úrovni měsíčních hodnot z období 1817–1840 (srážky 1821–1844) v publikaci Pokorneho (1852). Práce podává podrobnou popisnou charakteristiku těchto měření a uvádí výsledky statistické analýzy tlaku a teploty vzduchu, srážek, oblačnosti, směru a síly větru, mlhy, krupobití a bouřek. Key words: early instrumental measurements, pressure, temperature, precipitation, cloudiness, wind, meteorological phenomena, Andreas Sterly, Jihlava

INTRODUCTION

• Jihlava, July 1816–December 1826, Andreas Sterly,

The activities of the Imperial-Royal Moravian-Silesian Economic Society were important for the development of early instrumental meteorological observations in Moravia and Silesia. The establishment of the Society’s Meteorological Section in 1815 was a critical development in this respect. The Meteorological Section produced detailed printed instructions for observers and offered to buy barometers and thermometers for their use. In return, the observers were asked to send their observations to the Meteorological Section in Brno. Some of those records remain in the Society archives (see Meteorologická pozorování). According to those weather records, besides Brno observations (see Brázdil et al., 2005, 2006), other began to be made in some other places (Fig. 1): • location – Frýdek, period of observations – July–December 1815, observer – Martin Ehrmann • Klášterní Hradisko, August 1815–June 1816, Joseph Bayer, supervisor of the land cadaster • Jeseník, January 1816–November 1818, Jungnikel, syndicus

• Nové Město na Moravě, June 1817–June 1824, Johann

magistrate councillor Merfort, economic director • Opava, June 1819–December 1821, March 1823, Carl

• • • • • • •

Biela, secondary school principal, and his son Franz Biela, builder Leskovec nad Moravicí, January 1821–June 1828, Anton Saliger, economist and school supervisor Rosice, January–March 1820, 1821–1822, 1824–1827, Johann Jurende, mining office supervisor Opava, 1822–1827, Robert Jeník, grammar school principal Uhřice, May 1822, an observer is not mentioned Olomouc, May–June 1822, Andreas Baumgartner, Professor of physics at the lyceum Ivanovice na Hané, June 1822–1825, Andreas Altmann, official Žďár nad Sázavou, November–December 1822, Anton Prziborsky (Přiborsky), higher administrative official (Brázdil et al., 2005).

Meteorologický časopis, 10, 2007, 3–12

|

3

Figure 1. Meteorological stations in Moravia and Silesia, the readings from which were sent to the I.-R. MoravianSilesian Economic Society. The first year of records is given in brackets. Uhřice is not shown due to uncertainties regarding its location (Brázdil et al., 2005).

Unfortunately, some of the above records are only limited to a few months and there is no reliable indication of the true length of period over which observations were made. The present contribution links the analysis of meteorological observations of Ferdinand Knittelmayer in Brno of the years 1799–1812 (Brázdil et al., 2006) with that of meteorological observations of the magistrate councillor Andreas Sterly at Jihlava whose observations cover the longest span of those available. Sterly’s daily records for the period 1816–1826 remained partly preserved in the Moravian Land Archives in Brno and for the period 1817– 1840 (1844) are preserved in the form of monthly values published by Pokorny (1852).

ANDREAS STERLY Andreas Sterly (Fig. 2) was, during his day, one of the most important meteorological observers (d’Elvert, 1853; Wurzbach, 1879; Schwab, 1935). He was born on 21 November 1779 at Jihlava. At the age of five his father sent him to the primary school and in 1790 to the secondary school, from where, in the same year, he passed to the grammar school for five years. After grammar school Sterly spent another five years at Vienna University, where for the first three years he devoted his time to philosophy and for the remaining two years to legal sciences. He also attended lectures as diverse as politics, heraldry and numismatics. At the same time he studied French, Italian, English and also Czech (in the last year of his studies at Olomouc). At the beginning of October 1803 he began his career at Jihlava as court probationer and from 1807 as secretary. He soon applied his knowledge of French during the occupation of Jihlava by Napoleon’s army assisting in the negotiations concerning the community. In 1810 Sterly obtained a job at the town hall of Jihlava and three years later he was appointed magistrate councillor. In 1811 he got married and eventually became father of six children. In 1816 he started to

4

|


devote his time to meteorology, astronomy and electricity. In the same year he began his regular meteorological observations, which he carried out for more than 24 years. In 1820 he became the principal of the provisional land tax office for the Jihlava region. At that time he developed an interest in mineralogy, history and geography, and began working on the history of the town of Jihlava. Between 1828 and 1840 he worked as economics official before retiring in 1845. The fruitful life of Sterly ended at Jihlava (Fig. 3) at the age of 73 on 26 December 1852.

Figure 2. Andreas Sterly (1779–1852) and his signature (reproduction, State District Archives, Jihlava).

Figure 3. A view of Jihlava from the south, 1849. Drawing and lithography J. W. Zwettler, steel engraving F. Zastiera, print J. Rippl Jihlava (State District Archives, Jihlava).

Figure 4. An example of Sterly’s records of air temperature and the graph of the variation of mean monthly temperatures for March 1819 to February 1820 (Moravian Land Archives, Brno).

METEOROLOGICAL OBSERVATIONS BY ANDREAS STERLY As noted already, Andreas Sterly’s meteorological observations were connected with the activities of the Meteorological Section of the I.-R. Moravian-Silesian Economic Society in Brno, even though he was not mentioned among its original members, and only in 1819 did he become a corresponding member (Schwab, 1935; Brázdil et al., 2005). Sterly’s meteorological records of Jihlava from 1818 had already been used by the head of the Meteorological Section Dr. Josef Steiner, together with the observations from Brno and Opava in a lecture for the plenary session of the I.-R. MoravianSilesian Economic Society on 6 May 1819 (Steiner, 1819). Sterly’s meteorological observations are now preserved in the Moravian Land Archives in Brno and are stored in the section G 82 Economic Society Brno 1769–1937, sign. IV/3e,

Meteorological observations (see Meteorologická pozorování). In the box number 206 the following items are deposited: • monthly reports of daily meteorological observations

for July and September–December 1816, March–June 1817, April–August 1818, January–February and July– October 1819, April–September and December 1820, complete years 1821, 1822, 1824, 1825 and 1826, • annual text and tabulated weather overviews for the

period 1 March 1819 to 29 February 1820, 1 March 1820–28 February 1821 with additional material for the years 1822 and 1824 to 1826 (completed tables and text and graphical summaries of the annual variation of mean air temperature, pressure and relative humidity also survive for the first mentioned period – Fig. 4), • mean monthly values of air pressure and temperature

for the period 1817 to 1820.


|

5

Figure 5. Monthly report of meteorological observations of Andreas Sterly of Jihlava, October 1816 (Moravian Land Archives, Brno).

It is not known whether the reports for those months of missing observations were sent to the Society or whether they were lost at a later date. Sterly states, however, in his autobiography (Schwab, 1935) that from 1817 he sent his observations to the Meteorological Section every month. The monthly forms of meteorological observations conform largely to the structure recommended in the instructions of the Meteorological Section (Fig. 5). In the heading described as “Meteorologische Beobachtungen (Witterungsbeobachtungen), angestellt zu Iglau im Monate …” the word “Iglau” was initially completed in Sterly’s hand, and from 1819 the heading was printed in its entirety. The first two columns of the form were set aside for the day in the month and the hour of observation. In the third column were written the values of the air pressure (later he also added the temperature at the barometer) and in the fourth air temperatures. The following column contained

6 | Meteorologický časopis, 10, 2007

data for wind direction and strength. The next column was devoted to the observation of clouds and the seventh to the state of the weather (e.g. rain, fog, variable). The form ended with columns for notes derived from observing the Moon, for special events and general notes and finally for other measurements such as precipitation and air humidity. There is no doubt that Sterly observed not only in the months for which the monthly reports are now missing, but also for the period after 1826. According to Pokorny (1852) Sterly’s observations should exist for the years up to 1840 as part of quarterly reports of the Jihlava regional physicist. Sterly himself published a summary overview of his meteorological observations in the journal ‘Moravia’ (Sterly, 1846). This analysis was partly taken over by d’Elvert (1850) in the description of the history of Jihlava and to a greater degree by Pokorny (1852), who not only stated that he was able to use the report by Sterly (“durch die gewohnte Liberalität des Herrn Sterly, der mir sein Elaborat zur freiesten Benützung überliess”), but also that he published monthly summaries of meteorological observations for 1817 to 1840 (1844). It can be deduced from Sterly’s summary meteorological observations from 1 March 1819 to 29 August 1820 that he used an outdoor thermometer and the siphon barometer No. 34, made by Kassian Hallaschka to provide for the needs of the Meteorological Section (for details see Brázdil et al., 2005). According to the note in the monthly report, he first used these instruments on 13 December 1816. Sterly adds that an additional 80-part thermometer was placed in the shade in free air on the northern side of the building at a height of 4 ells (7.59 m) above the ground. In the summary of monthly observations for 1822 it is also possible to find a note making reference to the measurement of precipitation, which he found according to the volume of the caught water. Sterly, however, also measured the relative air humidity with the hair hygrometer according to Saussure’s hundred-part scale, but these data have not survived in the form of a continuous series of monthly values. As for the place where the observations were made, it can be assumed that it was either at the town hall or at Sterly’s houses located close to the main town square (Fig. 6). In the above mentioned summaries Sterly described the properties of the surroundings of the point of observation as “old mountains rising to the height of 263– 265 ells [499–503 m] above the sea, with visible forests on the southwest, west, northwest and north”. At the same time he also stated latitude (49°23´29´Ń) and longitude (33°16´00´´ east of Ferro, i.e. 15°36´14´É).

Figure 6. Jihlava and its surroundings in the map of 1825 (State District Archives, Jihlava). The possible locations of Sterly’s observatory: 1 – town hall (now Masaryk Square), 2 – house No. 145 (now Znojemská Street No. 5) where he stayed at the time of his marriage, 3 – house No. 261 (now Havířská Street No. 2, Sterly bought this property in 1821), 4 – house No. 65 (the house of Sterly’s wife, now Masaryk Square No. 15).

THE CHARACTER OF THE WEATHER AND CLIMATE IN JIHLAVA, 1817–1840 As the daily meteorological observations of Andreas Sterly cover only a part of the period 1816–1826, the summary of monthly values of meteorological elements and events for the period 1817–1840, published by Pokorny (1852) were used as the basis of this analysis. These summaries provide the following statistics: • the highest, the lowest and the mean air temperature on the Réaumur scale • the highest, the lowest and the mean air pressure in Vienna inches, lines and points • the volume of precipitation in cubic inches and lines, but only for the period 1821 to 1844 • observations of cloud cover and the number of cases with fog, hail or thunderstorm • observation of wind strength using the scale calm, weak, medium and strong • wind direction on an eight-point compass. From the point of view of statistical analysis of Sterly’s observations, a disadvantage is provided by the variable time of observation, which, according to the instruction of

the Meteorological Section should have been at sunrise (dependence of course on the season of the year and varying between 0400–0800 hours), at the time of its culmination (1200–1300 hours) and at sunset (1700–2100 hours). As a consequence in the years 1817 to 1819 observation times were 0800, 1500 and 2000 and/or 2100 hours but from July to the end of October 1819 the morning time changed by a quarter of an hour increments between 0700–0800 hours. The afternoon times varied between 1400 and 1500 hours, and the evening times drifted from 1954 to 1655 hours. In the years 1821 and 1822 the morning time of observation fluctuated between 0700 and 0800 hours, but in the period 1824 to 1826 it was consistently 0800 hours. The afternoon time, with the exception of 1330 hours during December 1821, was stabilised to 1400 hours, whereas the evening one shifted from 1800 (as a rule October–March) to 1900 (April, September) and 2000 hours (May–August). The fluctuating times of observation distort the statistical picture created by the daily mean values for air pressure and temperature; the degree of distortion varying itself during the year and within the individual months. Therefore stations Prague-Klementinum, Brno and Vienna-Hohe Warte were selected to act as reference points with which Sterly’s data were compared with the aim of removing such error values and consequently of homogenising the Jihlava series using the Standard Normal Homogeneity Test (SNHT) according to Alexandersson (1986). Sterly’s period from 1817 to 1840, when compared with the standard period 1961–1990, appears to be colder in the months of the winter half-year (particularly January to March), the summer months being only slightly warmer (Fig. 7). With the exception of the winter, all months had lower pressure in comparison with the standard period. The years 1821 to 1844 were drier in February and May but wetter in July and November to January. In other months the differences between the two periods are less distinct: in Brno wetter patterns prevailed from August to October, but in Prague these months with exception of September were drier. Air pressure The values of air pressure were given in Vienna inches, lines and points, and it was necessary to convert them to the SI units (1 inch = 12 lines = 26.34 mm, 1 line = 12 points = 2.195 mm). As Sterly also recorded the temperature at the barometer, it can be assumed that his values are reduced to the temperature of 0°C. The checking of the monthly values was based on the calculations of pressure differences with Prague, Brno and Vienna as well as on a comparison of the character of their annual variations year by year. Using these approaches it was possible to identify problematic monthly values taken from Pokorny (1852) which were consequently adjusted being calculated either from daily data or by use of linear regression methods. The Brno station was adopted as the reference data set because of the high correlation coefficients with those from Jihlava (Fig. 8b). After homogenisation the January pressure continued to present problems as a consequence of it providing the smallest correlation (r = +0.45) but no methods for improMeteorologický časopis, 10, 2007

|

7

Figure 7. Differences in mean air temperature (a), pressure (b) and precipitation (c) for the years 1817 to 1840 (and 1821 to 1844 for precipitation) and the reference period 1961 to 1990 for Brno (precipitation), Prague-Klementinum (temperature and precipitation) and Vienna-Hohe Warte (temperature and pressure) stations. 1

a Temperature [°C]

ving Sterly’s data were found. Conspicuously low correlations were also found in the data for May and June. Figure 8a shows the annual variation of monthly means of the homogenized air pressure with a maximum in October and minimum in April. This behaviour is consistent with pressure fluctuations of the Přibyslav station (536 m a.s.l.) for the period 1961 to 1990, this site being selected for comparison (Míková, Coufal, 1999). Months from the winter-half year in comparison with the summer half-year show higher variability expressed by a greater range of maximum and minimum values.

0

Prague - Klementinum

Air temperature

-2

The measured values of air temperature published by Pokorny (1852) were first converted from Réaumur degrees to degrees Celsius according to the relation 1°R = 1.25 °C. With a view to identifying and correcting erroneous observations Sterly’s monthly and annual data were calibrated against those for the sites noted above. The source of such anomalous data can be a lower number of observations in that month, the different observation times, errors in calculation (such as neglecting the sign) or printing errors. An example of neglecting the minus sign is found in the mean temperature of January 1826 with the value 7.3 °C in Pokorny (1852), compared with values of –5.7 °C in Prague and –6.1 °C in Vienna. As a result of high correlations for temperatures the Prague-Klementinum observatory was selected to provide the reference data set. Application of the SNHT procedures proposed by Alexandersson (1986) identified several months with inhomogeneities in 1823 and 1836 which were consequently adjusted. The homogenized Jihlava series shows high correlations with data from Prague-Klementinum with the lowest such value in September (Fig. 9b). But from comparison with the Jihlava series from 1901–1950 it is concluded that Sterly’s temperatures are too high in all months of the year (Fig. 9a). Without any metadata for his measurements it is difficult to find reason for this disparity although the greater differences from April to September compared to months of winter half-year might indicate a radiation effect on the thermometer. Similarly the August maximum in the annual variations instead of usual July might reflect exposure and adjustment influences rather than genuine climatic change. From these reasons temperature data should be treated with caution.

3

As already stated, the volume of water from liquid or solid precipitation was measured in Vienna cubic inches and lines (1 cubic inch = 18.275 cm3, 1 cubic line = 10.576 mm3), although no reliable indication is given of the area of the collection orifice. Conversions of the original data based on Sterly’s stated one Vienna square foot (999.07 cm2) yield results in millimetre depths of precipitation that are clearly in error. For this reason monthly volumes were converted only to cm3 and it is these values that are used for further analyses.


Pressure [hPa]

J

2

F

M

A

M

J

J

A

S

O

N

D

b Vienna - Hohe Warte

1 0 -1 -2 -3 J

F

M

A

M

J

J

A

S

O

N

D

A

S

O

N

D

Precipitation [mm]

20

c

Prague - Klementinum Brno

10 0 -10 -20 J

F

M

A

M

J

J

Figure 8. a) Comparison of the annual variation of mean air pressure in Jihlava according to Sterly’s measurements (1) for the period 1817 to 1840 (3 – mean of absolute maxima/ minima, 4 – absolute maximum/minimum) and the Přibyslav meteorological station (2) in the period 1961 to 1990; b) correlation coefficients between monthly pressure series of Jihlava and Brno for the period 1817 to 1840 (all values are statistically significant at α = 0.05). 1000

a

Pressure [hPa]

980 960 940 920 1

2

3

4

900 J

F

M

A

M

J

J

A

S

O

N

D

F

M

A

M

J

J

A

S

O

N

D

1.0

b 0.8

Correlation

Precipitation

Vienna - Hohe Warte

-1

0.6 0.4 0.2 0.0 J

Figure 9. a) Comparison of the annual variation of mean air temperature in Jihlava according to Sterly’s measurements (1) in the period 1817 to 1840 (3 – mean of absolute maxima/ minima, 4 – absolute maximum/minimum) and the Jihlava meteorological station (2) in the period 1901 to 1950; b) correlation coefficients between monthly temperatures in Jihlava and Prague-Klementinum over the period 1817 to 1840 (all values are statistically significant for α = 0.05). 40

Temperature [°C]

30

a

20 10 0 -10 -20 -30 -40 J

F

M

A

M

J

J

A

S

O

N

D

F

M

A

M

J

J

A

S

O

N

D

1.0

Correlation

0.8 0.6 0.4 0.2

b

0.0 J

Figure 10. a) Comparison of the annual variation of relative precipitation for Jihlava according to Sterly’s measurements (1) in the period 1821 to 1844 and the Jihlava composite series (2) in the period 1961 to 1990; b) correlation coefficients between the monthly precipitation series of Jihlava with Prague-Klementinum (3, 4) and Jihlava with Brno (5, 6) in the periods 1821 to 1844 (3, 5) and 1961 to 1990 (4, 6). 15

a Precipitation [%]

1

2

10

5

0 J

1.0

F

M

A

M

J

J

A

S

O

N

D

F

M

A

M

J

J

A

S

O

N

D

b

0.8 Correlation

vember, January and March, secondary minima in December, April and September. These results were compared with those for the standard period 1961 to 1990, compiled by reference to data from sites at Jihlava-waterworks (1961 to 1969), Jihlava-transmitter (1970 to October 1984) and Hubenov-dam (November 1984 to 1990). These precipitation series were integrated and the homogeneity checked using SNHT procedures (Alexandersson, 1986). Its consequent annual variation shows main maximum in July and main minimum in February and March. Secondary precipitation maxima occurred in January and November, minima in October and December. Sterly’s precipitation measurements can be further verified by reference to their correlation coefficients with series from Prague-Klementinum and Brno for 1821 to 1844 with same coefficients derived from the Jihlava series noted above and these two stations for the period 1961 to 1990 (Fig. 10b). When comparing the second series correlations with those above, correlation coefficients were higher in January–February, July and September–December for Prague and with exception of March, April, June and July were higher in all the other months for Brno. The greatest difference was found in the correlations for January.

0.6 0.4 0.2 0.0 J

Figure 10a shows annual variation of mean relative monthly values calculated as a proportion of the monthly to annual total volumes. Over the period 1821 to 1844, Sterly’s data indicate a principal maximum in June and minimum in February. Secondary precipitation maxima occurred in No-

Cloudiness Sterly’s cloudiness observations are based on a four-term vocabulary: clear (“heiter”), little cloud (“wenig bewölkt”), much cloud (“sehr bewölkt”) and overcast (“trüb”), the frequencies of which can be counted and used as the basis for analysis. Whilst the categories ‘clear’ and ‘overcast’ demonstrate a general decrease in frequency over the study period, the remaining two, after an initial increase reveal a conspicuous break after 1827 (Fig. 11), when the number of cases classified as ‘little cloud’ dropped conspicuously. Such a pattern might suggest an inconsistency in the observational procedures. The annual variation of cloudiness shows features typical of the Czech Republic (Fig. 12). However, in the case of clear sky, Sterly recorded an unexpected maximum in February, which replaced the conventional maxima that now occurs towards the end of the summer and in autumn (August to October). On the other hand, the maximum of overcast days noted for December and January, and the general tendency to more frequent overcast conditions in the winter half of the year is normal. Cases of partial cloud cover (little or much cloudy weather) dominate between March and October. Sterly also observed the type of clouds, which he recorded at first using whole words but after July 1819 with the following abbreviations: Cr – Cirrus, Cm – Cumulus, St – Stratus, Cc – Cirrocumulus, CrS – Stratocumulus (in Sterly’s terminology “Cumulostratus”), Nim – Nimbostratus (according to Sterly “Nimbus”) and St/Cum – Cumulus below Stratus. In addition he also noted the state of the sky, the colour of clouds and direction of their movement. Cloud types were absent, obviously, when skies were clear (“heiter, schön”) as well as when the weather record indicated rain, snow, fog or completely overcast sky (“trüb”).


|

9

40

1

2

3

4

1835

1837

2

3

30

Frequency

Figure 11. The mean monthly frequency of cloudiness based on three daily observing times at Jihlava for the period 1817 to 1840: 1 – clear, 2 – little cloudy, 3 – much cloudy, 4 – overcast.

20

10

0 1817

1821

1823

1825

1827

1829

1831

1833

1839

40

1

4

30

Frequency

Figure 12. The mean annual variation of the frequency of cloudiness based on three daily observing times for Jihlava in the period 1817 to 1840: 1 – clear, 2 – little cloudy, 3 – much cloudy, 4 – overcast.

1819

20

10

0 J

An analysis was carried out of the frequencies of different cloud types at the morning, noon and evening observations. The data were categorised into the winter (October to March) and summer (April to September) halfyears for the period 1820 to 1826 (although observations were missing for October to November 1820 and for all of 1823 and August to September 1826). The results are summarised in Figure 13. Cumulus clouds constitute the most abundant record with dominance in the summer half-year (65% of cases at noon and 49 % in the morning). In the winter half-year the frequencies varied between 51 % at noon and 39 % in the morning. The second most frequently recorded cloud type was that of stratocumulus (between 7.5 % and 11.1 % in the winter and between 7.4 % and 16.8 % in the summer half-year). Overcast skies were also frequently recorded, especially in the winter half-year.

F

M

A

M

J

J

A

S

O

N

D

correspond with the directional data for modern observations that also indicate a south-easterly dominance for the Českomoravská vrchovina Highlands (Brázdil et al., 2004).

Figure 13. Relative frequencies of the occurrence of types of clouds and the state of weather based on three daily observing times in the summer and in the winter half-years at Jihlava according to Sterly for the period 1820 to 1826: Cr – Cirrus, Cm – Cumulus, Cc – Cirrocumulus, CrS – Stratocumulus, Nim – Nimbostratus, St – Stratus; 1 – clear, 2 – nice, 3 – variable, 4 – overcast, 5 – fog, 6 – rain, 7 – snow; a – October– March, b – April–September. 70 60

Morning

50

a

b

40 30 20

Wind

10 0


Cr

Relative frequency [%]

In common with some other observations, those for wind strength (based on the simple scale of calm, weak, medium and strong) reveal a hiatus in the series in 1827 with a notable decrease in the frequencies (Fig. 14). Some individual events do, however, stand out amongst which is that of 4 March 1817. At half past six in the morning, according to Sterly, there appeared a heavy storm with dark clouds and rain, during which more than 500 trees were uprooted in the forests around Jihlava. Other cases of gales were described in daily records on several occasions, but without any similarly detailed accounts of the damage. The analysis of the wind directions records provides clear evidence of two dominant sectors – the north-western and the south-eastern (Fig. 15). The former prevails in March, June, July and October, whereas in the other remaining months it is south-easterlies that dominate. These results

Cm

Cc

CrS Nim

St

1

2

3

4

5

6

7

70 60

Noon

50

a

b

40 30 20 10 0 Cr

Cm

Cc

CrS Nim

St

1

2

3

4

5

6

7

70 60 50

Evening

a

b

40 30 20 10 0 Cr Cm Cc CrS Nim St

1

2

3

4

5

6

7

Figure 14. Fluctuation of the mean monthly frequency of wind strength based on three daily observing times at Jihlava for the period 1817 to 1840: 1 – calm, 2 – weak wind, 3 – medium wind, 4 – strong wind.

90

1

2

3

4

Frequency

60

30

0 1817

1819

Summer

N

NW

W

E 10 20 30 40 50

SW

E

SW

SE

Autumn

N

NW

SE

W

E 10 20 30 40 50

N

NW

NE

SW

10 20 30 40 50

S

S

Spring

NE

W

SE

NE

W

E

SW

10 20 30 40 50

SE

S

S

1825

1827

1829

1831

1833

1835

1837

1839

Sterly provides useful descriptions of phenomena and recorded the state of weather, giving an account of phenomena (such as fog, rain, hail and snow) or the state of the sky (clear and nice, changeable, overcast) at three observing times during the day. He also noted optical phenomena, such as the Moon’s halo or light columns in the atmosphere. Sterly also included some phonological data such as the emergence of buds on trees, blossom dates or the beginning of the harvest. Under astronomical phenomena he mentioned the solar eclipse (he also described the course of the weather during the eclipse on 7 September 1820), falls of meteorites as well as a comet. From such notes and records it is possible to determine the numbers of cases with fog, thunderstorm and hail. It needs, however, to be acknowledged (as is often the case) that such phenomena tend to be under-recorded. Furthermore, and in a manner noted already for other aspects of the record there exists a break in the observations after 1827, resulting in a systematic drop in the frequency with which phenomena were recorded (Fig. 16).

N

NW

NE

1823

Meteorological phenomena

Figure 15. Seasonal wind roses (%) for Jihlava in the period 1817 to 1840. Winter

1821

Figure 16. Fluctuation (a) and the mean annual variation (b) of the frequency of occurrence of fogs, thunderstorms and hail at Jihlava over the period 1817 to 1840. 80

Frequency

a

Fog

60

30

40

20

20

10

a

Thunderstorm

Hail

8 6 4

0

2 0

0 1817

1822

1827

1832

1837

10

1817

1822

1827

1832

b

Thunderstorm

8

4

0.8

6

3

0.6

4

2

0.4

2

1

0.2

0

0 J

F M A M J

J A S O N D

1822

1827

1832

1837

1.0

b

Fog

1817

1837

5

b

Frequency

10

40

a

Hail

0.0 J

F M A M J

J A S O N D

J

F M A M J

J A S O N D


|

11

With regard to the observations of fog, the data gathered by Pokorny (1852) raise some concerns regarding the frequencies of different phenomena when compared with the original observations. Thus, in January 1819 fog should have been observed in 26 occasions, but according to the available daily observations it was recorded only 23 times over 12 days. In December 1821 out of 32 observed times with fog the available daily records document indicates only 29 fogs, with one additional non-standard time entry but covering again 12 days. The numbers of fogs in the individual years fluctuate greatly, with, for example, 63 in 1821 but only two cases in 1829. In the annual rhythm the number of fogs decreases gradually from January to the period from May to August and then increases steadily to the annual maximum in December. The average number of fogs in November was higher than in January. Attention must, however, be drawn to the disagreements between the thunderstorms collated by Pokorny (1852) and the original daily meteorological observations. Such differences suggest inaccuracies in the data processing procedures. Thus, out of 15 thunderstorms cited for August 1819 only 13 days are documented in the daily records. Moreover, the lightning records (e.g. on 15 August 1822), as well as that of thunder (e.g. on 11 December 1824) or distant thunderstorms (such as on 9 October 1826) were not included in the statistics for thunderstorms. Sterly recorded days with hail at Jihlava but only in the months March to October. Nevertheless, within this limit some conclusions can be offered. There were seven hail days in 1819, whereas in 13 years there were as few as between 1–4 such days while 10 years passed with no record of hail. With regard to the annual patterns, the maximum hail incidence is in May followed by August. The frequency of hail in March is remarkable and is greater than in April and comparable to June. Out of five cases occurring in March only two can be wholly verified, because the daily observations of the cited three hailstorms of March 1818 are missing. Otherwise on 14 March 1817 hail is mentioned in conjunction with a thunderstorm, whereas on 25 March 1822 the observer recorded only rain with hail. The observation includes also a record of thunderstorm and rain mixed with hail on 22 May 1824 at Jihlava. Sterly noted that in the area between Telč and Batelov the hail was of extraordinary size. This, however, is the only reference in the extant to any possible damage.

CONCLUSIONS Andreas Sterly belonged to that class of investigator whose contributions to meteorological observations and meteorology made this branch of science so important and durable in the Czech Lands. Even though the preserved daily observations do not cover the whole period 1816 to 1826, its extension based on monthly values up to 1844 gives them a matchless importance in extending our knowledge of climatic patterns in the first half of the 19th century, i.e. in the period before the building of the national network of meteorological stations. At Jihlava, Sterly’s activities provided a foundation for later meteorological observations that began


in the 1860s (January 1864 to February 1865) and 1870s (after 1873), the records of which are deposited in the Archives of the Brno branch of the Czech Hydrometeorological Institute, and in the State District Archives of Jihlava. Acknowledgements The publication was brought to light thanks to the support of the Grant Agency of the Czech Republic for the solution of the grant No. 205/05/0858. Our cordial thanks for providing archive materials go to Mgr. Renata Písková, head of the State District Archives of Jihlava. We thank Dr. Bořivoj Herzlík, Brno for the English translation and Dr. Dennis Wheeler, University of Sunderland, UK for the English style correction.

REFERENCES Alexandersson, H., 1986, A homogeneity test applied to precipitation data. Journal of Climatology, 6, 6, 661–675. Brázdil, R.–Dobrovolný, P.–Štekl, J.–Kotyza, O.–Valášek, H.– Jež, J., 2004, History of Weather and Climate in the Czech Lands VI: Strong Winds. Masaryk University, Brno, 378 p. Brázdil, R.–Řezníčková, L.–Valášek, H., 2006, Early instrumental meteorological observations in the Czech Lands I: Ferdinand Knittelmayer, Brno, 1799–1812. Meteorologický časopis, 9, 2, 59–71. Brázdil, R.–Valášek, H.–Macková, J., 2005, Meteorologická pozorování v Brně v první polovině 19. století. Historie počasí a hydrometeorologických extrémů. Archiv města Brna, Brno, 452 p. + appendices. D’Elvert, C., 1850, Geschichte und Beschreibung der (königlichen Kreis-) und Bergstadt Iglau in Mähren. Stadtgemeinde Iglau, Brünn, 520 p. D’Elvert, C., 1853, Andreas Sterly. Ein Nekrolog. In: Schriften der historisch-statistischen Sektion der k. k. mähr. schles. Gesellschaft des Ackerbaues, der Natur- und Landeskunde, V. Heft, Brünn, 262–266. Meteorologická pozorování. MZA Brno, fond G 82 Hospodářská společnost Brno 1769–1937, sign. IV/3e, karton 206. Míková, T.–Coufal, L., 1999, Tlak vzduchu na území České republiky v období 1961–1990. Národní klimatický program Česká republika 28, Praha, 66 p. Pokorny, A., 1852, Die Vegetationsverhältnisse von Iglau. Ein Beitrag zur Pflanzengeographie des böhmisch-mährischen Gebirges. In Commission bei W. Braumüller, Wien, 164 p. Schwab, B., 1935, Selbstbiographie von Andre Sterly. Igel-Land, 2, 1931–35, 74. Steiner, J., 1819, Mittheilungen der k. k. M. S. Gesellschaft, etc. etc. Vortrag des meteorologischen Vereins zur General-Versammlung der k. k. M. S. Gesellschaft zur Beförderung des Ackerbaues, der Natur- und Landeskunde, am 6ten Mai 1819. Hesperus, 52, 409–414. Sterly, A., 1846, Die wesentlichen Resultate der in Iglau durch einen Zeitraum von 24 Jahren gemachten meteorologischen Beobachtungen, nebst den Ursachen, welche auf die Beschaffenheit des Klima dieses Ortes vorzüglich einwirken. Moravia, 9, 111, 442–443. Wurzbach, C., 1879, Biographisches Lexikon des Kaiserthums Oesterreich. 38. Theil. Druck und Verlag der k. k. Hofund Staatsdruckerei, Wien, 237–239.

Příloha č. 11

EARLY INSTRUMENTAL METEOROLOGICAL OBSERVATIONS IN THE CZECH LANDS III: FRANTIŠEK JINDŘICH JAKUB KREYBICH, ŽITENICE, 1787–1829 RUDOLF BRÁZDIL1, LADISLAVA ŘEZNÍČKOVÁ1, HUBERT VALÁŠEK2, OLDŘICH KOTYZA3 1

Institute of Geography, Masaryk University, Kotlářská 2, 611 37 Brno, Czech Republic Moravian Land Archives, Žerotínovo nám. 3–5, 656 01 Brno, Czech Republic 3 Regional Museum, Mírové nám. 171, 412 01 Litoměřice, Czech Republic 2

František Jakub Jindřich Kreybich (1759–1833) worked as parish priest in Žitenice, where he carried out meteorological observations in 1787–1829. Kreybich’s handwritten daily records are preserved in the Archives of the Academy of Sciences, Prague, the Czech Republic; they are fragmentary before 1800, then form a continuous series from November 1800 to December 1818. Data from Žitenice at the level of monthly values exist for 1790– 1793, published by Strnadt (1795) then, starting with 1817, they appeared regularly in the publications of the I. R. Patriotic-Economic Society. For the period November 1800–December 1818, the analysis of air pressure and temperature, wind direction and the frequency of selected meteorological phenomena is carried out on the basis of daily values. Monthly series of air pressure and temperature at Žitenice have been homogenised for the period November 1800–December 1829 through reference to measurements recorded at the Prague-Klementinum and Vienna-Hohe Warte stations. František Jakub Jindřich Kreybich (1759–1833) působil jako farář v Žitenicích, kde prováděl svá meteorologická pozorování v letech 1787–1829. Kreybichovy rukopisné denní záznamy jsou dochovány v Archivu AV ČR v Praze ve fragmentech před rokem 1800 a v souvislé řadě z období listopad 1800–prosinec 1818. Na úrovni měsíčních hodnot z let 1790–1793 jsou uvedeny údaje z Žitenic v publikaci Strnada (1795) a od roku 1817 byly pravidelně publikovány ve spisech c. k. Vlastenecko-hospodářské společnosti. Pro období listopad 1800–prosinec 1818 je na základě denních hodnot provedena statistická analýza tlaku a teploty vzduchu, směru větru a četností výskytu vybraných meteorologických jevů. Měsíční řady tlaku a teploty vzduchu v Žitenicích byly pro období listopad 1800– prosinec 1829 homogenizovány podle měření stanic Praha-Klementinum a Vídeň-Hohe Warte. Key words: instrumental meteorological measurements – air temperature – air pressure – wind – meteorological phenomena – meteorological singularities – hydrometeorological extremes – František J. J. Kreybich – Žitenice

INTRODUCTION Although the first instrumental meteorological measurements for the region were taken at Zákupy (northern Bohemia) for 21 December 1719–31 March 1720 (Brázdil, Valášek, 2002), the development of meteorological observations in Bohemia was most closely associated with the existence of the Klementinum astronomical observatory in Prague. The measurement of precipitation, air pressure and temperature began there in 1752, under Josef Stepling, its first director (Observationes, 1753; Kreil, 1865). Whereas observations for the subsequent period 1753–1768 have not been found, those for the following years survived, largely thanks to the work of the third observatory director, Antonín Strnad, who published the monthly means of air pressure for the years 1752 and 1769–1793 and the monthly means of air temperature for the years 1770–1793 (Strnadt, 1794a, 1794b). As well as this, Strnad also published summary monthly overviews of visual daily observations in Prague for 1769–1771 (Strnadt, 1788, 1789, 1790) and a part of the measurements

for 1774 (Strnadt, 1775). On 1 January 1775, Strnad began regular daily observations (Meteorologická pozorování, 1976), through which, after 1 August 1781, Klementinum came to be included in the network of 39 meteorological stations that make up the Societas Meteorologica Palatina (Seydl, 1954; Pejml, 1975). Antonín Strnad was aware of the importance of meteorological observations from the very beginning and made an appeal that they be more widespread in Bohemia (Strnadt, 1776): “Many thanks would be due to anyone who, without great daily exertion, records changes in the air, the rise and fall of warmth, changes in the winds, dry or rainy weather; anyone who can procure the necessary instruments for the country and prepare suitable people for the task will later be remembered with glory. Eventually this may also have useful consequences over the course of the years, something that could lead us to a better knowledge of the state of the country. I herewith invite all Czech patriots to lend their hands, together with me, to realise this useful enterprise in the future.” Probably thanks to Strnad’s

Meteorologický časopis, 10, 2007, 63–74

|

63

activities, further meteorological observations appeared in Bohemia in the 1780s and 1790s for (Prague aside), Žitenice (Pejml, 1985), Choceň, Planá, Teplá and Boleboř (Strnadt, 1791, 1795). The meteorological observations for Žitenice made by František J. J. Kreybich are among the best-known and of the longest duration for their era. Stöhr (1920) characterised them as second-oldest only to the measurements at Klementinum. The analysis of Kreybich’s observations, which is the subject of the present paper, is a thematic continuation of the analysis of the earliest instrumental measurements in the Czech Lands, made by Ferdinand Knittelmayer in Brno (Brázdil et al., 2006) and Andreas Sterly at Jihlava (Brázdil et al., 2007).

FRANTIŠEK J. J. KREYBICH

Dr. Kottenauer. To convalesce, he spent considerable time at a spa in Teplice (Grunert, 1888; A. H., 1925). In the course of Kreybich’s residence at the Žitenice rectory, the establishment became an important cultural and scientific centre. Guests at the rectory included such notables as Alois David, director of the Klementinum observatory and Kassian Halaschka, a professor at Prague University. Kreybich also kept up a correspondence with Strnad and David (see e.g. Ankert, 1902), to whom he sent the results of his meteorological observations. Kreybich received a number of acknowledgements for his productive work. In 1806 he was appointed honorary canon of the chapter, then dean in 1817. In 1810–1811 he was invited as a consultant expert to a frontier committee negotiating claims made by Saxony to certain enclaves in the Rumburk region (Mucha, 1959). In 1820 he was appointed a corresponding member of the I. R. PatrioticEconomic Society and in 1827 an honorary member of the Society of the Patriotic Museum. Kreybich’s work in Žitenčice ended in 1829, after some 43 year. The reason for his retirement and his move

František Jindřich Jakub Kreybich (also spelt Kreibich) (Fig. 1) was born on 25 July 1759 at Kamenický Šenov, into the family of a linen merchant (for Kreybich’s further biographical data see Hackel, 1842; Grunert, 1888; A. H., 1925, 1933; Mucha, 1959; Pejml, 1970; Bechyně, 1972; Pejml, 1985). He spent his school days with relatives at Hoštka and went on to study Figure 1. František Jindřich Jakub in the choir of a Jesuit grammar school Kreybich (1759–1833) in Chomutov. It was here that his inte(O. Kotyza archives). rests in the exact sciences, natural history and technical drawing were kindled and nurtured by the rector of the establishment. He completed college education with success at the public examinations in mathematics and physics in Prague in 1779–1780; he was also interested in astronomy. Kreybich was then admitted to the Figure 2. A view of Žitenice Prague general seminary and subsequentbefore 1850 (Regional ly spent a year at the Litoměřice seminaMuseum, Litoměřice). ry, where he was ordained in 1786. His first post was curate to Žitenice (Fig. 2). In 1795 he was appointed parish priest and in the same year qualified as Doctor of Philosophy. He also taught catechism at the agricultural school in Žitenice (1791– 1801). In 1794 Kreybich completed, after three years of work, a detailed map of the Litoměřice diocese, required by the bishop as a template for dividing his administrative area into parishes. He then participated in the creation of a further series of maps and became known as a cartographer (Mucha, 1959, 1986; Pejml, 1985). However, his career was interrupted when, on 31 May 1806, he was struck by lightning in a thunderstorm during a trip to Poustevna near Skalice. His life was saved only by the prompt intervention of the regional physician, one 64 | Meteorologický časopis, 10, 2007

to Litoměřice may perhaps be found in social unpleasantness arising out of a sermon in which he publicly reproached the dissolute life of a certain official (Riegrův slovník naučný, 1864). In Litoměřice, he lived on Michalská Street in what is known as the “Budův dům” at No. 38. He died of double pneumonia there on 17 December 1833, at the age of 74.

Strnadt (1795) published the results of Kreybich’s observations of air pressure and temperature for the individual months of the period 1790–1793 (for 1791 see also Strnadt, 1793b). Figure 3. Summary of the meteorological observations of František J. J. Kreybich for 1787, title page (Wetterbeobachtung, catalogue no. 766).

METEOROLOGICAL OBSERVATIONS MADE BY FRANTIŠEK J. J. KREYBICH According to Katzerowsky (1887), Kreybich’s meteorological observations were available at the State Secondary Technical School at Litoměřice. However, all trace of them ends in the late 19th century. At the latest, the observations started in the year 1787 (Fig. 3), for the first ten months of which there exist summaries of Kreybich’s data in the Archives of the Academy of Sciences, Prague, Czech Republic (Wetterbeobachtung, catalogue no. 766). Meteorological observations for 1788, 1790 and 1794 are also deposited in the same place (Meteorologische Beobachtungen, catalogue no. 767) together with brief comparative barometric measurements for 1793, 1797 and 1798 (Observationes, catalogue nos. 712, 717, 721, 747). Kreybich even complemented his observations between November 1787 and June 1788 (Fig. 4) by graphical expression. Strnad used Kreybich’s observations on the meteorological character of the autumn and December of 1788 and January of 1789 in Žitenice in his relatively detailed compilation analysis of an extremely cold December in 1788 (Strnadt, 1791, 1793a). As well as this, Figure 4. Overview of 150 meteorological observations for June 1788 processed by Kreybich (Meteorologische Beobachtungen, catalogue no. 767).


|

65

Kreybich’s handwritten daily records of meteorological observations from November 1800 until December 1818 inclusive, sent to David of the I. R. Patriotic-Economic Society in the Bohemian Kingdom (Meteorologische Beobachtungen 1800–1818, catalogue nos. 768–785) remain preserved in the Archives of the Academy of Sciences, Prague, Czech Republic (Šlechtová, 1981). Kreybich’s monthly overviews consist of three parts corresponding to three daily readings, in addition to columns for the day and the phases of the moon (Fig. 5). Each of these parts includes the measurement of air temperature in Réaumur degrees (with a separate column for positive and negative temperatures), air pressure in Paris measures, temperature at the barometer and symbols expressing the character of the weather. Below the table for any given month appears a word characteristic of his evaluation of the values measured. Every year is then completed with a summary of the values measured for the individual months. Starting with 1802, the observations are further complemented by an agrometeorological overview of the year with mention of any impacts on the harvest. Starting in 1817, Kreybich’s Žitenice observations were regularly published in the writings of the above-mentioned society (Nachricht, 1825, 1826), continuing in a new series until the end of his observations (Neue Schriften, 1828–1832; Resultate, 1828). The writings give the monthly and the annual means of air pressure, calculated from the highest and the lowest values reduced to temperatures in 0°R. Monthly and annual means of air temperature are calculated as the means of the sums of all positive and all negative air temperatures. The means of the two characteristics were also completed with the absolute annual maxima and minima and their dates of occurrence. Žitenice also appears in the table of prevailing monthly wind directions. As well as contributing to this, Kreybich’s work also proved useful to the general economic overviews of Bohemia for the given year, always published after the meteorological overview proper. No detailed information about the installation of the thermometer and barometer at the Žitenice rectory is available from Kreybich’s records. In 1796, the I. R. Patriotic-Economic Society had thermometers and siphon barometers produced for its observers, but it is not clear whether or not Kreybich used instruments already in his possession. It is noteworthy that, in reports from 1801 (Meteorologische Beobachtungen, catalogue no. 768), he states that the society thermometer agrees with his, but that society barometer shows values 1.2 Parish lines lower. New instruments were later available from the Society in 1817 and 1827 but it is not clear whether Kreybich was involved. The times of day at which Kreybich made his observations in the period


1800–1818 are not all perfectly defined. There were stable readings at 1400 and 2100 hours, but the exact timing of the morning observations is problematical. In simplified form one might say that Kreybich measured at 0700 hours in November–February, at 0600 hours in March–April and September–October and at 0500 hours from May to August. In a number of cases, however, this scheme is impaired by observation at the time of the sunrise, which appears for the first time in 1805 (thus, in the whole of 1816, with the exception of September), or further deviations from the above time (thus, in May 1802 at 0600 hours, in December 1813 at 0800 hours). Thus the recent standard observation times, at 0700, 1400 and 2100 hours, were observed in the period 1800–1818 in only a quarter of all months (of which only 1803 completely). With respect to the daily air temperature variation, measurement at an earlier time than 0700 hours leads to underestimation, and a later term overestimation, of the temperature from the morning reading. From 1819 onwards, detailed information on Kreybich’s measurements has not survived. Despite all these problems with the morning observations, there is no doubt that the measurements at Žitenice were carried out by Kreybich three times a day, not leaving out any evening readings, as maintained by Stöhr (1930) or Gattermann (1935 – quotation in Pejml, 1985). With respect to the quality of judgement, Pejml (1985) posed another question, namely that of who stood in to measure for Kreybich when he was away on his frequent journeys, since there is not a single omission in the time observations for the above period. Figure 5. A specimen of the daily meteorological records made by František J. J. Kreybich: June 1801 (Meteorologische Beobachtungen, catalogue no. 768).

ANALYSIS OF THE METEOROLOGICAL OBSERVATIONS MADE BY FRANTIŠEK J. J. KREYBICH AT ŽITENICE Kreybich’s measurements at Žitenice (50º33´12´Ń, 14º09´44´´ E, 223 m.a.s.l.) have been used in a number of papers to date, particularly in the analysis of temperature patterns, since this place was once considered to be the warmest in Bohemia (e.g. Fritsch, 1851; Kreil, 1865; Katzerowsky, 1887, 1890, 1895; Stöhr, 1920; Gattermann, 1924, 1926). The period 1801–1829, for which with Kreybich’s observations survive was, according to measurements at the secular Prague-Klementinum and Vienna-Hohe Warte stations, colder from October to April and warmer from May to September in comparison with the years 1961–1990. In the case of air pressure, according to the Vienna measurements, an almost regular alternation of positive (highest in February) and negative (highest in October) events occurs in the difference between the two periods (Fig. 6).

2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0 -2.5

Vienna - Hohe Warte J

F

M

A

M

J

J

A

S

O

N

D

Temperature [°C]

1.5 1.0 0.5 0.0

Figure 7. a) Correlation coefficients of monthly means of air pressure between the Žitenice and Vienna-Hohe Warte stations in the period November 1800–December 1829 before (1) and after (2) homogenisation; b) box plot of mean monthly air pressure at Žitenice in the period November 1800–December 1829; c) mean annual variation of air pressure in Žitenice (3) in comparison with the Prague-Klementinum station (4) (data for Prague in Fritsch, 1850) in the period November 1800–December 1829 and with the Brandýs nad Labem-Stará Boleslav station (5) in the period 1961–1990 (data in Míková, Coufal, 1999). 1.0

-0.5

Prague - Klementinum Vienna - Hohe Warte

-1.0 -1.5 J

F

M

A

M

J

J

A

S

O

N

D

Correlation

Pressure [hPa]

Figure 6. Differences in mean air pressure and temperature for the years 1801–1829 and the normal period 1961–1990 for Prague-Klementinum and Vienna-Hohe Warte stations.

A further common monthly pressure series from Žitenice, consisting of both the series mentioned, has been created and the Standard Normal Homogeneity Test (SNHT – Alexandersson, 1986) used for homogeneity testing of it. In the first step, based on differences in monthly pressure means between Žitenice, Prague-Klementinum and ViennaHohe Warte, erroneous values were identified and then corrected by linear regression (Prague – October 1807; Vienna – January 1816, September 1812, December 1817). Erroneous values from March 1820–1822, September 1821– 1822, October 1825 and December 1821–1822 in Žitenice were filled in after application of SNHT which did not exhibit any inhomogeneity in the Žitenice monthly data (Vienna-Hohe Warte as reference station). But correlation coefficients between both series are rather weak, mainly in May (0.67) and August (0.62), while in the other months they fluctuated between 0.78 (July) and 0.92 (March) (Fig. 7a). The homogenised monthly series for air pressure in November 1800–December 1829 were further used to express annual variation by means of a box plot (Fig. 7b), which also yields information on the variability of individual monthly means. The annual variation demonstrates a main maximum in September (with secondary maxima in February and June) and a main minimum in April. The highest

2

0.8 0.6

1

0.4 0.2

a

0.0 J

Air pressure

M

A

M

J

J

A

S

O

N

D

F

M

A

M

J

J

A

S

O

N

D

Pressure [hPa]

995 990 985 980 975 970 965

b

960 J 1000

Pressure [hPa]

The values of air pressure were measured at Žitenice in Paris inches and lines (1 inch = 27.07 mm, 1 line = 2.256 mm), from which they were converted to mm Hg and subsequently to hPa. Two different pressure series for Kreybich’s Žitenice measurements are available: • series of monthly means calculated as the average of three daily readings for the period 1800–1818 (Meteorologische Beobachtungen 1800–1818, catalogue nos. 768–785) which is difficult to convert to 0 ºC because knowledge of Kreybich’s barometer is limited, • series of monthly means calculated from maximum and minimum daily values converted to 0 ºR for the period 1816–1829 (Nachricht, 1825, 1826; Neue Schriften, 1828–1832).

F

1000

995

5

990

4 3

985

c 980 J

F

M

A

M

J

J

A

S

O

N


D

|

67

mean monthly value occurred in November 1805 (998.9 hPa), the lowest in February 1823 (969.6 hPa). The variation range of monthly extremes is highest between November and February, and lowest in May. A comparison of Žitenice data with the annual variation of air pressure at the Brandýs nad Labem-Stará Boleslav station (50°11´17´´ N, 14º39´57´´ E, 180 m.a.s.l.) in the period 1961–1990 shows a movement of maxima from September to October while the minimum conforms in April (Fig. 7c). On the other hand, the Žitenice observations are closely comparable to those of the Prague-Klementinum station for 1801–1829 (Fritsch, 1850). Air temperature In similar fashion to that applied to air pressure, testing of the homogeneity of series of mean monthly air temperatures was also carried out, in this case with respect to the Prague-Klementinum station. With very high correlation coefficients between the two series, homogenisation was done for measurements taken in June, September and October from 1823, and were corrected in the November data from 1800–1802. As well as this, disputable means in February 1808, March 1801 and August 1823 were corrected with respect to Prague-Klementinum (Fig. 8a). The annual variation in mean air temperatures expressed by the box plot indicates a simple annual wave with a minimum in January and a maximum in July, from which the temperature for August differs very little (Fig. 8b). The highest monthly mean occurred in August 1807 (25.3ºC), the lowest in January 1823 (–9.3ºC). For comparison, the annual variation of temperatures at the Doksany station (50º27´16´´ N, 14º09´42´´ E, 158 m.a.s.l.) in the period 1961–1990 is also shown in Fig. 8b. With respect to this station, temperatures in Žitenice were higher between April and October.

Wind In 1806–1818 Kreybich included the frequency of wind directions in his overview of monthly observations, on the basis of which it has been possible to compile wind roses for the individual seasons and for the year (Fig. 9a). The prevalence of a western flow is recognizable for winter, autumn and the year, while in summer this direction was overtaken by the frequencies for the north-westerly wind. In spring the two wind directions are almost equal, with only the frequency of the occurrence of the south-easterly wind approaching them. For the period 1819–1827, each month at Žitenice is characterised only by the prevailing wind direction (in 1828–1829 only for the whole year). This information was collated for the whole period 1806– 1827, showing the frequencies of prevailing seasonal winds (Fig. 9b); south-westerly winds prevailed in winter and autumn, north-westerly in spring and summer. Figure 9. a) Relative frequencies (%) of wind direction for the seasons and the year at Žitenice in the period 1806–1818; b) seasonal relative frequencies (%) of prevailing wind direction at Žitenice in the period 1806–1827: 1 – winter, 2 – spring, 3 – summer, 4 – autumn. a

Winter

N NW

NE

W

SW

Correlation

W

SE

SW

Summer

N

0 5 10 15 20 25

SW

SE

Autumn

N

NE

W

E

0 5 10 15 20 25

S

NW

NW

E

NE

0 5 10 15 20 25

W

SE

SW

S

E

SE S

N

Year

NW

NE

2

a

0 5 10 15 20 25

W

1

0.9

SW

0.8 J

NE

S

Figure 8. a) Correlation coefficients of the monthly means of air temperature between the Žitenice and Prague-Klementinum stations in the period November 1800–December 1829 before (1) and after (2) homogenisation; b) mean annual temperature variation at Žitenice in the period November 1800–December 1829 expressed by box plot in comparison with the station Doksany (3) in the period 1961–1990 (data for Doksany see Květoň, 2001). 1.0

NW

E

0 5 10 15 20 25

Spring

N

F

M

A

M

J

J

A

S

O

N

D

E

SE S

30

b 100

20 15

Frequency [%]

Temperature [°C]

25

10 5 0 -5 -10

3 F

M

A

M

J

J

A

S

O

N

b

1

3

4

40 20

D

N

NE

E

SE

S

Wind direction


2

60

0

-15 J

80

SW

W

NW

Figure 10. Fluctuation (a) and the mean annual variation (b) of the frequency of occurrence of precipitation days (1 – heavy rain, 2 – light rain, 3 – snow or hail), thunderstorms and fogs at Žitenice over the period 1806 to 1818. 50

400

a Frequency

300

Precipitation days 1

2

40

3

a

Fog

150

30

200

100

20

100

50

10 0

0

0

1806 1808 1810 1812 1814 1816 1818

1806 1808 1810 1812 1814 1816 1818 8

30

b Frequency

200

Thunderstorm

a

Precipitation days 1

20

2

20

Thunderstorm

b

3

10

J

F M A M J

J

A S

O N D

b

6

15

4

10

2

5

0

0

1806 1808 1810 1812 1814 1816 1818

Fog

0

J

F M A M J

Kreybich also recorded the number of days with stürmisch, “stormy wind” between 1806 and 1818. The storminess frequency in Žitenice showed a smoothed annual variation with a maximum in February (secondary peak in August) and a minimum in July. A total of 63 days with stormy wind occurred in 1813, but in 1816 this total was only 17 days. Meteorological phenomena Kreybich’s monthly overviews for 1806–1818 also include the frequency of precipitation, fog, thunderstorms, stormy wind and variable weather, frosts and the state of the sky. In the case of precipitation days (Fig. 10), Kreybich distinguished between heavy and light rains and solid precipitation (snow, hail). In the period processed he recorded the most precipitation days in 1812 (199), the least in the year before (132). In the annual variation, frequencies for December to March dominate, with minima falling to October and September. Kreybich observed thunderstorms (Fig. 10) at Žitenice for every month of the year except November and December. The maximum occurred in July, followed by May and June. Their annual number fluctuated between 34 days (1807) and 17 days (1813 and 1818). The number of fogs (Fig. 10) was evidently underestimated in 1806–1808 (maximum in 1814). In terms of annual variation, the maximum occurred in October, followed by the frequencies of fogs in November–January; with minima recorded in April–May. Meteorological singularities Kreybich’s daily records from November 1800 to December 1818 have also made it possible to address the study of meteorological singularities as calendar-bound deviations in a given meteorological element from its mean long-term variation. Singularities were studied for air pressure and temperature employing the methodology established by Řezníčková et al. (2007), which makes it possible to evaluate statistically significant deviations in the annual variation of those elements.

J

A S O N D

J

F M A M J

J

A S O N D

In the case of air pressure in the period November 1800– December 1818, altogether 23 singularities were found with a mean duration of 3.0 days; the longest singularities extended to the 7 days of 1–7 January and 13–19 September (Table 1, Fig. 11). For air temperature, altogether 29 singularities were found with a mean duration of 2.8 days (Table 1, Fig. 12). In 9 cases a statistically significant deviation from the mean was only recorded for a period of one day. The longest temperature singularity, with duration of 8 days, fell to 2–9 May. Examination of Table 1 demonstrates close agreement between these events and the singularities revealed by analysis of Knittelmayer’s observations in Brno in 1799– 1812 (Brázdil et al., 2006), pointing to a certain temporal stability and spatial extent. In terms of the best-known meteorological singularities traditionally mentioned for the Czech Republic (Řezníčková et al., 2007), the situation in Žitenice in the period November 1800–December 1818 was as follows: • “deep winter” – expressed only on days 11–14 January, bound to a negative singularity of air pressure (12– 14 January), not to the prevalence of an anticyclonic weather regime • “May cooling” – usually associated with the so-called “Ice-men” did not appear at all; on the contrary, the longest warm anomaly occurred on 2–9 May (further 21–22 May and, surprisingly, the onset of “Médard Weather” on 8–12 June) as well • “high summer” – there was a perceptible singularity in the case of air temperature, with positive anomalies on 22–25 July, 27 July and 29 July–1 August • “Indian summer” – this was well expressed by two episodes of high air pressure on 13–19 September and 3–6 October, which were not reflected in the air temperature singularities (positive deviation only for 7–8 October) • “Christmas thaw” – this positive temperature singularity appears with a delay only on 29–30 December, the more significant preceding cooling falling to 19– 21 December. Meteorologický časopis, 10, 2007

|

69

Table 1. Singularities in air pressure and temperature in the period November 1800–December 1818 in Žitenice after Kreybich’s observations and in comparison with singularities in Brno for the period 1799–1812 (Brázdil et al., 2006). Key: MD – H – L – C – W –

mean deviation higher pressure lower pressure cold warm

Pressure No

Type

Period

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

H L L H L H H L H H L H L H L H L H L L H L L

1-7 Jan 12-14 Jan 20-22 Jan 19-21 Feb 8-10 Mar 25-27 Mar 5 Apr 15-17 Apr 2-3 June 11-13 June 2-4 July 9 July 17-20 July 17-18 Aug 11 Sep 13-19 Sep 22-24 Sep 3-6 Oct 8-12 Oct 29 Oct 28-29 Nov 6 Dec 8-10 Dec

Figure 11. Annual variation of mean daily air pressure in Žitenice for the period November 1800–December 1818: a) three-day running means smoothed by 60-day low-pass filter, b) three-day running deviations with intervals of reliability (80%) and marking of singularities with a duration of ≥ 2 days (for numbers, see Table 1).


MD [hPa] 3.2 -2.7 -2.5 4.0 -3.0 2.2 1.6 -3.8 0.9 1.2 -2.5 0.8 -1.2 1.0 -1.4 2.2 -2.0 3.0 -1.9 -2.0 2.0 -2.0 -2.8

Temperature Brno

No

Type

Period

2, 4-8 Jan 12-14 Jan 20-22 Jan 19-21 Feb 23-27 Mar 4-6 Apr 2-5 July 14-19 Sep 23-24 Sep 2-6 Oct 8-11 Oct 2-6 Dec 8-9 Dec

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

C W C C W C W C W W W C C W C W C W W W C W W C W C C C W

11-14 Jan 21 Jan 26-27 Jan 20 Feb 3-5 Mar 23-27 Mar 7-9 Apr 20-23 Apr 2-9 May 21-22 May 8-12 June 18-20 June 23-25 June 1 July 5-6 July 9-11 July 17 July 22-25 July 27 July 29 July-1 Aug 17 Aug 29 Aug-1 Sep 7-8 Oct 11-15 Oct 23 Oct 3 Nov 11 Dec 19-21 Dec 29-30 Dec

MD [°C] -1.3 1.1 -0.8 -0.9 0.9 -0.8 1.0 -1.4 1.2 0.9 0.8 -0.7 -0.9 0.6 -0.7 1.1 -0.8 0.6 0.6 1.0 -0.8 1.0 0.5 -0.8 0.6 -0.6 -0.8 -1.1 1.1

Brno 12-13 Jan 21-22 Jan 27 Jan 3-4 Mar 23-26 Mar 2-4, 9-10 May 22-23 May 6-8 June 18-20 June 23-25 June 9-10 July 17-19 July 24 July 28 July-3 Aug 16-18 Aug 28 Aug-1 Sep 6-9 Oct 11-15 Oct 2-3 Nov 17-22 Dec 24-29 Dec

Figure 12. Annual variation of mean daily air temperature in Žitenice for the period November 1800–December 1818: a) three-day running means smoothed by 60-day low-pass filter, b) three-day running deviations with intervals of reliability (80%) and marking of singularities with a duration of ≥ 2 days (for numbers, see Table 1).

EXTREME HYDROMETEOROLOGICAL PHENOMENA IN THE OBSERVATIONS OF FRANTIŠEK J. J. KREYBICH As a sensitive observer, František J. J. Kreybich also included in his records information on the occurrence of extreme hydrometeorological phenomena. Thus, in 1787 he drew attention to thunderstorms during which lightning started fires, namely on 13 April near Lovosice, 12 June in Terezín, 14 June and 7 July in Roudnice nad Labem, and on 16 June and 31 July in several places. From 1802 onwards, information about extremes appears in the economic overviews for given years with which he rounded off his own meteorological observations. These dealt particularly with the evaluation of impacts of the weather on agricultural production of grains, vegetables, fruit and grapevines. Kreybich recorded a certain amount of information on floods, often with reference to the movement of ice and the formation of ice barriers (for floods on the Elbe and the Ohře see Brázdil et al., 2005). Thus, on the night of 13/14 February 1805, the ice started moving on the Elbe at Litoměřice and stopped at Lovosice, when water with ice spread over fields. Substantially worse was, however, a sudden rise of water when the course of the river near Děčín was blocked on 26 February. In 1810, the ice on the Elbe started moving on 28 February and on the following day the water rose in both the Ohře and the Elbe, to such an extent that the roads and surroundings of Brozany, Doksany, Brňany, České Kopisty, Terezín and further communities were under water. During this long-lasting flood, the water culminated on 4 March and particularly after rain on 15 March. The water also left the course of the river in 1811 during movement of ice around Litoměřice. In 1812 the waters of the Ohře flooded the territory from Doksany up to its confluence with the Elbe, which also rose above its banks later (on 3 and 6 April). Kreybich devoted a particularly detailed description to the long and hard winter of 1813/14, when thick ice formed on the rivers and there was a flood associated with the movement of ice during which a bridge at Litoměřice was destroyed on 24 March (Fig. 13) and five people drowned. From 10 to 15 May 1815, the Elbe in flood damaged cereals, potatoes and further field crops. Another movement of thick ice was recorded at the beginning of January 1816. Kreybich

mentioned damage to hay as the Elbe flooded meadows near Mělník at the beginning of June 1817. Kreybich also recorded cases of drought. For example, in autumn 1810, great heat and more than two months of drought dried out brooks and streams, while the Elbe at Litoměřice fell to the lowest levels in living memory. Mills could not operate, leading to high prices for flour. Frequent rains in November rectified the situation. Kreybich also noticed frosts. Thus, for 1 June 1810 he mentioned great frost damage to young trees in higher positions, also speaking in a similar way of frosts around 15 November. On 29 June 1813, potatoes, beans and cucumbers froze in a strong frost and on 4–5 July snow fell in the mountains. Frosts on 16–22 April 1815 caused damage to walnut trees and grapevines. Repeated frosts towards the end of May of the same year again harmed grapevines and emerging grains. Grapevines suffered from the first frosts at the beginning of October 1817 and particularly in strong frosts on 16–18 October. For 15 May 1811, Kreybich recorded considerable hail damage to field crops around Žatec. Further thunderstorms with hail on the evening of 2 July did less damage, but a strong west by south-west wind across Třebenice, Lovosice, Litoměřice and Žitenice did great damage to roofs and windows, and uprooted fruit trees. The specification of the trajectory in Kreybich’s report indicates that it might have been a tornado. Kreybich’s information on the summer months of 1816 is particularly interesting. This season was characterised by unusually thick fogs and dangerous thunderstorms, during which lightning often struck and started fires. These thunderstorms were accompanied by heavy downpours and hailstorms, particularly on 5 August (Podbořany, Minice, Břvany in the Žatec region and the area between Libčeves and Libochovice in the Litoměřice region), and again on 14 August (the surroundings of Libčeves). Cold and rainy weather from May throughout the summer led Kreybich to rank this year as far from productive and, in a number of places, to place it among years with very bad crops. The year 1816, in the wake of the mighty volcanic eruption of Tambora (Lesser Sunda Islands, Indonesia; VEI = 7) in April 1815 (Sigurdsson, Carey, 1992), is marked as “the year without a summer” (see e.g. Stommel, Stommel, 1983; Harington, ed., 1992; Písek, Brázdil, 2006).

Figure 13. The Litoměřice bridge, destroyed during the Elbe flood of 24 March 1814 (Regional Museum, Litoměřice, catalogue no. SV H 3820).


|

71

CONCLUSIONS The meteorological and cartographic activities of František J. J. Kreybich lent him a status that significantly extended beyond the region in which he lived and worked. Kreybich’s observations stand out for their precision, comparability to the standards of modern climatological measurements, and their duration, even though they are not complete for the whole period of 1787–1829 in terms of daily and monthly values. Despite this shortcoming, they make it possible to obtain a considerable amount of valuable climatological information from a period in which analysis for the territory of the Czech Lands is otherwise based almost exclusively on the work of Prague-Klementinum. No less important are Kreybich’s reports of hydrometeorological extremes and their impacts, including the agrometeorological characteristics he assigned to individual years. He thus made a significant contribution to the completion of the mosaic of early instrumental meteorological measurements in the Czech Republic. Acknowledgements The paper was made possible by financial support from the Grant Agency of the Czech Republic, Grant No. 205/05/0858. Sincere thanks must be extended to the Archives of the Academy of Sciences of the Czech Republic in Prague for making available the meteorological observations of F. J. J. Kreybich, particularly to Mgr. Vlasta Mádlová. Thanks are due for the digitalisation of meteorological observations to Mgr. Pavel Zahradníček (Brno). Acknowledgement is made to PhDr. Bořivoj Herzlík (Brno) for English translation and Tony Long (Auchrae, Scotland) for English style corrections.

REFERENCES A. H. [Ankert, H.], 1925, Dr. Franz Kreibich. Unsere Heimat – Blätter für Heimatkunde des Leitmeritzer Gaues (Beilage zur Leitmeritzer Zeitung), 6, 1, 1–2. A. H. [Ankert, H.], 1933, Der 100. Todestag des Meteorologen und Kartographen Kanonikus Dr. Franz Jakob Kreibich. Leitmeritzer Zeitung, 63, 145, 4. Alexandersson, H., 1986, A homogeneity test applied to precipitation data. Journal of Climatology, 6, 6, 661–675. Ankert, H., 1902, Aus Leitmeritz (David, Director der Prager Sternwarte, an Canonicus Kreibich in Leitmeritz). Mittheilungen des Nordböhmischen Excursions-Clubs, 25, 192–193. Bechyně, J., 1972, Meteorolog a kartograf ze Žitenic. Kulturní měsíčník města Litoměřic 1972, únor, 2–4. Brázdil, R.–Dobrovolný, P.–Elleder, L.–Kakos, V.–Kotyza, O.– Květoň, V.–Macková, J.–Müller, M.–Štekl, J.–Tolasz, R.–Valášek, H., 2005, Historical and Recent Floods in the Czech Republic. Masaryk University and Czech Hydrometeorological Institute, Brno, Prague, 370 pp. Brázdil, R.–Řezníčková, L.–Valášek, H., 2006, Early instrumental meteorological observations in the Czech Lands I: Ferdinand Knittelmayer, Brno, 1799–1812. Meteorologický časopis, 9, 2, 59–71. Brázdil, R.–Řezníčková, L.–Valášek, H., 2007, Early instrumental meteorological observations in the Czech Lands II:


Andreas Sterly, Jihlava, 1816–1840 (1844). Meteorologický časopis, 10, 1, 3–12. Brázdil, R.–Valášek, H., 2002, Meteorologická měření a pozorování v Zákupech v letech 1718–1720. Geografie – Sborník České geografické společnosti, 107, 1, 1–22. Fritsch, K., 1850, Grundzüge einer Meteorologie für den Horizont von Prag, entworfen aus den an der k. k. Universitäts-Sternwarte daselbst in den Jahren 1771 bis 1846 angestellten Beobachtungen. Druck der k. k. Hofbuchdruckerei von Gottlieb Haase Söhne, Prag, 179 pp. Fritsch, K., 1851, Ueber die Temperatur-Verhältnisse und die Menge des Niederschlages in Böhmen. Sitzungsberichte der math.-naturw. Classe der kaiserl. Akademie der Wissenschaften, Wien, 22 pp. + 4 appendices. Gattermann, E., 1924, Die meteorologischen Beobachtungen des Dr. F. J. H. Kreibich in Schüttenitz (1788 bis 1829). Unsere Heimat – Blätter für Heimatkunde des Leitmeritzer Gaues (Beilage zur Leitmeritzer Zeitung), 5, 12, 45–46. Gattermann, E., 1926, Die meteorologischen Beobachtungen in Schüttenitz. Mitteilungen des Nordböhmischen Vereines für Heimatforschung und Wanderpflege, 49, 89–90. Grunert, J. R., 1888, Kanonicus Kreybich (Ein Gedenkblatt zu dessen 55. Sterbetag). Leitmeritzer Wochenblatt, 33, 98, 1–3. Hackel, J., 1842, Franz Jakob Heinrich Kreibich. Libussa, 1, 371–381. Harington, C. R., ed., 1992, The Year Without a Summer? World Climate in 1816. Canadian Museum of Nature, Ottawa, 576 pp. Katzerowsky, W., 1887, Übersicht der Maxima und Minima der Temperatur von Schüttenitz und Leitmeritz. In: Jahres-Bericht des k. k. Staats-Ober-Gymnasiums zu Leitmeritz in Böhmen für das Schuljahr 1887. Verlag des k. k. Staats-Ober-Gymnasiums, Leitmeritz, I–IX. Katzerowsky, W., 1890, Maxima und Minima des Luftdruckes von Schüttenitz und Leitmeritz. In: Jahres-Bericht des k. k. Staats-Ober-Gymnasiums zu Leitmeritz in Böhmen für das Schuljahr 1890. Verlag des k. k. Staats-Ober-Gymnasiums, Leitmeritz, 29–33. Katzerowsky, W., 1895, Meteorologische Nachrichten aus den Archiven der Stadt Leitmeritz. In: Jahres-Bericht des k. k. Staats-Ober-Gymnasiums zu Leitmeritz in Böhmen für das Schuljahr 1895. Verlag des k. k. Staats-Ober-Gymnasiums, Leitmeritz, 3–32. Kreil, C., 1865, Klimatologie von Böhmen. Druck und Verlag von Carl Gerold’s Sohn, Wien, 450 pp. Květoň, V., 2001, Normály teploty vzduchu na území České republiky v období 1961–1990 a vybrané teplotní charakteristiky období 1961–2000. Národní klimatický program Česká republika 30, Praha, 217 pp. Meteorologická pozorování v Praze-Klementinu 1775–1900, I. Hydrometeorologický ústav, Praha 1976, 258 pp. Míková, T.–Coufal, L., 1999, Tlak vzduchu na území České republiky v období 1961–1990. Národní klimatický program Česká republika 28, Praha, 66 pp. Mucha, L., 1959, František Jakub Jindřich Kreibich. Lidé a země, 8, 8, 356–358. Mucha, L., 1986, Mapy 16 česhkých krajů od Fr. J. J. Kreibicha. In: Z dějin geodézie a kartografie 6. Národní technické muzeum, Praha, 61–93. Nachricht von den Witterungsbeobachtungen, welche die kais. koen. Patriotisch-Oekonomische Gesellschaft in den Kreisen Böhmens veranstaltet hat. Verfasst von Professor Aloys David. 1ste Lieferung vom Jahre 1817–1819. Prag, 1825. 2te Lieferung vom Jahre 1820 u. 1821. Prag, 1826.

Neue Schriften der kais. königl. Patriotisch-ökonomischen Gesellschaft im Königreiche Böhmen. Band I/1–II/2. Gottlieb Haase Söhne, Prag, 1828–1832. Observationes baro-scopicae, thermo-scopicae, hyeto-metricae ad anum 1752 factae per Jos. Stepling, Soc. Jesu Sacerdotem, Caesareo-Regium Studii Philosophici Pragensis Directorem, et lectae in concessu Philosophico X. Calendarum Junii, Anno 1753 celebrato. Pejml, K., 1970, Z dějin přírodních věd. František Kreibich. Kulturní měsíčník [Roudnice n. L.], 6, 36–38, 43. Pejml, K., 1975, 200 let meteorologické observatoře v pražském Klementinu. Hydrometeorologický ústav, Praha, 78 pp. Pejml, K., 1985, PhDr. P. František Jindřich Jakub Kreybich (1759–1833), kartograf a meteorolog. In: Vlastivědný sborník Litoměřicko 17–20. Okresní vlastivědné muzeum, Litoměřice, 129–148. Písek, J.–Brázdil, R., 2006, Responses of large volcanic eruptions in the instrumental and documentary climatic data over Central Europe. International Journal of Climatology, 26, 4, 439–459. Resultate aus den Witterungsbeobachtungen von den Jahren 1822 bis einschliessig 1826, nebst Jahresberichten; als Fortsetzung der in der ersten und zweiten Lieferung dargestellten Witterungsbeobachtungen. Gottlieb Haase Söhne, Prag 1828. Riegrův slovník naučný, 1864, Frant. Jak. Jindř. Kreibich. Díl V, 18, 979–980. Řezníčková, L.–Brázdil, R.–Tolasz, R., 2007, Meteorological singularities in the Czech Republic in the period 1961– 2002. Theoretical and Applied Climatology, 88, 3–4, 179–192. Seydl, O., 1954, Mannheimská společnost meteorologická (1780–1799). Meteorologické zprávy, 7, 1, 4–11. Sigurdsson, H. S.–Carey, S., 1992, The eruption of Tambora in 1815: Environmental effects and eruption dynamics. In: Harington, C. R., ed.: The Year Without a Summer? World Climate in 1816. Canadian Museum of Nature, Ottawa, 16–45. Stommel, H.–Stommel, E., 1983, Volcano weather: the story of 1816, the year without a summer. Seven Seas Press, Newport RI, 177 pp. Stöhr, A., 1920, Schüttenitz, der wärmste Ort Böhmens. Unsere Heimat – Blätter für Heimatkunde des Leitmeritzer Gaues (Beilage zur Leitmeritzer Zeitung), 1, 12, 45–46. Stöhr, A., 1930, Luftschwere, Luftwärme und Wind von Schüttenitz. Unsere Heimat – Blätter für Heimatkunde des Leitmeritzer Gaues (Beilage zur Leitmeritzer Zeitung), 11, 9, 33. Strnadt, A., 1775, Witterungsbeobachtungen für das Jahr 1774. In: Abhandlungen einer Privatgesellschaft in Böhmen, zur Aufnahme der Mathematik, der vaterländischen Geschichte, und der Naturgeschichte. Erster Band. Im Verlage der Gerlischen Buchhandlung, Prag, 389–394. Strnadt, A., 1776, Meteorologische Beobachtungen auf das Jahr 1775. In: Abhandlungen einer Privatgesellschaft in Böhmen, zur Aufnahme der Mathematik, der vaterländischen Geschichte, und der Naturgeschichte. Zweyter Band. Im Verlage der Gerlischen Buchhandlung, Prag, 392–406. Strnadt, A., 1788, Physikalischer Witterungskalender. K. k. Normalschulbuchdruckerey, Prag, 152 pp. Strnadt, A., 1789, Physikalisches Taschenbuch auf das Jahr 1789. Für Freunde der Oekonomie und Witterungskunde. K. k. Normalschulbuchdruckerey, Prag, 183 pp. Strnadt, A., 1790, Chronologisches Verzeichniss der Naturbegebenheiten im Königreiche Böhmen vom Jahre Christi 633 bis 1700 mit einigen ökonomischen Aufsätzen samt

der periodischen Witterung auf das Jahr 1790. Gerlische Buchhandlung, Prag, 259 pp. Strnadt, A., 1791, Meteorologische Resultate der in Prag und einigen andern Orten in Böhmen gemachten Luftbeobachtungen und andern Erscheinungen. In: Neuere Abhandlungen der k. Böhmischen Gesellschaft der Wissenschaften. Erster Band. J. V. Degen, Wien und Prag, 235–256. Strnadt, A., 1793a, Beyträge zu der Geschichte des Winters im Christmonate 1788. In: Mayer, J., ed.: Sammlung Physikalischer Aufsätze, besonders die Böhmische Naturgeschichte betreffend, von einer Gesellschaft Böhmischer Naturforscher. Dritter Band. In der Waltherischen Hofbuchhandlung, Dresden, 39–88. Strnadt, A., 1793b, Meteorologischer Beytrag auf das Jahr 1792. In: Riem, J., ed.: Neue Sammlung vermischter oekonomischen Schriften. Dritter Band. Dresden. Strnadt, A., 1794a, Ueber die mittlere Barometerhöhe von Prag. In: Mayer, J., ed.: Sammlung Physikalischer Aufsätze, besonders die Böhmische Naturgeschichte betreffend, von einer Gesellschaft Böhmischer Naturforscher. Vierter Band. In der Waltherischen Hofbuchhandlung, Dresden, 41–60. Strnadt, A., 1794b, Bestimmung des mittlern Grads der Wärme von Prag. In: Mayer, J., ed.: Sammlung Physikalischer Aufsätze, besonders die Böhmische Naturgeschichte betreffend, von einer Gesellschaft Böhmischer Naturforscher. Vierter Band. In der Waltherischen Hofbuchhandlung, Dresden, 61–68. Strnadt, A., 1795, Resultate der in Prag und einigen andern Orten in Böhmen 1790, 1791, 1792, 1793 gemachten meteorologischen Beobachtungen. In: Neuere Abhandlungen der k. Böhmischen Gesellschaft der Wissenschaften. Zweyter Band. J. G. Calve, Prag, 249–253. Šlechtová, A., 1981, Státní hvězdárna. Inventář archivního fondu. Ústřední archiv Československé akademie věd, Praha, 202 pp.

Archival sources All sources are located in the Archives of the Academy of Sciences, Czech Republic, Prague in the Státní hvězdárna – State Astronomical Observatory section. Meteorologische Beobachtungen in Schüttenitz. Catalogue no. 767 (previous catalogue no. VI C 8, No. 2). Meteorologische Beobachtungen vom Jahre 1801 nebst den Monaten Novemb und December vom Jahre 1800 nach Reaumurschen Thermometer und Barometer der oekonomischen Gesellschaft zu Prag. Schüttenitz im Leitmeritzer Kreise von Franz Ja. Kreybich Pfarrer. Catalogue no. 768 (VI C 8, No. 3). Meteorologische Beobachtungen vom Jahre 1802 Fr. Jac. von Kreybich Pfarrer zu Schüttenitz. Catalogue no. 769 (VI C 8, No. 4). Meteorologische Beobachtungen nebst Tabellarischer Uibersicht und Oekonomischen Bemerkungen vom Jahre 1803 beobachtet im Leitmeritzer Kreise von Franc. Jac. Kreybich, Pfarrer zu Schüttenitz. Catalogue no. 770 (VI C 8, No. 5). Meteorologische Beobachtungen vom Jahre 1804 von Fr. Ja. H. Kreybich Pfarrer zu Schüttenitz nächst Leitmeritz. Catalogue no. 771 (VI C 8, No. 6). Meteorologische Beobachtungen nebst einigen oekonomischen Bemerkungen vom Jahre 1805 beobachtet in Schüttenitz unweit Leitmeritz von Fr. Jac. Kreybich Pfarrer. Catalogue no. 772 (VI C 8, No. 7). Meteorologický časopis, 10, 2007

|

73

Meteorologische Beobachtungen nebst 2 Summarischen Tabellen und einer oeconomischen Uibersicht vom Jahre 1806 von Franz Jac. Kreybich Canon. u. Pfarrer zu Schüttenitz 1/2 Stunde NOstlich von Leitmeritz. Catalogue no. 773 (VI C 8, No. 8). Meteorologische Beobachtungen vom Jahre 1807 beobachtet in Schüttenitz bei Leitmeritz von Fr. Jac. Hein. Kreybich Canonicus zu Leitmeritz und Pfarrer zu Schüttenitz. Catalogue no. 774 (VI C 8, No. 9). Meteorologische Beobachtungen vom Jahre 1808 in Schüttenitz nahe bei Leitmeritz gemacht von Fr. Jac. H. Kreybich Canonic zu Leitmeritz u. Pfarrer zu Schüttenitz. Catalogue no. 775 (VI C 8, No. 10). Meteorologische Beobachtungen nebst berechneten Tabellen vom Jahre 1809 in Schüttenitz 1/2 Stunde NO von Leitmeritz beobachtet von Franz Jac. Hein. Kreybich Phil. Doct. Canonic. zu Leitmeritz und Pfarrer zu Schüttenitz für die Prager oekonomische Gesellschaft. Catalogue no. 776 (VI C 8, No. 11). Meteorologische Beobachtungen nebst berechneten Tabellen und oekonomischen Bemerkungen vom Jahre 1810 für die K. K. ökonomisch patriotische Gesellschaft in Prag von Franz Jak. Heinr. Kreybich … Catalogue no. 777 (VI C 8, No. 12). Meteorologische Beobachtungen vom Jahre 1811 für die K. K. ökonomisch-patriotische Gesellschaft in Schüttenitz beobachtet von Fr. Jac. Hr. Kreybich … Catalogue no. 778 (VI C 8, No. 13). Meteorologische Beobachtungen nebst oekonomischen Bemerkungen vom Jahre 1812 in Schüttenitz 1/2 Stunde NOstlich von Leitmeritz beobachtet von Franz Jac. Hr. Kreybich … Catalogue no. 779 (VI C 8, No. 14). Meteorologische Beobachtungen vom Jahre 1813 im Leitmeritzer Kreise zu Schüttenitz beobachtet vom Fr. Jac. Hr. Kreybich … Catalogue no. 780 (VI C 8, No. 15). Meteorologische Beobachtungen vom Jahre 1814 in Schüttenitz beobachtet unweit der K. Kreisstadt Leitmeritz von


Fr. Jac. Heinr. Kreybich … Catalogue no. 781 (VI C 8, No. 16). Meteorologische Beobachtungen nebst ökonomischen Bemerkungen vom Jahre 1815 in Schüttenitz unweit der K. Kreisstadt Leitmeritz beobachtet von Fr. Jac. H. Kreybich … Catalogue no. 782 (VI C 8, No. 17). Meteorologische Beobachtungen mit oekonomischen Bemerkungen vom Jahre 1816 beobachtet in Schüttenitz 1/2 Stunde Nordöstlich von der K. Kreisstadt Leitmeritz enfernt … von Franz Jacob Heinrich Kreybich … Catalogue no. 783 (VI C 8, No. 18). Meteorologische Beobachtungen vom Jahre 1817 in Schüttenitz 1/2 Stunde Noestlich von Leitmeritz beobachtet von Franz Jac. H. Kreybich … Catalogue no. 784 (VI C 8, No. 19). Meteorologische Beobachtungen vom Jahre 1818 für die K. K. patriotische oekonomische Gesellschaft in Böhmen beobachtet in Schüttenitz 1/2 Stunde Nordöstlich von Leitmeritz von Franz Jac. Heinr. Kreybich … Catalogue no. 785 (VI C 8, No. 20). Observationes barometricae et thermometricae factae 1793 mense Majo Schuttenitzii prope Litomericium. Catalogue no. 712 (VI C 4, No. 7). Observationes barometricae et thermometricae habitae Schuttenicii 1797 mense Octobri. Catalogue no. 717 (VI C 4, No. 12). Observationes barometricae et thermometricae Schuttenicii 1798 sub finem Maji et initio Junii. Catalogue no. 721 (VI C 4, No. 16). 1793 Observationes barometro Societatis Scientiarum Bohemae portatili et thermometro meo minori, institutae Schuttenitzii prope Litomerizium mense Majo una cum Resultatis calculo inde deductis. Catalogue no. 747 (VI C 6, No. 2). Wetterbeobachtung der 1ten zehn Monate des 1787ten Jahrs aus der Gegend bey Leitmeritz. Catalogue no. 766 (VI C 8, No. 1).

Seznam příloh. Příloha č. 9:

Recommend Documents