Memo
Aan
LS Datum
Kenmerk
Aantal pagina's
1 januari 2014
1207726-000-ZKS-0032
8
Van
Doorkiesnummer
E-mail
Katherine Cronin
+31 (0)88 33 58 027
[email protected]
Onderwerp
DELWAQ-SPM_Noordzee-Zuno-DD_j03-11_v01 metadata Waterkwaliteitsmodelschematisatie
METADATA
Page break
DELWAQ-SPM_Noordzee-Zuno-DD_MER Zandwin
May 2013
1.
Datum
Ons kenmerk
Pagina
1 januari 2014
1207726-000-ZKS-0032
2/8
Basisbeschrijving gebiedsmodel 1.1. Watersysteem 1.2. Naamgeving gebiedsmodel (zie Naamgeving conventies modellen Rijkswaterstaat)
1.3. Modelconfiguratie
1.4. Datum laatste versie 1.5. Datum eerste versie 1.6. Binnen welk (beleids)kader is het model ontwikkeld? 1.7. Ter beantwoording van welke vraag of vragen is het model ontwikkeld?
1.8. Belangrijkste berekende grootheden (inclusief eenheid en eventueel statistiek) 1.9. Voor welke toepassing mag het model worden ingezet? Voor welke toepassing niet (zonder aanpassing)?
2.
Modelonderdelen 2.1. Modelrooster (plaatje, .png .jpg etc.)
Zuidelijke Noordzee DELWAQ-SPM_Noordzee-Zuno-DD_MER Zandwinning Kies uit: Hydrodynamica Slib Tracers/Zout/Temperatuur Nutriënten, Primaire productie, Zuurstof Draagkracht, Schelpdieren Habitats/Ecotopen september 2013 2008 MER Zandwinning (+ details) Het model is (door)ontwikkeld in het kader van het project MER Zandwinning. De vraag was in welke mate zandwinning in de kustzone leidt tot een verhoging van de slibconcentratie en vertroebeling in de kustzone. Slibconcentratie in de waterkolom (mg/l). Slibfluxen langs de Hollandse kust en naar de Waddenzee (MT/jaar). Slibpercentage in de bodem (% slib) Het model is specifiek ontwikkeld om het grootschalige verspeidingsgedrag van een groot aantal slibpluimen in de kustzone te berekenen en de bijdrage hiervan ten opzichte van de natuurlijke slibdynamiek te kwantificeren. Het model is niet bedoeld om het lokale effect van individuele pluimen te berekenen en ook niet om een vertaling te maken van grootschalige naar kleinschalige effecten.
Datum
Ons kenmerk
Pagina
1 januari 2014
1207726-000-ZKS-0032
3/8
2.2. Geografische locatie (KML) 2.3. Beschrijving rooster en resolutie
2.4. Beschrijving van situatie 2.5. Gesimuleerde periode 2.6. Kenmerk gesimuleerde periode – Waarom is deze periode gekozen? (bijv. gemiddeld jaar) 2.7. Gebruikte software en versienummer
3.
Top panel shows the hydrodynamic grid, bottom panel shows the aggregated DELWAQ grid Zie repository There were three domains in the Delft3D-FLOW model that were aggregated to one DELWAQ in the horizontal. In the vertical there are 12 nonequidistant sigma layers (the same as in Delft3DFLOW). Yearly simulations 2003-2011 (2003-2008 calibration; 2009-2011 validation/forecast) Period before and after the construction of the Maasvlakte-2 Delft3D-WAQ version 4.5212 17-06-2011
Model in data, beschrijving en bronnen 3.1. Koppeling met andere gebiedsmodellen (bijv. hydrodynamica, slib, …) 3.2. Geografische gebiedsdata 3.3. Bathymetrie
Delft3D-FLOW_Noordzee-ZUNO-DD_j03-j11_v01
As in hydrodynamic model. (For convenience repeated here: Base bathymetry was taken from the Simona-ZUNO Model V4 as samples extracted and re-interpolated on the fine grid of ZUNO-DD_j03-j11_01. Other data was taken from the latest soundings of the Nederlands Continentaal Plat (25 by 25 m). In some locations these soundings date from 2009 but can be as old as 2002 in other areas. For all three domains of Zuno-DD the option ‘data in cell centre’ was chosen when the samples were interpolated to the grid.
Datum
Ons kenmerk
Pagina
1 januari 2014
1207726-000-ZKS-0032
4/8
3.4. Schematisatie gelijk aan hydrodynamica? 3.5. Indien nee, wat is reden voor wijziging? Wat is de gewijzigde schematisatie (horizontaal en verticaal)? Hoe is aanpassing gecontroleerd en wat is informatieverlies door aggregatie?
3.6. Randvoorwaarden (boundary conditions)
3.7. Lozingen van stoffen
3.8. Overige forceringen 3.9. Onderbouwing van keuze van belangrijke modelprocessen
In addition to the base bathymetry, three modifications were made to improve the representation of channels in the grid: 1. Deepening of the channel near Brouwersdam 2. Deepening of the southern channel near Haringvliet and 3. Deepening of the northern channel near the Haringvliet) Ja / Nee: Two of the three grids (fine and intermediate) were aggregated 2x2 Grid aggregation in the horizontal only to speed computational time. The number of vertical layers were kept the same. In previous phases of the project (Blaas et al., 2012) checks were performed to assess the influence of reducing the horizontal resolution on parameters such as salinity and temperature. The differences were deemed acceptably small to justify the improvement in computational time Concentrations of SPM were prescribed at the channel and northern model boundaries based on a derived climatology Suspended sediment concentrations from the CEFAS KnowSeas database (stored on Deltares servers) were used. The data includes daily riverine loads, optimised from (where possible) daily flow data and weekly, bi-weekly or monthly flow and nutrient data. The data covers riverine loads for countries bordering the Baltic Sea, Norway, Denmark, Germany, the Netherlands, France (north of 47 deg. N or the Loire), Great Britain, Northern Ireland and Ireland. The data was processed by CEFAS (S.M. van Leeuwen,
[email protected]) and the University of Hamburg (J. Paetsch, H.-J. Lenhart: German and Dutch data). In addition, representation of erosion from bordering cliffs and banks, such as the Flemish banks was taken from the literature Temperature, salinity and wind forcings from Delft3D-FLOW model. The simulation of SPM was the primary objective, therefore a good prediction of salinity (representing the transport capabilities of the model) was necessary from the Delft3D-FLOW model. In addition, the prescription of bed shear stress, important for the resuspension of sediments from the bed was calculated offline in a method described in Cronin (2013a). A new process description has been added to the model: in order to reproduce the development of an easily erodible fluffy sediment layer near the bed the concept of deposition efficiency has been
Datum
Ons kenmerk
Pagina
1 januari 2014
1207726-000-ZKS-0032
5/8
3.10. Onderbouwing van keuze van belangrijke modelparameters
3.11. Gebruikte procesmanager/ workflowmanager en databases
4.
introduced. The deposition efficiency constant is an extra calibration parameter which is a multiplication factor of the settling flux into the bed. A value less than 1 results in a higher average concentration at the bottom. This will include the effect of estuarine circulation on net sediment transport. This process alters the vertical gradients of SPM in the water column and bed composition in relation to the surface values. The key model parameters related to SPM are chosen as part of the buffer model approach taken here. A good parameter setting for the entire period (2003-2011) was required and moreover because the modified grid and subsequent re-aggregation and the updated bed shear stresses, the parameterisation of the DELWAQ model required recalibration. The most influential parameters on resuspension of sediment from the bed and deposition to the bed include: • Critical Shields stress for sand mobilization in the buffer layer (TauShields) • Van Rijn Pickup factor from the buffer layer (FactResPup) • Critical bed shear stress (per fraction) in the fluff layer (TcrS1,IMi) • First order resuspension rate from the fluff layer (VResIMi) • Setting velocity (per fraction) Not applicable
Kalibratie/validatie 4.1. Kalibratie: Hoe is gekalibreerd? Welke parameters zijn gekalibreerd? Welke gegevens zijn gebruikt voor kalibratie?
4.2. Gevoeligheidsanalyse
A good parameter setting for the entire period (2003-2008) was required and moreover because the modified grid and subsequent re-aggregation and the updated bed shear stresses, the parameterisation of the DELWAQ model required recalibration. As the bed shear stress due to waves and currents is significantly lower than in previous versions of the model (Blaas et al., 2012, an update of the coefficients of resuspension and deposition was required (see point 3.10). Recalibration was performed by adjusting the parameters for erosion and resuspension for the two layers of the buffer model, the fluff layer and buffer layer. These and other parameters (Cronin, 2013a)were further adjusted when comparing with field measurements to ensure that both calm conditions and stormy conditions were well represented by the model. no
Datum
Ons kenmerk
Pagina
1 januari 2014
1207726-000-ZKS-0032
6/8
4.3. Validatie
4.4. Betrouwbaarheid (objectieve maat)
4.5. Betrouwbaarheid (expert oordeel)
5.
A set of surface in-situ validation data from MWTL (Rijkswaterstaat), MUMM and CEFAS were used. In addition, vertical profiles of model results were also compared to the vertical SPM and turbidity profile measurements by the Port of Rotterdam (PoR). Finally, the deterministic model was plotted against MERIS data that was interpolated onto the model grid (Gaytan & Blaas 2012). Using both in-situ and MERIS data gives a better impression of the success of the model in capturing seasonal and inter-annual variability, especially in areas where only a few in-situ measurements per year are taken. The bias and root mean square errors of measurements versus model were assessed. The bias convention was model minus measurement. Nearshore concentrations had a lower bias than offshore concentrations. The model captured both the daily and seasonal fluctuations of SPM as well as the capturing the increased concentrations during storm peaks. In addition, the amount of sediment present in the bed should remained equilibrium after each year simulation
Uitgevoerde aanpassingen (indien van toepassing) 5.1. De reden van de aanpassing
5.2. Het doel van de aanpassing 5.3. Een omschrijving van de aanpassing 5.4. Het effect van de aanpassing. Indien er meerdere aanpassingen tegelijkertijd zijn doorgevoerd moet het effect van de verschillende aanpassingen afzonderlijk en cumulatief in beeld worden gebracht. Dus aparte beschrijvingen van de effecten voor programmatuur, resolutie, schematisatie, schematisatiemethode, randvoorwaarden, parameters. Daarnaast een beschrijving van het totale effect van alle veranderingen.
The main differences between this version of the DELWAQ-SPM_Noordzee-Zuno-DD (v2) and the previous version (v1) are: 1. The prescription of bed shear stresses due to waves. In V1 (MER-Zandwinning) a fetch length approach was used. Here in v2 the bed shear stresses generated offline using SWAN yearly averaged wave fields 2. Grid differences 3. Different calibration objectives – in v1 the Wadden Sea was of greater importance and the near bed concentrations were also examined during calibration
Datum
Ons kenmerk
Pagina
1 januari 2014
1207726-000-ZKS-0032
7/8
5.5. Leidt de aanpassing tot een zodanige wijziging van de kalibratieresultaten dat er opnieuw gekalibreerd moet worden? 5.6. Consequentie van de aanpassing voor het huidige model en voor toekomstige modellen 5.7. Het effect van het aangepaste model op de relaties met de andere basismodellen.
6.
Projecten waarin gebiedsmodel ontwikkeld/gebruikt is 6.1. Projecttitel Projectnummer Opdrachtgever Contactpersoon Deltares 6.2. Projecttitel Projectnummer Opdrachtgever Contactpersoon Deltares 6.3. Projecttitel Projectnummer Opdrachtgever Contactpersoon Deltares 6.4. Uitbreiden indien nodig
7.
Beheer van de modellen 7.1. 7.2. 7.3. 7.4.
Eigenaar van gebiedsmodel Functioneel beheerder Rijkswaterstaat Applicatiebeheerder Deltares Status gebiedsmodel
7.5. Locatie bestanden
8.
MOS2-II en MOS2-III (the last version of the model) 1204561 Havenbedrijf Rotterdam (HBR) Meinte Blaas
Deltares n/a Katherine Cronin and Thijs van Kessel Kies uit: Ongeschikt / Verouderd In ontwikkeling / Niet gevalideerd Gereed voor gebruik
https://repos.deltares.nl/repos/mcmodels/trunk /ZUNO-DD/delwaq/
Referenties 8.1. Kalibratie- en validatierapport
8.2. Overige rapporten
Cronin, K. and Blaas, M. (2013a) MoS2-II Deterministic Model calibration; Updates of the hydrodynamic and SPM model, Deltares Report 1204561-000--0025, 97 pages Cronin et al., (2013b) Setup and evaluation fo baseline hydrodynamic and SPM models, Deltares Report 1208242-002, 104 pages Blaas, M., Cronin, K., Serafy, G.Y.E., Friocourt, Y.F., Triana, I.D.T.F.G., Aguilar, S.G. and Keetels, G.H. (2012) MoS2: Model setup, data assimilation and skill
Datum
Ons kenmerk
Pagina
1 januari 2014
1207726-000-ZKS-0032
8/8
assessment; Model supported Monitoring of SPM in the Dutch coastal zone, p. 235, Deltares El Serafy, G., Gaytan Aguilar, S., Cronin, K., vanBeek, J., Dijkstra, J., Stuparu, D. & Blaas, M., (2013). SPM Model optimization and forecast analysis. Deltares Report, 1208242-002, 138 pages. 8.3. Overige producten 8.4. Achtergrondinformatie
9.
Wensen voor B&O en Ontwikkeling 9.1. Wensen voor Beheer en Onderhoud
1.
2.
3.
9.2. Wensen voor Ontwikkeling 9.3. Mogelijkheden voor verbeteringen 9.4. Mogelijkheden voor uitbreiding
To assess and subsequently improve SPM simulation in the Wadden Sea and offshore further as this calibration focused on the area of interest to the Port of Rotterdam, the Dutch coastal zone To make a detailed comparison between the different model versions to assess their strengths and weaknesses To run an updated wave simulation for the model domain and a coupled flow-wave simulation in order to examine the differences in bed shear stresses generated through different methods
Further calibration should be done to improve the offshore concentrations of SPM
