GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
The use of GIS to determine changes in Caraulun River Delta, Timor-Leste. Juno Rouwenhorst, Dr. Guy Boggs, Prof. Bob Wasson , Dr. Waqar Ahmad Charles Darwin University, Darwin, Australia.
[email protected] Raimundo Mau The Agriculture and Land use Geographic Information Systems Unit Narciso Almeida de Carvalhoc Ministry of Agriculture, Forestry, and Fisheries, Timor-Leste
Abstract
Integrated remote sensing and GIS techniques are commonly used to determine and map changes in a landscape. In this study, such techniques are used to establish changes in the Caraulun River Delta, Timor-Leste. Deltas are useful indicators of change in catchments because they are the integrated outcome of the processes and changes that affect sediment transport and hydrology in a catchment. Anthropogenic activities are often a cause for change and can have great effects on the geomorphology of river channels, cause changes in coastline positions, and can impact coastal ecosystems. The land cover and land use of the Caraulun catchment in Timor-Leste has been heavily modified over centuries, with potentially significant consequences for water quality, river geomorphology, and biodiversity. In order to determine whether these modifications have had an effect on the river delta, Landsat 5 TM imagery from 1986, 1996, and 2006 is being used to examine the evolution of the Caraulun river delta. Special focus is given to determining changes in the coastline positions, land cover, and river morphology. Although final results are not available for these proceedings, remote sensing and GIS methods and techniques for the project are discussed. The proposed methodology for establishing coastline change will consist of density slicing techniques; object-based classifications will be used to determine changes in the river morphology; and independent unsupervised classifications of each image will be followed by post-classification change detection for the determination of land cover change. These methods will be followed by GIS analyses to verify and quantify change. 50
Penggunaan GIS Untuk Mendeteksi Perubahan di Caraulun River Delta, Timor-Leste Juno Rouwenhorst, Dr. Guy Boggs, Prof. Bob Wasson, Dr. Waqar Ahmad, Raimundo Maub & Narciso Almeida de Carvalhoc
Abstrak
Teknik-teknik GIS dan penginderaan jarak jauh terpadu biasanya digunakan untuk menemukan dan memetakan perubahan-perubahan yang terjadi pada landscape. Dalam penelitian ini, teknik-teknik seperti itu digunakan untuk membuat perubahan – perubahan di Caraulun River Delta, Timor Leste. Delta adalah indikator yang baik dalam mendeteksi perubahan yang terjadi dalam daerah tangkapan air karena delta adalah hasil dari proses-proses yang terpadu dan perubahan-perubahan yang mempengaruhi perpindahan sedimen dan hidrologi dalam daerah tangkapan air. Kegiatankegiatan antropogenik sering menjadi penyebab terjadinya perubahan dan dapat memberikan efek-efek yang besar terhadap geomorfologi aliran-aliran sungai, menyebabkan perubahan terhadap posisi garis pantai dan mempengaruhi ekosistem pesisir Permukaan tanah dan penggunaan tanah daerah tangkapan air Caraulun telah mengalami banyak perubahan selama berabad-abad, dengan efeknya terhadap kualitas air, geomorfologi sungai, dan keragaman hayati. Untuk mengetahui apakah perubahan-perubahan ini mempunyai efek terhadap delta sungai, gambar Landsat 5 TM dari tahun 1986, 1996, dan 2006 digunakan untuk meneliti evolusi delta sungai Caraulun. Fokus yang besar adalah mengetahui perubahan-perubahan posisi garis pantai, permukaan tanah, dan morfologi sungai. Meskipun tidak ada hasil akhir dari kegiatan-kegiatan ini, penginderaan jauh dan teknik-teknik GIS untuk proyek ini dibicarakan. Metodologi yang disusulkan untuk mengetahui perubahan garis pantai akan terdiri dari teknik-teknik pemilahan kepadatan; klasifikasi berdasarkan obyek akan digunakan untuk mendeteksi perubahan-perubahan morfologi sungai yang terjadi; klasifikasi Remote sensing applications for mapping and monitoring 51
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
Introduction It is generally accepted that upland land uses affect the lowlands of a catchment (e.g. Wasson et al., 2008). Following this hypothesis, deltas can be considered to be the integrated outcome of all processes and changes that affect sediment transport in a catchment. Consequently, delta change can be a useful indicator of the net result of change in a catchment. The study of changes in a delta can provide us with information that not only allows us to improve coastal management, but also catchment management. Changes in the geomorphology of deltas generally give us important information about natural coastal dynamics, including the effects of wave action, tides, coastal currents, river flow, sea-level changes, storms, and wind action (Woodroffe, 2003; Restrepo & Lopez, 2007; Syvitski & Saito, 2007), but uncharacteristic changes in rates of delta evolution could also indicate upland land use changes. Hence the monitoring of delta change is an important issue not only for the delta itself, but for the whole catchment as well. Timor-Leste has had a very turbulent history (e.g. Silove, 2000; Nevins, 2002, 2003), many of the effects of which are reflected in the landscape. Deforestation due to logging is considered to be a serious environmental problem in most parts of Timor-Leste (e.g. Sandlund et al., 2001; Bouma & Kobryn, 2004), and the effects of the displacement of people are also noticeable through abandoned houses, adjacent rice paddies and agricultural plots. Anthropogenic activity can have a significant effect on the natural coastal dynamics, such as by altering sediment flow and water discharge. For example, increases and changes in land use have been shown to increase sedimentation (e.g. Boix-Fayos et al., 2007; Syvitski & Saito, 2007) and affect river and stream morphology (e.g. Vanacker et al., 2003); the construction of dams can decrease sediment and water 52
discharge to the coast (e.g. Chu et al., 2006); and the development of irrigation systems can decrease both water flow and sediment loads to the coast. In 1996, the construction of an irrigation system was finalized in the Caraulun catchment in the south of Timor-Leste. The irrigation system was built to supply the rice paddies in the centre of the delta with water. However, this irrigation system has since been destroyed by a flood, and re-building of the structure was completed in 2007. Apart from this fact, the Caraulun catchment has also been identified by the government of Timor-Leste as being heavily modified by human activity. Integrated remote sensing and geographic information systems (GIS) techniques have been successfully used to map, analyse, and monitor delta change (e.g. Kaufmann & Seto, 2001; Yue et al., 2003), coastline change (e.g. Frihy et al., 1998; White & Asmar, 1999; Chen et al., 2005; Foody et al., 2005), and changes in delta river morphology (e.g. Yang et al., 1999; Fan et al., 2006). Hence, the aim of this project is to use integrated remote sensing and GIS techniques to determine whether there has been any change in the Caraulun River delta over the past 20 years, and to ascertain what these changes are, focusing on coastline change, changes in river morphology, and land cover change. Spatial and temporal environmental factors are included in the analyses to attempt to establish some of the potential drivers for change. Description of study area The Caraulun River delta occupies an area of approximately 55 km2, and is located on the south coast of Timor-Leste, south of the capital Dili (Fig.1). There are several hamlets scattered throughout the delta, with the main village, Betano, located on the eastern coastline of the delta. The Caraulun catchment coastline is characterized by wave-exposed, fairly steep, darkgrey sandy beaches. The slopes of the beach vary according to position on the beach and
independent dan tanpa pengawasan setiap gambar akan diikuti oleh klasifikasi selanjutnya yaitu untuk mendeteksi perubahan permukaan tanah. Metode-metode ini akan diikuti oleh analisa-analisa GIS untuk memastikan dan menghitung perubahan yang terjadi. Pendahuluan Kita semua sepakat bahwa pemanfaatan tanah di dataran tinggi mempengaruhi daerah tangkapan air di dataran rendah (e.g. Wasson et al., 2008). Mengikuti hipotesa ini, delta dianggap sebagai hasil dari semua proses dan perubahan yang mempengaruhi transportasi sediment di daerah tangkapan air. Efeknya adalah perubahan delta bisa merupakan indikator yang baik untuk mengetahui hasil perubahan yang sebenarnya terjadi dalam daerah tangkapan air. Penelitian terhadap perubahan-perubahan di sebuah delta dapat memberikan kita informasi yang bukan hanya membantu kita untuk meningkatkan manajemen pantai, tapi juga manajemen daerah tangkapan air. Perubahan-perubahan pada geomorfologi delta memberikan kita informasi penting tentang dinamika alam pantai, perubahan permukaan air laut, badai, dan gerak angin (Woodroffe, 2003; Restrepo & Lopez, 2007; Syvitski & Saito, 2007), tapi perubahan-perubahan aneh yang terjadi di delta juga bisa mengindikasikan perubahanperubahan pemanfaatan yang terjadi di tanah dataran tinggi. Oleh karena itu pemantauan terhadap perubahan pada delta adalah hal yang penting bukan hanya untuk delta itu sendiri tapi juga untuk seluruh daerah tangkapan air. Timor Leste mempunyai sejarah kekacauan (e.g. Silove, 2000; Nevins, 2002, 2003), banyak dari efek-efek tersebut terefleksi di landscape. Penggundulan hutan karena aktifitas logging adalah masalah serius di hampir seluruh daerah di Timor Leste (e.g. Sandlum et al., 2001; Bouna & Kobryn, 2004), dan efek relokasi penduduk juga terlihat dari rumahrumah yang ditinggal, lahan sawah di daerah sekitar dan bekas perkebunan.
Kegiatan antropogenik bisa memberikan efek yang besar terhadap dinamika alam pantai seperti mengubah aliran sedimen dan aliran air. Contohnya peningkatan dan perubahan pemanfaatan tanah telah terbukti meningkatkan sedimentasi (e.g. Boix-Fayos et al., 2007; Syvitski & Saito, 2007) dan mempengaruhi sungai dan morfologi arus (e.g. Vanacker et al., 2003); pembangunan bendungan bisa mengurangi sedimentasi dan aliran air ke pantai (e.g. Chu et al., 2006); dan pembangunan system-sistem irigasi bisa mengurangi sedimentasi dan aliran air ke pantai. Pada tahun 1996 pembangunan sistem irigasi diselesaikan di daerah tangkapan air Caraulun di Selatan Timor Leste. Sistem irigasi dibangun untuk mensuplai air ke lahan persawahan di tengah delta. Namun sistem irigasi ini telah hancur oleh banjir, dan pembangunan kembali sistem ini selesai pada tahun 2007. Terlepas dari masalah ini, pemerintah Timor Leste telah mengidentifikasi bahwa telah terjadi perubahan besar di daerah tangkapan air Caraulun karena aktifitas manusia. Penggunaan penginderaan jarak jauh terpadu dan teknik-teknik GIS telah berhasil memetakan, menganalisa, dan memantau perubahan pada delta (e.g. Kaufmann & Seto, 2001; Yue et al., 2003). Perubahan garis pantai (e.g. Frihy et al., 1998; White & Asmar, 1999; Chen et al., 2005; Foody et al., 2005), dan perubahanperubahan pada morfologi delta sungai (e.g. Yang et al., 1999; Fan et al., 2006). Oleh karena itu tujuan proyek ini adalah untuk menggunakan penginderaan jarak jauh terpadu dan teknik-teknik GIS untuk mengetahui apakah ada perubahan pada delta Caraulun River selama 20 tahun terakhir, dan memastikan perubahan apa yang sebenarnya terjadi, fokus
Remote sensing applications for mapping and monitoring 53
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
Table 1: Landsat 5 TM imagery used in this study Date of observation
Sensor
Size
Tide level at Suai at time of image observation
1986 – Sept.12
Landsat 5 TM
Quarter scene
2.71 m
1996 – Sept. 7
Landsat 5 TM
Quarter scene
1.40 m
2006 – Sept. 3
Landsat 5 TM
Full scene
1.15 m
along the coast. Intertidal reef platforms can be found in several places along the coast. According to local fishermen, the benthos becomes dominated by a muddy substrate approximately 50 m offshore (see Wyatt, 2004). The offshore gradient is low, and the edge of the continental shelf lies approximately 410 km from the delta coast. The vegetation fringing the shoreline varies from patches of mangrove to lagoon swamp to palm forest and shrubbery. Land use consists mainly of agriculture, with rice being the main crop produced. Other crops include bananas and maize. In 1996 an irrigation system was developed in the delta, diverting water from the river to irrigate a large area of rice paddies in the centre of the delta. Slash-and-burn agriculture is commonly applied in upland areas of the catchment, with farmers rotating their fields between every 2–7 years (pers. comm. with local farmers). As a result, much of the vegetation has been affected by human activity. Vegetation types that are found in the delta include Casuarina forests along the river banks, whilst eucalypt and palm forests can be found inland. Along the coast there are small patches of mangrove forest. The weed Chromolaena odorata is wide-spread. The Caraulun River bed ranges from approximately 200m to 700m in width, and consists of sand, pebbles and boulders. During the dry season, a few small channels run through the river bed, the courses of which alter much from year to year. Large flows occur during the wet season, driven by tropical thunderstorm and monsoon rainfall activity. The climate of Timor-Leste is of the tropical monsoonal 54
type, with the south coast having a bimodal rainfall pattern. It has its first heavy rainfall from December-February and the second one between May-June, with the mean annual rainfall ranging between 1300-2400 mm (see Asian Development Bank, 2004). Stream powers during the wet season are higher and as a consequence cause more changes in the river bed than dry season river flow. Erosion of the river banks can be observed in various places.
Fig.1: The location of Caraulun catchment in Timor-Leste Methods Data & Software The data used for this study includes a DEM downloaded from the CGIAR website (http://srtm.csi.cgiar.org), which was produced using 90m Shuttle Radar Topography Mission (SRTM) data, and three Landsat 5 TM images dating from 12th September 1986, 7th September 1996, and 3rd September 2006 (see Table 1). The images were selected based on having no cloud cover in the study area, and for having been observed at similar dates during the same month (Lillesand & Kiefer, 2000). Software being used for the project includes ArcGIS®, Erdas IMAGINE®, and Definiens®.
terhadap perubahan garis pantai, perubahanperubahan pada morfologi sungai, dan perubahan pada permukaan tanah. Faktor-faktor spasial dan bersifat sementara termasuk di dalam analisa-analisa yang dilakukan untuk menemukan potensi penyebab terjadinya perubahan. Penjelasan tempat penelitian Luas delta Caraulun River kira-kira 55 km2 dan terletak di pesisir Selatan Timor Leste, daerah Selatan ibukota negara Dili (Gambar. 1). Ada beberapa desa tersebar di sekitar delta, dengan desa utama, Betano, yang terletak di daerah Selatan delta. Garis pantai daerah tangkapan air Caraulun ditandai pantai-pantai yang berpasir gelap, cukup curam dan tidak terlindung dari ombak. Kemiringan pantai berbeda-beda menurut posisi pantai dan posisinya di sepanjang pesisir. Menurut petani setempat dasar laut didominasi oleh Lumpur kira-kira sejauh 50 m dari pantai (lihat Wyatt, 2004). Kemiringan pantai rendah, dan sisi landas kontinen membentang kira-kira 4-10 km dari pesisir delta. Tumbuhan yang mengelilingi daerah lepas pantai berbeda-beda dari kumpulan-kumpulan bakau, rawa sampai hutan palem dan semak-semak. Pemanfaatan tanah umumnya adalah untuk pertanian dan beras adalah produk utamanya. Produk-produk pertanian lain adalah pisang dan jagung. Pada tahun1996 sistem irigasi dibangun di delta yang menggunakan air sungai untuk mengairi areal persawahan yang luas di tengah delta. Sistem pertanian dengan membuka lahan umum dilakukan di daerah tangkapan air dataran tinggi dengan rotasi yang dilakukan oleh petani antara 2-7 tahun (pers.comm. dengan petani setempat). Sebagai hasilnya adalah banyak daerah vegetasi yang terkena dampak dari aktifitas manusia. Jenis-jenis vegetasi yang ditemukan di delta adalah hutan-hutan Casuarina di sepanjang tepi sungai, sementara hutan Eucalypt dan
palem ditemukan di daerah yang agak jauh dari tepi sungai. Di sepanjang pesisir ada sejumlah kecil kelompok hutan bakau. Rumput liar Chromolaena odorata banyak tersebar di daerah tersebut. Lebar dasar Caraulun River membentang kira-kira 200 sampai 700 m yang terdiri dari pasir, baru kerikil dan batu-batu besar. Selama musim kemarau hanya sedikit aliran sungai yang mengalir di sungai tersebut yang arahnya berubah-rubah dari tahun ke tahun. Aliran sungai yang besar terjadi pada musim hujan yang disebabkan oleh hujan lebat dan angin monsoon. Iklim di Timor Leste adalah tipe monsson tropis dengan pesisir selatan mempunyai dua pola hujan. Periode hujan lebatnya yang pertama dari Desember sampai February sedangkan yang kedua dari Mei sampai Juni dengan rata-rata curah hujannya adalah 13002400 mm per tahun (lihat Bank Pembangunan Asia, 2004). Kekuatan arus selama musim hujan lebih besar dan menyebabkan lebih banyak perubahan yang terjadi di dasar sungai daripada di musim kemarau. Erosi yang terjadi di tepi sungai dapat terlihat di banyak tempat. Metode-metode Data dan Software Data yang digunakan dalam penelitian ini adalah DEAM yang di download dari website CGIAR (http://srtm.csi.cgiar.org), yang dibuat dengan menggunakan 90m data Shuttle Radar Topography Mission (SRTM), dan 3 gambar Landsat 5 TM dari tanggal 12 September 1986, 7 September 1996 sampai 3 September 2006 (lihat tabel 1). Gambar-gambar tersebut dipilih karena tidak tertutup awan dan diambil pada bulan yang sama (Lillesand & Kiefer, 2000). Software yang digunakan adalah ArcGIS®, Erdas IMAGINE®, dan Definiens®. Pra-proses Sebelum analisa dan klasifikasi gambar, semua dataset mentah melalui teknikal proses untuk memastikan pixel over time ke pixel
Remote sensing applications for mapping and monitoring 55
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
Pre-processing Prior to image classification and analysis, all raw data sets were subjected to pre-processing techniques to ensure over time pixel to pixel comparison and to highlight changed areas. This commenced with the orthorectification of the raw images using Geocover Landsat 7 ETM+ data for xy control and the SRTM 90m DEM for z control. The RMS error was less than 50 m (for more information on the Geocover Landsat orthorectification process see Tucker et al., 2004). This technique was used to find field sites to allocate land cover labels to the spectral clusters generated out of the image classification process. This was followed by atmospheric path correction using standard dark area subtraction technique (Ahmad & Hill, 2003). The resultant images were cleaned using image masking techniques and enhanced to ensure maximum possible spectral variation in the raw images.
Fig2: Determination of the Caraulun River Delta
56
The DEM was used to determine the catchment and delta study areas. First, basins were created using the hydrology tools in ArcMap, then those basins covering the intended study area were manually selected to create a single polygon to outline the whole of the Caraulun catchment. In order to create the primary delta study area, those elevations up to and including 50 m above sea level and located within the catchment were selected. However, this created a study area that included areas that were not of interest, and excluded areas that were of interest (for example, the coastline was extended seaward so that no areas of coastline are lost when clipping out the delta for image processing). Therefore, the delta area was manually edited to include and exclude these areas (Fig.2). Discussion of methods to be used Analysis for the determination of change was focused on three sections: coastline changes (Fig.4), river morphology changes (Fig.5), and land cover changes (Fig.6). Fig.3 shows a general summary of the methods. Coastline A number of studies have examined the extraction of waterlines and changes in coastlines (e.g. White & El Asmar, 1999; Ryu et al., 2002; Chen et al., 2005; Chu et al., 2006; Zhao et al., 2008). Commonly used methods in these studies have included digitizing of the coastline using visual investigation (Zhao et al., 2008); density slicing or edge detection methods using a single band (Yamano et al., 2006); and classifications using multiple bands (White & El Asmar, 1999; Ryu et al., 2002). Preliminary investigation of the imagery indicates that density slicing will be the most suitable, as turbidity within the water column is low and the near-infrared (NIR) band is highly absorbed. Coastline mapping is vulnerable to tidal variations, however the imagery has low to mid tides and this error is considered to be equivalent to the scale of mapping.
Geocover Landsat 7 ETM+ data
Landsat 5 TM imagery: 1986 1996 2006
Pre-processing: orthorectification DOS Enhancing Subsetting
1986, 1996, 2006
Coastline: density slicing using NIR band overlay analysis (GIS)
1986, 1996, 2006
River morphology: object-based classification to map river bed and channels GIS analyses
DEM (SRTM 90 m)
Determination of catchment and study areas using ArcMap tools
1986, 1996, 2006
Land cover: Independent pixel-based unsupervised classifications Post-classification change detection using GIS analysis
Fig.3: A generalized overview of the project methods. A more detailed perbandingan dan memastikan area-area menggambarkan seluruh daerah tangkapan yang berubah. Aktifitas ini dimulai dengan air Caraulun. Untuk membuat area penelitian orthorectification gambar-gambar mentah utama delta, dipilih daerah dengan ketinggian dengan menggunakan Geocover Landsat 50 m atau lebih di atas permukaan laut dan 7 ETM+ data untuk xy control dan SRTM berada di dalam daerah tangkapan air. Namun 90m DEM untuk z kontrol. Kesalahan RMS area penelitian yang dibuat ini meliputi areakurang dari 50 m (untuk informasi lebih lanjut area yang tidak menarik untuk diteliti dan mengenai proses Geocover Landsat orthorecmengesampingkan area-area yang menarik tification lihat Tucker et al., 2004). Teknik ini untuk diteliti (contohnya garis pantai ditamembantu kita untuk menemukan tempatrik menjorok ke laut sehingga tidak ada area tempat terbuka dan untuk mengalokasikan garis pantai yang luput ketika memilah-milah label-label permukaan tanah ke kluster-kluster delta untuk pengolahan gambar). Oleh karena spektral yang dihasilkan dari proses klasifikasi itu, area delta secara manual diedit untuk gambar. Kemudian dilanjutkan dengan korememasukkan dan mengeluarkan area-area ini ksi jejak atmospheric dengan menggunakan (Gambar. 2). teknik standar substraksi area gelap (Ahmad Pembahasan Metode & Hill, 2003). Hasil-hasil gambar dari proses Analisa pendeteksian perubahan yang terjadi tersebut dibersihkan dengan menggunakan berfokus pada 3 bagian yaitu: perubahan garis teknik masking dan enhanced untuk mendapantai (Gambar 4), perubahan morfologi sunpatkan hasil maksimal variasi spektral gambargai (Gambar 5), dan perubahan permukaan gambar mentah tersebut. tanah (Gambar 6). Gambar 3 memberikan DEM digunakan untuk menemukan areainformasi tentang metode-metode secara area daerah tangkapan air dan delta untuk ringkas. penelitian. Pertama, daerah-daerah cekungan Garis Pantai dibuat dengan peralatan hidrologi di ArcMap, Sejumlah penelitian telah meneliti tandadaerah-daerah yang meliputi area penelitian tanda jalur air dan perubahan-perubahan garis tersebut kemudian secara manual dipilih unpantai (e.g. White & El Asmar, 1999; Ryu et tuk menciptakan sebuah single polygon untuk al., 2002; Chen et al., 2005; Chu et al., 2006; Remote sensing applications for mapping and monitoring 57
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
After establishing the coastlines, an overlay analysis will be carried out to highlight any changes. Then, the linear aspect of coastline change will be examined, which will include determining distances between linear features, at, for example, distances of 30 m, or the equivalent of one pixel. Changes in total coastline length and complexity will also be examined, as well as where the changes are located in relation to the river mouth. This will be followed by a comparison of the coastline changes with any changes in temporal and spatial environmental factors, such as rainfall,
agricultural development, and, if possible, changes in sea-level. As can be seen in Fig. 4, changes in the coastline can be readily observed in the raw imagery. From 1986 to 1996 there is a clear retreat of the edge of the river mouth, which expands again from 1996 to 2006. A number of other retreats and expansions of the coast have also occurred, in particular on either side and to the west of the river mouth, and near lagoon areas. Since the direction of the along shore current is from east to west, these changes may be related to changes in sediment supply coming from the Caraulun river. River morphology In order to manage a river system and the many cultural practices that exist within a catchment, an understanding of the evolution of large river systems is essential, and this includes knowledge of river channel migration on the delta (Yang et al., 1999). Various methods exist for extracting river channels from remotely sensed imagery. The digitization of the channels through visual investigation is commonly used (e.g. Fan et al., 2006; Boix-Fayos et al., 2007), often when the image is displayed in a certain band. Syvitski & Saito (2007) found that for most deltas the red or NIR wavelengths are better suited for establishing channel dimensions. Another frequently used technique is the pixel-based classification using either single bands, or a combination of bands. Yang et al. (1999) found that the combination of Landsat bands 1, 6, and 7 was the most effective for mapping river channels at low water discharge, whereas bands 1, 5, and 7 were the best for mapping river channels at high water discharge.
Fig.4: Arrows indicate some of the coastline changes of the Caraulun river delta from 1986-2006. 58
However, given the linear characteristics of river channels, object-based classification is likely to be an effective technique for automatic mapping of river channel characteristics. Object-based classification involves the initial segmentation of an image into objects,
Zhao et al., 2008). Metode-metode yang biasa dilakukan dalam penelitian ini adalah digitasi garis pantai dengan menggunakan investigasi visual (Zhao et al., 2008); metode-metode pemisahan kepadatan atau pendeteksian tepi menggunakan single band (Yamano et al., 2006); dan klasifikasi menggunakan multiple band (White 7 El Asmar, 1999; Ryu et al., 2002). Penelitian awal terhadap gambar menunjukkan bahwa pemisahan kepadatan akan menjadi metode yang paling tepat karena kekeruhan dalam kolom air rendah dan near-infrared (NIR) band banyak diserap. Pemetaan garis pantai rentan terhadap variasi air pasang, namun gambar menunjukkan air pasang rendah dan sedang dan kesalahan ini dianggap setara dengan skala pemetaan. Setelah membuat garis pantai, analisa terhadap permukaan akan dilakukan untuk melihat perubahan-perubahan yang terjadi. Kemudian aspek linear perubahan garis pantai akan diteliti termasuk menentukan jarak antara fitur-fitur linear contohnya pada jarak 30 m atau setara dengan satu pixel. Perubahan-perubahan pada panjang garis pantai dan kompleksitas juga akan diteliti, begitu juga lokasi perubahanperubahan yang terjadi yang berhubungan dengan mulut sungai. Hal ini diikuti dengan sebuah perbandingan antar perubahan-perubahan garis pantai yang bersifat sementara dan disebabkan oleh faktor-faktor lingkungan spasial seperti hujan, perkembangan pertanian dan jika mungkin perubahan-perubahan pada level permukaan laut. Seperti yang bisa dilihat pada Gambar 4 perubahan-perubahan pada garis pantai bisa diteliti melalui gambar yang masih mentah. Dari tahun 1986 sampai 1996 terlihat jelas terjadi penyusutan mulut sungai yang meluas lagi dari tahun 1996 sampai 2006. Sejumlah penyusutan dan perluasan lain di pantai juga terjadi, yaitu di kedua sisi dan bagian barat mulut sungai, dan dekat area lagoon. Karena arah arus sungai dari timur ke barat, perubahan-perubahan ini mungkin berhubungan dengan suplai sedimen yang datang dari sungai Caraulun.
Morfologi Sungai Untuk mengatur sistem sungai dan banyaknya adat istiadat yang ada di daerah tangkapan air, pengertian akan evolusi sistem-sistem sungai sangatlah penting yang meliputi pengetahuan tentang migrasi aliran sungai di delta (Yang et al., 1999). Berbagai macam metode dilakukan untuk mengekstraksi aliran-aliran sungai hasil dari penginderaan jarak jauh. Proses digitalisasi aliran-aliran sungai dengan investigasi visual sudah umum dilakukan (e.g. Fan et al., 2006; Boix-Fayos et al., 2007), ketika gambar ditampilkan di band tertentu. Menurut Syvitski & Saito (2007) sebagian besar delta berwarna merah atau panjang ombak lebih cocok untuk membuat dimensi-dimensi aliran sungai. Teknik lain yang sering dilakukan adalah klasifikasi berdasarkan pixel dengan menggunakan single band atau band kombinasi. Menurut Yang et al. (1999) kombinasi Landsat band 1, 6 dan 7 adalah yang paling efektif untuk pemetaan aliran-aliran sungai dengan arus air yang kecil, sedangkan band 1, 5, dan 7 adalah yang paling efektif untuk pemetaan aliranaliran sungai dengan arus air yang besar. Namun dengan karakteristik linear aliranaliran air sungai, klasifikasi berdasarkan obyek sepertinya lebih efektif untuk pemetaan otomatis karakteristik aliran-aliran air sungai. Klasifikasi berdasarkan obyek meliputi segmentasi awal sebuah gambar ke dalam obyek dengan variannya yang lebih kecil dari obyekobyek disekitarnya. Teknik ini bisa dilakukan untuk gambar heterogen atau pemfokusan fitur-fitur yang tersembunyi seperti gedung, jalan, dan kanopi pohon. Sama dengan analisa terhadap garis pantai dan permukaan tanah, GIS akan menganalisa perubahan-perubahan dalam aspek linear saluran-saluran air sungai dengan mengukur fitur-fitur seperti lebar aliran setiap 30 m (setara untuk setiap pixel) dan lengkungan aliran setiap 500 m, dan menganalisa bentuk ‘permanen’ baru. (i.e. bukan musim kemarau)
Remote sensing applications for mapping and monitoring 59
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
Fig.5: Arrows indicate some of the changes in the Caraulun river banks and flow from 1986-2006. within which variance is less than compared with neighbouring objects. This technique is particularly applicable to heterogeneous imagery or the extraction of discrete features such as buildings, roads, and tree canopies. Similar to the coastline and land cover analyses, the GIS analysis will examine changes in the linear aspect of the river channels by measuring features such as channel width at every 30 m (equivalent to every pixel) and sinuosity at every 500 m, and examining the formation of new ‘permanent’ (i.e. not dry season) channels and shifts in channels. The location of these changes will be evaluated in relation to such factors as distance from coast and the surrounding land cover types. Temporal and spatial environmental factors such as catchment –and agricultural development (and the construction of the irrigation system in particular) and rainfall will be used in the analysis to determine whether these have affected any change in the river morphology. In Fig. 5 both widening and narrowing of different parts of the river can be observed, as is indicated by the arrows. The most obvious change in the river flow, however, is the new channel breaking through the west bank in the 60
2006 image. The river has also become more sinuous near the river mouth in the same image. Land cover A myriad of land cover classification and change detection techniques exist, and these are constantly being altered to suit the specific study and study site. The most common approaches for change detection include either post-classification (Helmschrot & Flugel, 2002; Yuan et al., 2005) or image differencing techniques (Singh, 1989; Lunetta, 1999; Phua et al., 2007). However, many variants of these have been carried out over the years (see for example Vogelmann et al., 2001; Chen et al., 2005; Cakir et al., 2006). In the present study, independent pixel-based unsupervised classifications will be carried out for each image, the results of which will be analysed in a GIS. Likely classes to come out of these classifications include – but are not restricted to – mangroves, coastal forest, beach, saltpans, agricultural areas, and urban areas. The GIS analysis will examine which classes have changed, and where they have changed. Certain locations might be more prone to land cover change, such as, for example, coastal or inland areas; areas near or away from river channels; and agricultural or urban areas. How these areas and classes have changed will also be explored, including what the cover
aliran-aliran dan perubahan arah aliran-aliran sungai tersebut. Lokasi perubahan-perubahan ini akan dievaluasi sehubungan dengan faktor-faktor seperti jarak dari pantai dan jenis-jenis permukaan tanah disekitarnya. Faktor-faktor lingkungan spasial dan bersifat temporal seperti daerah tangkapan air dan pembangunan pertanian (dan kususnya pembangunan sistem irigasi) dan hujan akan digunakan dalam analisa untuk menentukan apakah faktor-faktor ini menyebabkan perubahan pada morfologi tanah. Dalam Gambar 5 bagian-bagian sungai yang melebar dan menyempit bisa dilihat seperti yang ditunjukkan oleh tanda panah. Perubahan arus sungai yang paling jelas terlihat adalah aliran baru yang menerobos tepi barat sungai yang terdapat pada gambar tahun 2006. Sungai kemudian menjadi lebih berlikuliku di daerah dekat mulut sungai yang terlihat pada gambar yang sama. Permukaan tanah Ada banyak teknik-teknik klasifikasi permukaan tanah dan pendeteksian perubahan yang terjadi di mana teknik-teknik tersebut terus disesuaikan dengan penelitian dan tempat penelitian. Pendekatan yang paling umum dipakai untuk mendeteksi perubahan adalah pasca klasifikasi (Helmschort & Flugel, 2002; Yuan et al., 2005) atau teknik membedakan gambar (Singh, 1989; Lunetta, 1999; Phua et al., 2007). Namun teknik-teknik ini sudah digunakan selama bertahun-tahun (lihat Vogelman et al., 2001; Chen et al., 2005; Cakir et al., 2006). Dalam penelitian ini klasifikasi tanpa supervisi dengan pixel independen akan dilakukan dalam setiap gambar, hasilnya akan dianalisa dalam GIS. Biasanya hasil-hasil dari klasifikasi ini adalah gambar-gambar bakau, hutan pantai, pantai, daerah endapan garam, daerah pertanian, dan daerah urban.
Fig.6: Areas of prominent land cover changes in the Caraulun river delta during the period 1986-2006. GIS akan menganalisa hal-hal apa yang berubah dan lokasi terjadinya perubahan tersebut. Beberapa lokasi tertentu mungkin akan lebih mudah mengalami perubahan di permukaannya seperti daerah pantai atau daerah yang jauh dari pantai, daerah yang dekat atau yang jauh dari aliran sungai; dan daerah pertanian atau perkotaan. Bagaimana perubahan-perubahan ini terjadi juga akan diteliti misalnya; berubah menjadi apa daerah-daerah tersebut dan apakah terjadi peningkatan atau penurunan. Faktor-faktor lingkungan spasial dan temporal seperti yang disebutkan di atas dan aktifitas antropogenik akan digunakan untuk
Remote sensing applications for mapping and monitoring 61
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
types have changed to, and whether the classes have increased or decreased in area. Temporal and spatial environmental factors, such as those mentioned above, and anthropogenic activity will be taken into account in order to determine whether any relationships exist, or whether they could be drivers for the change. Fig. 6 shows the areas of major land cover change in the delta. The area in the largest circle is the irrigated agricultural area, with the irrigation system being located on the east bank of the river in the top left corner of the image. The change from forest/wetland to agriculture is evident and the construction of roads and buildings from 1996 to 2006 can also be observed. The smaller circle encompasses the village of Betano, which has been expanding since 1986, although an increase in forest can be observed from 1986 to 1996. The arrow indicates changes in vegetation type and forest cover, where the new river channel has broken through in the 2006 image. At this stage it is unclear what effect, if any, the flood that destroyed the irrigation system had on the land cover and flow of the river in this area.
62
Conclusion In summary, a range of remote sensing and GIS techniques is being used to determine change in the Caraulun river delta from 1986-2006, focussing on changes in the coastline, land cover, and river morphology. This project is ongoing, so the final methods used will depend on the success of use and the accuracy of their results. It is hoped that the results from this study can create an overall picture of how the delta has changed over the past 20 years and perhaps indicate how these changes might relate to spatial and temporal environmental factors, natural processes, and anthropogenic activity; and eventually aid in the overall picture of change and management for the whole of the Caraulun catchment.
menentukan apakah ada hubungan–hubungan atau apakah mereka bisa menjadi penyebab perubahan. Gambar 6 menunjukkan perubahan besar di permukaan tanah delta. Area di lingkaran terbesar adalah daerah pertanian yang teririgasi dengan sistem irigasi yang terletak di tepi timur sungai di sebelah kiri atas gambar. Perubahan di hutan/tanah basah dan daerah pertanian serta pembangunan jalan dan gedung dari tahun 1996 sampai 2006 dapat dilihat. Lingkaran kecil mengelilingi desa Betano, yang berkembang sejak tahun1986, meskipun peningkatan perubahan dalam hutan juga dapat terlihat dari tahun 1986 sampai 1996. Tanda panah menunjukkan perubahan-perubahan pada jenis tumbuhan dan permukaan hutan di mana terdapat aliran sungai baru dalam gambar tahun 2006. Pada tahap ini akibat banjir (yang menghancurkan sistem irigasi) terhadap permukaan tanah dan arah aliran sungai belum jelas.
sedang berlangsung, jadi metode terakhir yang digunakan akan tergantung pada keberhasilan dan akurasi hasil-hasil dari metode tersebut. Diharapkan bahwa hasil-hasil dari penelitian ini dapat memberikan informasi menyeluruh tentang bagaimana delta telah berubah dalam 20 tahun terakhir dan mungkin dapat menunjukkan hubungan perubahan-perubahan ini dengan faktor-faktor lingkungan spasial dan temporal, proses-proses alam, dan aktifitas antropogenic; dan akhirnya membantu dalam memberikan gambaran menyeluruh tentang perubahan yang terjadi dan manajemen di seluruh daerah tangkapan air Caraulun.
Kesimpulan Sejumlah teknik-teknik penginderaan jarak jauh dan GIS sedang digunakan untuk menemukan perubahan yang terjadi di delta sungai Caraulun dari tahun 1986-2006 yang berfokus pada perubahan garis pantai, permukaan tanah dan morfologi sungai. Proyek ini
Remote sensing applications for mapping and monitoring 63
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
Acknowledgments This project is part of the author’s Masters project, which is affiliated with the “Timor-Leste coastal/marine habitat mapping for tourism and fisheries development” project – a collaborative project between the Ministry of Agriculture, Forestry, and Fisheries (MAFF) of the government of Timor-Leste, Charles Darwin University (CDU), the Northern Territory government, Australian National University, James Cook University, and the Australian Institute of Marine Science. Acknowledgments for help in the field go out to Alexio Leonita Amaral (MAFF), Joao Guterres (MAFF), Francisco da Silva (MAFF), and Oscar da Silva (MAFF). References Ahmad W. and Hill G.J.E. 2003. Use of varied resolution remotely sensed data for the management of tropical savannas, north Australian case study. Proceedings of the South East Asian Natural Resources and Environment Management. Sabah, Malaysia. (University of Malaysia) Asian Development Bank 2004. ‘Water for all – Water for growth’. Asian Development Bank, TA:TIM 3986. Boix-Fayos C., Barbéra G.G., López-Bermúdez F. and Castillo V.M. 2007. Effects of check dams, reforestation and land-use changes on river channel morphology: Case study of the Rogativa catchment (Murcia, Spain). Geomorphology 91: 103-123. Bouma G.A. and Kobryn H.T. 2004. Change in vegetation cover in East Timor, 1989-1999. Natural Resources Forum 28: 1-12. Cakir H.I., Khorram S. and Nelson S.A.C. 2006. Correspondence analysis for detecting land cover change. Remote Sensing of Environment 102: 306-317. Chen S.-s., Chen L.-f., Liu Q.-h., Li X. and Tan Q. 2005. Remote sensing and GIS-based integrated analysis of coastal changes and their environmental impacts in Lingding Bay, Pearl River Estuary, South China. Ocean & Coastal Management 48: 65-83. Chu Z.X., Sun X.G., Zhai S.K. and Xu K.H. 2006. Changing pattern of accretion/erosion of the modern Yellow River (Huanghe) subaerial delta, China: Based on remote sensing images. Marine Geology 227: 13-30. Fan H., Huang H., Zeng T.Q. and Wang K. 2006. River mouth bar formation, riverbed aggradation and channel migration in the modern Huanghe (Yellow) River delta, China. Geomorphology 74: 124-136. Foody G., Muslim A. and Atkinson P. 2005. Super-resolution mapping of the waterline from remotely sensed data. International Journal of Remote Sensing 26: 5381-5392. Frihy O.E., Dewidar K.M., Nasr S.M. and El Raey M.M. 1998. Change detection of the northeastern Nile Delta of Egypt: shoreline changes, Spit evolution, margin changes of Manzala lagoon and its islands. International Journal of Remote Sensing 19: 1901-1912. Helmschrot J. and Flügel W.-A. 2002. Land use characterisation and change detection analysis for hydrological model parameterisation of large scale afforested areas using remote sensing. Physics and Chemistry of the Earth 27: 711-718.
64
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
Kaufmann R.K. and Seto K.C. 2001. Change detection, accuracy, and bias in a sequential analysis of Landsat imagery in the Pearl River Delta, China: econometric techniques. Agriculture, Ecosystems and Environment 85: 95-105. Lillesand T.M. and Kiefer R.W. 2000. Remote sensing and image interpretation, 4th Ed. John Wiley & Sons, New York. 724pp Lunetta R.S. 1999. Applications, project formulation, and analytical approach. In Remote Sensing Change Detection: Environmental Monitoring Methods and Applications. Eds. R.S. Lunetta and C.D. Elvidge. Taylor & Francis Press, London. pp. 1-19. Nevins J. 2002. (Mis)representing East Timor's past: structural symbolic violence, international law, and the institutionalization of injustice. Journal of Human Rights 1: 523-540. Nevins J. 2003. Restitution over coffee: truth, reconciliation, and environmental violence in East Timor. Political Geography 22: 677-701. Phua M.-H., Tsuyuki S., Furuya N. and Lee J.S. 2007. Detecting deforestation with a spectral change detection approach using multitemporal Landsat data: A case study of Kinabalu Park, Sabah, Malaysia. Journal of Environmental Management doi:10.1016/j.jenvman.2007.04.011. Restrepo J.D. and López S.A. 2007. Morphodynamics of the Pacific and Caribbean deltas of Colombia, South America. Journal of South American Earth Sciences doi:10.1016/ j.jsames.2007.09.002. Ryu J.-H., Won J.-S. and Min K.D. 2002. Waterline extraction from Landsat TM data in a tidal flat. A case study in Gomso Bay, Korea. Remote Sensing of Environment 83: 442-456. Sandlund O.T., Bryceson I., de Carvalho D., Rio N., da Silva J. and da Silva M.I. 2001) Assessing Environmental Needs and Priorities in East Timor: Issues and Priorities. UNOPS Report. Silove D. 2000. Conflict in East Timor: Genocide or Expansionist Occupation? Human Rights Review 1: 62. Singh A. 1989. Digital change detection techniques using remotely sensed data. International Journal of Remote Sensing 10: 989-1003. Syvitski J.P.M. and Saito Y. 2007. Morphodynamics of deltas under the influence of humans. Global and Planetary Change 57: 261-282. Tucker C.J., Grant D.M. and Dykstra J.D. 2004. NASA's Global Orthorectified Landsat Data Set. Photogrammetric Engineering & Remote Sensing 70: 313-322. Vanacker V., Vanderschaeghe M., Govers G., Willems E., Poesen J., Deckers J. and De Bievre B. 2003. Linking hydrological, infinite slope stability and land-use change models through GIS for assessing the impact of deforestation on slope stability in high Andean watersheds. Geomorphology 52: 299-315. Vogelmann J.E., Helder D., Morfitt R., Choate M.J., Merchant J.W. and Bulley H. 2001. Effects of Landsat 5 Thematic Mapper and Landsat 7 Enhanced Thematic Mapper Plus radiometric and geometric calibrations and corrections on landscape characterization. Remote Sensing of Environment 78: 55-70.
65
GIS APPLICATIONS FOR SUSTAINABLE DEVELOPMENT AND GOOD GOVERNANCE
Wasson R.J., Juyal N., Jaiswal M., McCulloch M., Sarin M.M., Jain V., Srivastava P. and Singhvi A.K. 2008. The mountain-lowland debate: Deforestation and sediment transport in the upper Ganga catchment. Journal of Environmental Management 88: 53-61. White K. and El Asmar H.M. 1999. Monitoring changing position of coastlines using Thematic Mapper imagery, an example from the Nile Delta. Geomorphology 29: 93-105. Woodroffe C.D. 2003. ‘Coasts: Form, process and evolution.’ Cambridge University Press, Cambridge, UK. 623pp Wyatt A.S.J. 2004. ‘Preliminary survey of the nearshore coastal marine environment of the south coast of East Timor: a baseline for assessing the impacts of a developing nation.’ Honours Thesis thesis, University of Western Australia. Yamano H., Shimazaki H., Matsunaga T., Ishoda A., McClennen C., Yokoki H., Fujita K., Osawa Y. and Kayanne H. 2006. Evaluation of various satellite sensors for waterline extraction in a coral reef environment: Majuro Atoll, Marshall Islands. Geomorphology 82: 398-411. Yang X., Damen M.C.J. and van Zuidam R.A. 1999. Satellite remote sensing and GIS for the analysis of channel migration changes in the active Yellow River Delta, China. International Journal of Applied Earth Observation and Geoinformation 1: 146-157. Yuan F., Sawaya K.E., Loeffelholz B.C. and Bauer M.E. 2005. Land cover classification and change analysis of the Twin Cities (Minnesota) Metropolitan Area by multitemporal Landsat remote sensing. Remote Sensing of Environment 98: 317-328. Yue T.X., Liu J.Y., Jorgensen S.E. and Ye Q.H. 2003. Landscape change detection of the newly created wetland in Yellow River Delta. Ecological Modelling 164: 21-31. Zhao B., Guo H., Yan Y., Wang Q. and Li B. 2008. A simple waterline approach for tidelands using multi-temporal satellite images: A case study in the Yangtze Delta. Estuarine, Coastal and Shelf Science 77: 134-142.
66