Overlay analysis of GIS Layers from an Integrated Geospatial Database to Evaluate Changes on AL Sammalyah Island Salem Essa Geology Department U.A.E. University, Al-Ain, P.O. Box: 17555, U.A.E. salem.essa@uaeu.ac.ae Ronald Loughland Environmental Research Department Emirates Heritage Club, Abu Dhabi P.O. Box: 41464, U.A.E. Mohamed E. Khogali Environmental Research Department Emirates Heritage Club, Abu Dhabi P.O. Box: 41464, U.A.E. Abdulmunem Darwish Environmental Research Department Emirates Heritage Club, Abu Dhabi P.O. Box: 41464, U.A.E. Abstract GIS is extensively used to study and analyze different surface features of ALSammalyah Island. 15 GIS layers were created to host spatial information about land surface. Layers include large-scale historical aerial photographs from the seventies, the eighties, the nineties, as well as recent aerial photographs from 1999/2000 and 2005 years. These two dates were chosen to conduct a change detection analysis during the first phase of this study. The remaining 11 layers of the database include thematic vector layers derived from the 1999/2000 and 2005 aerial photographs. During the first phase of the study, change detection analysis was performed to evaluate the changes affecting the island on six years period between 1999 and 2005. The results demonstrated total vegetation cover increase from 3.742 km2 in 1999 to 5.101 km2 in 2005, an increase of 36.3% in six years. The study also showed that this increase in vegetation extent is mostly attributed to the increase in mangrove planted areas with an amount of 2.256 km2 in 1999 to 3.568 km2 in 2005, a percentage increase of 58.2% in six years. At this stage of the project the focus is on establishing an integrated Geographic Information System (GIS) database including multi dates imageries, -six different dates, as well as derived vector thematic layers from two specific dates, namely the 1999 and the 2005. GIS overlay analyses are being implemented to investigate eventual spatial relationships between the different environmental variables. 1. INTRODUCTION This paper covers the first and second phases of a research project for building a GIS database for AL Sammalyah Island. The AL Sammalyah GIS database will assist in planning for the conservation and management of the natural resources and the development of ecotourism activities on the island. The island has witnessed high rates of change in land use in the past few years; environmental management approaches have been implemented since the mid 1990s with a vision for improving the decision making process and providing quick and accurate services to researchers and the ecotourism community in general. The building of an integrated spatial database, during the current phase, is an important step towards the achievement of the actual research objectives. A spatial database containing 15 GIS layers was built; both raster and vector layers were integrated; the data is projected in the UTM zone 40N. The outcome of the change detection analysis task, conducted during the first phase of the project, was also added to the database. The integration of multi-scale aerial photographs in a GIS database is of paramount importance and has a high cost to benefit ratio. The building of the actual GIS database is essential for sustainable development and management of the island in the long term [9], [20]. The objectives of the present study are summarized below:
2. BACKGROUND In recent years geographic databases have become increasingly large and complex. For example, AirPhoto USA's US National Image Mosaic is 25 terabytes (TB) in size, EarthSat's global Landsat mosaic at 15 m resolution is 6.5 TB, and Ordnance Survey of Great Britain has approximately 450 million vector features in its MasterMap database covering all of Britain [21]. The creation of a spatial GIS database is a major objective of the actual study. To emphasize the issues related to the management of the island resources, multi date large scale aerial photographs were scanned, georectified and integrated into the spatial database. Six different dates were considered, the only available data!, mid-1970s, mid-1980s, 1999, 2000 and 2005 aerial photographs with spatial resolution 1 to 2 meters. Vector data were derived from the 1999 and 2005 by on-screen digitization to evaluate the change occurred on the island during this active period of management and laying down of infrastructures on the island. A total of 11 different GIS layers were produced namely: roads, roundabouts, footpaths/tracks, buildings, breaks/bays/ports/petrol station, palm trees, mangroves, shrubs, water bodies, water channels, and barren land. A DEM will be generated and added to the database on a later stage because cuts, filling, and the creation of new artificial hills continue to be done during our study. Studying land resources management using spatial related technology starts by assessing the amount of changes occurred during the past active period of heavy work and the establishment of a modern roads network as well as the construction of many heritage buildings. Change detection is defined as the process of identifying differences in the state of an object or phenomenon by observing it at different times. The basic principle in using remotely sensed data for change detection is that: changes in the objects of interest will result in changes in reflectance values or local textures that are separable from changes caused by other factors such as differences in atmospheric conditions, illumination and viewing angles, and soil moistures [29], [22], [1], [6], [2], [13], [7], [26], [30], [16], [6], [4], [24], [25]. 3. STUDY AREA AL Sammalyah is located at approximately 24o 26'10"N - 24o 28'56?N and 54o 29'22?E - 54o 34'12?E (figure1). Situated in the Arabian Gulf, about 12 km north east of Abu Dhabi Island near (Um al-Nar Island) area and just opposite Shati’-Al Raha beach, is characterized by its rich ecosystems and marine life. The Island area covers about 14 square kilometers. Soil texture is dominated by sand, with high salt content reaching 31.5% TSS on the surface, being mostly Chloride soluble salts giving a white color to the soils of the island, which produces high brightness levels on satellite imagery operating in the visible portion of the EMR, however the very narrow mangrove soils surrounding the island have dark color because of high content of organic carbon content. The Island is a natural reserve containing high biodiversity, particularly mangroves (Avicenna marina), which grow throughout the coastal areas of Abu Dhabi Emirate and is associated with varieties of plant communities including: Arthrocnemum macrostachyum, Seidlitzia rosmarinus, Suaeda vermiculata, and Cyperus conglomerates, [11], [9].  figure 1. Location of the study area A change detection analysis using multi-sensor and multi-temporal aerial photographs and satellite imagery can help in emphasizing changes in Land Use/Land Cover (LULC) that has occurred since the establishment of the DER in 1996, and to assess the degree of success in the implementation of the DER's mandate in protecting and developing the ecosystems of the island. 4. DATA A temporal series of remote sensing data was acquired for the study area. It was composed of aerial photographs from mid-1970s, mid-1980s, 1999, 2000, 2005, and 2006. For conducting the change detection study, aerial photographs from 1999 and 2005 were used; this is because changes were visible on available aerial photographs only from the 1990s. No major activities were noticed from the mid-1980s aerial photographs. Most aerial photographs were procured at the MSD (Military Survey Department), and Abu-Dhabi Municipality. The software used includes ESRI ARCGIS9 software for vector processing, and ERDAS Imagine 8.4 for image processing. The hardware used includes PC Pentium IV 3.2 GH speed, and HP color LaserJet printers. The selection of data, used in this study was largely governed by availability and accessibility of archived data especially at the MSD archives; the selection of hardware, and software was governed by the UAEU geology department's facilities. 5. METHODOLOGY The methodology applied to the change detection analysis in this study is based on integrating GIS with remote sensing method, including the advantage of using GIS ability to incorporate different source data into change detection applications. Available high resolution aerial photographs of 1999 and 2005 were used, and a post-classification comparison method is implemented. Visual analysis method was used for image interpretation and simultaneously an on-screen digitizing of changed areas was performed. Texture, shape, size and patterns of the images were used for identification of LULC change through visual interpretation. A unified land cover classification scheme was established for classification of images. Eleven land cover classes were identified in the image interpretation process (Table 1). The classified images were then used to derive class area statistics and class patterns over the past 6 years. The output layers are brought into a GIS database [29], [30], [4], [5], [23], [8], [13], [1], [14], [25], [15]. The methodology applied to the GIS database building follows the following general steps:
 6. RESULTS Production of a set of imagery spanning the period from the mid-1970s to the very new 2006 aerial photo. Imageries are geometrically corrected, co-registered, resampled to a pixel resolution of 2 meters and projected to the UTM projection zone 40N. Eleven land cover classes were produced and converted to vector format for the 1999 and 2005. Two land cover maps were produced for 1999 and 2005. Eleven land cover classes were mapped and categorized into four groups, these include:
 Figure 2: Green vegetated areas  Figure 3: Building construction  Note: Decrease carries negative sign while increase carries positive sign. 7.1. Change Detection Analysis 7.1.1. Urbanization: changes in roads, roundabouts, and building (1999 - 2005) Table 2 shows that there has been an increase of 23.6%, 300% and 220% in roads length, number of roundabouts and buildings areas respectively between 1999 and 2005. This high increase of the level of urbanization in the island required large amounts of investment from the managing organization known as the Emirates Heritage Club, to satisfy the increased demand to build new touristic, scientific and sports facilities; e.g. horse riding halls, Olympic shooting range, Olympic swimming pools, laboratories, green houses, offices, dormitories, and the establishment of a modern road network. These facilities are used to host increasing numbers of students from all around the country as well as researchers and to emphasize ecotourism. This is attributed to the high-level political will of H.H Sheikh Sultan Bin Zayed, Chairman of the Emirates Heritage Club to transform the island from a desertified area into a well-developed reserve for protecting wild life and ecosystems and for developing ecotourism and preserving local heritage. 7.1.2. Forestation, agricultural activities and environmental development: changes in vegetation cover extent (1999 - 2005) Total vegetation cover extent has increased from 3.742 km2 in 1999 to 5.101 km2 in 2005, an increase of 36.3% between 1999 and 2005. The density of the vegetation has also increased giving increase to the overall biomass production on the island. This biomass production can be quantified using vegetation indices calculated from satellite imagery and will eventually, be one of the objectives of future studies using satellite imagery. Furthermore, this increase in vegetation extent is mostly attributed to the increase in mangrove planted areas; indeed mangrove planted area has increased from 2.256 km2 in 1999 to 3.568 km2 in 2005, an increase of 58.2% in about six years. This gain in the mangrove –occupied areas has occurred on previously barren land which has been reclaimed and irrigated using sea water hence indicating the success of procedures applied to combat desertification, enhance biodiversity and sustain the environment on the island. Mangrove is highly adapted to the conditions of the island and is known for its high salt-tolerance, as a soil stabilizer, an ideal marine and bird habitat, and can sustain bees' elevation giving uniqueness to mangrove honey production. This success in the extent of vegetated areas especially mangrove plantations, gives evidences that the UAE is wisely investing in the domain of scientific research oriented at producing species that are adapted to its climatic conditions and hence increasing green areas and combating desertification locally and in the region. 7.1.3. Desertification and land disturbance: changes in barren land extent and footpath lengths (1999 - 2005) Barren land extent was studied as an indicator of desertification and abundance of green vegetation throughout the island, while tracks and footpaths lengths are used as an indicator of land disturbance and hence another indicator of desertification. There has been a decrease of 15 % in barren land areas between the two dates; while tracks and footpaths lengths have decreased by 32% between 1999 and 2005. This is another indicator of the decrease in desertification levels and the development of the ecosystems on the island during the study period. The loss in the bare and disturbed areas is due primarily to new vegetated areas (85%), urbanization (11%), and water bodies and water channels (4%). 7.1.4. Conservation and land reclamation: changes in water bodies and water channels (1999 - 2005) The spreading of water bodies and water channels is clear evidence of conservation and land reclamation. Table 2 shows that there has been an increase of 13.7% in the channel networks between the two data sets while water surfaces appear on the 2005 imagery only, with a large artificial salt water lagoon of 15,000 m2 (figure 4). Water surfaces are designed to attract migrant birds during the winter season and to develop aquatic life on the island, while channels are used to bring high-tide sea current generated water into mangrove areas. 7.2. Database building Al Sammalyah GIS database contain both raster and vector data, large scale aerial photographs were scanned and converted to ArcGIS geodatabase format. Vector data were obtained from on-screen digitization of the 1999 and 2005 datasets, and are added to the geodatabase. An integrated geodatabase of six different dates of large scale aerial photographs, 11 vector layers spanning 3 different dates were obtained totalizing 33 vector layers showing the status of the land cover on the island during 1999 and 2005 dates, and assessing the rate and nature of change occurring on the island. Results indicate good progress in the levels of greening of the island, especially in the increase of the salt-tolerant mangrove plantation during the study period. A geodatabase of about 40 GIS layers is now available to the decision makers of the island for the best management of its land resources. CONCLUSION On the island large-scale reclamation started in the early 1990s and has increased very rapidly since then. This is confirmed by the decrease in bare land and the increase in vegetated areas especially plantations of salt-tolerant mangroves and palm trees. Urbanization and the spread of water bodies is testimony to the development of the island for enhancing scientific research and developing the ecosystem. ACKNOWLEDGEMENT The authors are grateful to the following: (i) the Emirates Heritage Club, DER, for the financial support to carry out this work; (ii) the Research Affaires at the UAE University for their assistance and continuous support to this research; and finally (iii) Mr. G. Abdul Fattah for his assistance in digitizing vector data, preprocessing of aerial photos and in preparing some figures for this paper. REFERENCES
|