Investigation of Buyukcekmece water basin by using active remote sensing data and GIS-supported hydrologic model

Together with changing conditions of the rapidly developing world, use of our water resources for various purposes is also on the rise. Increase in population as well as increase in migration to cities have major impacts on our water resources. Furthermore, the composition of the air together with water and soil has significantly been spoiled due to damages that occurred in the last century in nature. The environment and the atmosphere is heavily polluted by greenhouse gases released from settlement and industrial regions that grow rapidly and on a global scale the temperatures tend to rise more and more rapidly. As a consequence, today we are facing global climate changes. Together with the speeding climate change particularly after the industrial revolution, the existing envvironmental conditions including hydrological and ecological ones are undergoing a rapid change. With the effect of the warming in the atmosphere together with climate change, the hyrdological cycle is also subject to change. The change in composition of climate and hydrological cycle also affect water resources. Therefore, identification, protection and sustainable management of water resources have gained substantial importance. When planning on water resources, it is a must to conduct good basin management planning as well. This study aims at planning the use of the soil, one of the most significant natural resource, monitoring its changes and developing a model to provide spatial data on components of soil, particularly soil humidity to decision makers. Furthermore, with the help of the hydrological model applied within this work, various data was obtained from water budget components of Büyükçekmece Reservoir that meets a significant amount of water need in Istanbul today including real evapotranspiration, percolation, ground humidity, potential evapotranspiration, percolation, ground floww, total flow, annual precipitation, surface flow and snowmelt values and their temporal changes. With the help of the model applied in the thesis,a basis was povided to foresee changes in water budgets in cases of different scenarios that may occur in the future due to various reasons including population increase and use of terrain. At the first stage of th study, a total of 42 soil samples taken from different parts of Büyükçekmece basin, one of the largest potable water basins of İstanbul, were analysed for their soil humidity, lime, pH, carbon, nitrogen and salinity rates among soil components in laboratory. While number and distribution of the spots on the field were identified together with terrain work and coordinates data, certain parameters including terrain plantation and use, transportation infrastructure to the sample spots, altitude, soil type and positional distribution were evaluated within the Geographical Information System (GIS). Soil samples were taken on 16 July 2010, 09 August 2010 and 02 September 2010 but RADARSAT-1 satellite data was received only on 16 July and 2 September simultaneous to the days above. The type of the relation between values of the soil components obtained upon laboratory analyses and those from backscattered (betanought) from RADARAST-1 data was examined. The regression analyses conducted showed a significant relation between RADARSAT-1 and soil humidity from soil components, while other components of lime stone, pH, carbon, nitrogen and salinity have not shown any significant relation Furthermore, soil humidity maps for different dates were formed based on the RADARSAT-1 data by applying equations obtained upon regression analysis. It is foreseen that increasing the frequency of sampling locations collected locationally and using remote sensory data in identification of soil humidity in larger areas and particularly areas with spatial challenges will provide more sensitive and reliable results. At the second stage of the study, maps on soil components were generated by using different interpolation methods. Initially spatial soil component maps were generated by using the spoits where soil samples were taken with the Ordinary Kriging method, one of the geostatistical methods, to define soil pH, lime and salinity content. Soil humidity results on these points, backscatter values of Radarsat-1 data and different methods of interpolation were used to generate spatial soil humidity map of the study area and different spatial anaylses were conducted with the integration of these maps with the GIS. Consequently, 11 control points - 20% of the total points - were identified to conduct accuracy analysis of the model; first with 25 points, then secondly 10 homogeneously distributed points from the 25 points were identified as controlpoints to apply regression analyses and different interpollation methods Inverse Distance Weighting (IDW), Global Polynominal Interpolation (GPI), Local Polynominal Interpolation (LPI), Radial Basis Functions (RBF) Kriging, Kokriging were used to compare results among each other. The evaluations showed that generally the most accurate results were obtained with Ordinary Kriging method while other interpolation methods also gave good results when spatial evaluation was conducted, that the higher number of points gave more optimum results and that RMS errors were lower. Furthermore Inverse Distance Weighting (IDW), Local Polynominal Interpolation (LPI) ve Radial Basis Functions (RBF) methods also showed to calculate similar results to Ordinary Kriging while results obtained by radar data gave lower accurate results compared to other methods in this study area. At the final stage of the study, a hydrological modelling was conducted for Büyükçekmece basin which is the work field of the study by using Soil and Water Assesment Tool (SWAT) model software. The terrain cover from 1990 and 2006 generated from CORINE as model terrain cover / usage data to be converted into FAO data set and the impact of terrain cover / usage changes on the model outcomes were evaluated. At the outcome of the model values for both dates on real evapotranspiration, percolation, base flow, annual precipitation, surface flow and snowmelt water budget components were generated. As the real ET, ground humidity, percolation were identified to be more effective on base flow, graphics and maps of hydrological components for all the basin in two different periods were formed. The outcomes of hydrological component maps that were generated by using the 1990 and 2006 terrain cover maps were formed by taking annual averages of 1985-1990 and 1991-2010 periods. The purpose of such an evaluation is to examine the impact of terrain cover / use changes encountered in 1990 and 2006 to the components of the same periods of time. In order to interpret the data generated in the study, terrain cover / use change maps for lower basins as well as graphics for two different water budget component for each period were generated. While changes in terrain cover / use caused significant changes on the basis of water budget component; it was seen to have no such change in other ones. Again,the Total Flow data and basin areas was used to calculate water potential in order to foresee the water potential of Büyükçekmece basin with the water budget components generated as an outcome of the model.