DeutschClear Cookie - decide language by browser settings
Haider, Khaled (2011): Using Spatial Data for Geo-Environmental Studies. Dissertation, LMU München: Faculty of Geosciences



The physically-based spatially-distributed model PROMET (Processes of Radiation, Mass and Energy Transfer) is applied to the Greater Damascus Basin, which is considered as one of the most important basins in Syria, to serve as a case study of using spatial data for Geo-environmental studies. Like most areas of the Middle East, the study area is characterized by large temporal and spatial variations in precipitation and by limited water resources. Due to the increasing water demand caused by the economic development and the rapid growth of population, the study area is expected to suffer from further water shortages in the future. This highlights the necessity of developing an integrated Decision Support System (DSS) to evaluate strategies for efficient and sustainable water resources management in the basin, taking into consideration global environmental changes and socio-economic conditions. The work presented here represents the first steps toward achieving this goal through applying a distributed hydrological model (an important component of any integrated DSS for water resources management) to the Greater Damascus Basin utilizing different types of spatial data used as time-dependent (e.g., meteorology) and time-independent (e.g., topography and soil) input parameters. The model PROMET, which was developed within the GLOWA-Danube project as part of the decision support system DANUBIA, is run on an hourly time step (for the period from 1991 to 2005) and a 180*180m spatial resolution to simulate the water and energy fluxes in this basin. The model is embedded within a raster-based GIS-structure which facilitates the integration of the diverse types of spatial data. The spatial information related to topography (such as elevation, slope, and exposition) as well as those related to runoff routing (such as upstream-area, channel width, and downstream proxel) are automatically extracted from Digital Elevation Model (Shuttle Radar Topography Mission, SRTM-90m DEM). The spatial patterns of the different land use/land cover classes are derived from remote sensing data (classification of a cloud-free LANDSAT 7 ETM+ image using the supervised classification algorithm). The spatial fields of meteorological input data are provided on an hourly basis through spatiotemporal interpolation of the measurements of the available weather stations. Spatial information about the soil texture is provided through generalization and aggregation of the soil type classes of the Soil Map of Syria (prepared by USAID) and transferring the soil types to texture classes. Several pedotransfer functions are then used to estimate the soil hydraulic properties for each soil texture class (and each soil layer) found in the study area. While plant physiological parameters (which are assumed to be static, such as minimum stomatal resistance) are estimated for each vegetation class using information taken from literature sources, the temporal evolution of Albedo and Leaf Area Index (LAI) are derived from five cloud-free LANDSAT-7 images acquired at different seasons of the year. The goodness of the results obtained by the model PROMET are verified and/or validated by comparing them either with their corresponding data observed in the filed or with remote sensing-derived information (e.g., snow cover). Two subcatchments are selected for the purpose of calculating the spatially-distributed annual water balances. The results indicate that the modelled mean annual runoff volume fits well with the measured discharge for both chosen subcatchment. In addition, the simulated discharge is compared to the observed one (at seven gauge stations) on a monthly basis, covering the whole simulation period (15 years). The results of the regression analysis for each of these gauge stations (with slope of regression line ranges from 0.79 to 1.04; coefficient of determination 0.69-0.90; and Nash-Sutcliffe Coefficient 0.73-0.95) indicate that there is a good correlation between simulated and observed monthly mean discharge volumes.