REMOTE SENSING-BASED INPUTS TO RAINFALL RUNOFF MODELS: A CASE STUDY FROM THE TIGRIS-EUPHRATES WATERSHED

Monitoring systems which are needed to estimate runoff are absent on a regional scale for the majority of the Earth’s surface and it is unlikely that such systems will be in place in the near future given the efforts and resources needed to construct and maintain such systems. Readily available global remote sensing data sets is providing avenues to address these deficiencies. In the Tigris-Euphrates watershed, our test area, the paucity of stream flow and rain gauge data is limiting our ability to construct calibrated rainfall - runoff models. Existing rain gauge data is being supplemented by TRMM data that provides global (50°N – 50°S) rainfall data, every three hours with a 0.25° x 0.25° footprint. We are examining the utility of various satellite-based (e.g., SSM/I, T/P Altimeter, and ERS-1 SAR) techniques under varying topographic (e.g., mountains, foothills, valley flat) and climatic (e.g., temperate, arid, semi-arid) conditions, and river morphologies (e.g., braided stream) to estimate river discharge. We extracted empirical expressions that relate in-situ river discharge to satellite-based, river discharge-dependant variables (SSM/I: Brightness temperature, T/P Altimeter: River level, ERS-1 SAR: River width). Specifically, satellite-based above variables were correlated to in-situ flow data (1964 -1997) from four gauging stations: the Keban (Lat: 38.7925°N, Long: 38.7439°E), Hit (Lat: 33.6436°N, Long: 42.8215°E), Baghdad (Lat: 33.3157°N, Long: 44.3922°E), and Mosul (Lat: 36.3397°N, Long: 43.1425°E). Preliminary results show a good correspondence between in-situ stream flow data from the Baghdad and Hit gauge stations and brightness temperatures obtained from SSM/I data (pixel size: 25x 25 km) over the Baghdad and Hit gauge stations and neighboring pixels downstream from the two stations (Baghdad: R²=0.7213; Hit: R² =0.6347). Current and future research will focus on: (1) determining which of the applied methodologies, and under which settings/conditions, produce optimum discharge results, (2) extracting discharge and precipitation data for time periods where in-situ data is absent, (3) using the extracted stream flow and precipitation data to construct and calibrate continuous rainfall-runoff models for the watershed.