REMOTE SENSING, GEOGRAPHIC INFORMATION SYSTEM AND HYDROLOGICAL MODEL FOR RAINFALL-RUNOFF MODELING

摘要

Flood is an unavoidable natural hazard. Flooding causes a lot of damage such as decay infrastructure, transportation, economic lost, human health and death. The study of rainfall-runoff relationship is important in flood forecasting and water resources planning. The aim of this study is to develop an operational framework for flood forecasting by integration of remote sensing, geographic information system (GIS) and hydrological model. Study area located at Kota Tinggi, Johor. This study is divided into three main parts. The first part is preparation of data sets for the study, such as topographic map, soil map, land use map, Landsat 5 image, DEM (Aster) and hydrological data. Flood event of 2006 rainfall and stream flow data obtained Rantau Panjang rain gauge station. The second part involved automatic catchment delineation, flow direction mapping and stream order mapping by Arc Hydro model. Landsat image was classified by maximum likelihood method and gives overall accuracy 80.84% and kappa coefficient 0.7498. Manning's roughness coefficient for each sub-catchment extract from land use map generated as input into the model. Understanding and analysis of catchment characteristic is important in better rainfall runoff modeling. The final part is concern about rainfall-runoff modeling, simulation rainfall-runoff discharge of catchment for single event and analyzed the runoff hydrograph. The observed and simulated peak stream flow recorded at 21st DEC 2006 5.00am and 0.00am are 375.266m~3/s and 370.163m~3/s respectively. In the validation part, linear regression method was used to correlate observed and simulated stream flow data and gives correlation coefficient 0.8842. The results of the study show that integration of Remote Sensing, GIS and hydrological models can solve hydrological problems.