Simulating rice yield by integrating GIS data and MODIS-LAI for Jeolla-do region in South Korea

Paddy rice is the main crop and food in South Korea. Rice yield estimation is of great significant to obtain timely and accurately at early stages, which plays the important role in decision making and agricultural policy. Rice yield is resulted from the integration of weather, soil, water supply, cultural practice and rice phonological properties at a certain field. In this study, we aim to examine Water Accounting Rice Model (WARM) for rice yield estimation by integrating GIS data and remote sensing data. The experimental was conducted on Jeolla-do region, South Korea in 2007. GIS data including meteorological data and soil characteristics; and remote sensing data (MODIS-LAI) were the major input for WARM model. Rice yield information at the same year from Korea Statistic was collected to verify WARM's results. As a consequence, it indicated that WARM's estimation is overestimated compared to Korea Statistic data. Root Mean Square Error (RMSE) and Relative Root Mean Square Error (%RMSE) of WARM were 35.11 kg/10A and 7.88%, respectively.