Abstract
The objective was to map landslide hazards throughout Thailand in future climates. To achieve this goal, a logistic regression model of landslides was developed using topographic and hydrologic data as its explanatory variables. For the future prediction of extreme rainfalls, precipitation data from 3 GCMs of CMIP5 (MIROC5,MRI-cgcm3,and GFDL-esm2g) in 4 RCP scenarios (RCP2.6, 4.5, 6.0, and 8.5) was retrieved and analyzed in 3 climatic periods (near-future, intermediate-future, and far-future climate). The results were as follows. 1) In a case of MIROC5 in RCP8.5 in the intermediate-future climate, the estimated extreme daily rainfall value of a 5-yr return period was 18% more than the current climate. 2) Northern and middle-western mountainous areas, and western and central Malay Peninsula, showed the largest hazards in Thailand. 3) In a case of MIROC5 in RCP8.5 in the intermediate-future climate, the estimated landslide hazard was 27% more than the current climate.