SENSITIVITY EXPERIMENTS OF GLOBAL RIVER MODELS TO DIFFERENT PHYSICAL PROCESSES AND ELEVATION DATA AND CHANGES IN FLOOD RISK

Abstract

 Recent studies have projected the future flood risk caused by climate change, but there is uncertainty in these projections. In particular, previous studies have pointed out the variability of flood risk due to differences in the climate models and scenarios used, but it is also necessary to evaluate the uncertainty caused by model simulations. In this study, we conducted sensitivity experiments of a river model, CaMa-Flood, to investigate the impact of recent advances in satellite observation and numerical methods on global river flood simulation and global flood risk estimation. The results showed that the difference in elevation data caused by errors in satellite observations have a significant impact on the inundation distribution. Regarding the physical processes of the model, the temporary flow of flood waters along the river has the greatest impact on the difference in inundation area, with a difference of about 5.5% at most. It is also found that the global population exposed to floods differs by up to 14% depending on the physical processes of the model and the input data.