Water-related disasters are often caused by the influence of local characteristics such as topography. Therefore, downscaling is necessary to estimate water-related disaster risk. However, high-resolution dynamic downscaling requires a huge amount of computation, while statistical downscaling is difficult to reflect local effects. Here, a downscaling method that can reflect local effects is developed and apply the products to an integrated land simulator (ILS) to improve the accuracy of regional water-related disaster risk assessment during climate change. Using this method, we estimated the spatial distribution characteristics of precipitation, temperature, surface wind, surface pressure, relative humidity, downward short- and long-wave radiation, and cloud cover, which are required as external forces for the ILS, and succeeded in reproducing changes in river discharge corresponding to heavy rain associated with meso-scale disturbances using ILS. In the future, we will conduct estimation using multiple models and multiple scenarios of climate model simulation to assess the risk of water-related disasters due to climate change.