Uncertainty in simulated streamflow using runoff driven by the outputs of a high-resolution regional climate model

Abstract

Estimating river discharge using climate model output can aid in analyzing the potential impacts of climate change on water-related disasters. This study aimed to explore the uncertainty in simulated streamflow using the non-hydrostatic regional climate model (NHRCM) outputs in Thailand. The NHRCM was simulated at 5- and 2 km resolutions. To estimate runoff, two land surface models (LSMs) were employed: the Meteorological Research Institute–Simple Biosphere Model (MRI-SiB) in NHRCM and the Simple Biosphere including Urban Canopy (SiBUC). The NHRCM rainfalls captured the seasonal pattern of rainfall in the upper Ping River Basin, although they were underestimated. The 2 km NHRCM had less rainfall, but it captured the local topography better than the 5 km model. This difference in rainfall affected the simulated streamflow. Furthermore, the uncertainty of the simulated streamflow was influenced by the different runoff schemes used by the LSMs. For instance, MRI-SiB incorporates a direct infiltration structure from the surface to the second soil layer and estimates subsurface runoff using hydraulic diffusion and gravitational flow, while SiBUC uses a gravitational-only subsurface runoff approach. These variations led to significant disparities in surface and subsurface runoff. Future work should enhance the accuracy of rainfall from climate models and runoff from LSMs for assessing the potential impacts of climate change on water-related disasters.