Uncertainty of climate projections and an approach utilizing climate model outputs for hydrologic computation in the Ba River Basin

Abstract

A top-down approach begins with Global Climate Models (GCMs) is a common method for assessing climate change impacts on water resources in river basins. To overcome the coarse resolution of GCMs, dynamic downscaling by regional climate models (RCMs) with bias-correction procedures is utilized with the aim to reflect the meteorological features at the river basin scale. However, the results still entail large uncertainties. This paper examines the ability to capture the observed baseline temperature and precipitation (1986-2005) in the Ba River Basin from GCM outputs, RCM outputs, bias-corrected GCM outputs and bias-corrected RCM outputs by analyzing statistical indicators between historical simulations and observed data in 4 temperature and 6 rainfall stations. Bias-corrected results of both GCM and RCM have significantly smaller errors compared to the unbias-corrected ones. The uncertainty of future climate projection for the mid and late 21^th century of the bias-corrected GCMs and RCMs are evaluated. It is found that there is still uncertainty in projected results. A concept of “Decision-Scaling” which combines top-down and bottom-up approaches is proposed to assess the climate change impacts on hydrological system to take into account uncertainties of climate projections by models.