Abstract
The drought-risk effects of climate change are estimated using 5-km regional climate ensemble experiments, which are dynamical downscaling datasets of “Database for Policy Decision-Making for Future Climate Change, d4PDF” having 20-km resolution large ensemble climate simulations. The Yodo River Basin study area has water resources managed using highly developed water utilization facilities.
For climate ensemble experiments, we applied a moving average method using a “moving-window” and a quantile-mapping bias-correction method applicable to large ensemble datasets. The storage volume of Lake Biwa, the largest lake in Japan, was evaluated to estimate drought risk in the Yodo River Basin. And the storage volume of other many water resource development facilities are evaluated too. For this research, water cycle processes such as snowfall, snowmelt, evaporation and water use in the Yodo River Basin are calculated using a distributed rainfall-runoff model. Furthermore, all amounts of demand set in the water resource development plan are supplied in calculations.
Frequency distributions of the annual minimum level of Lake Biwa in the present climate and future climate are estimated as a safe degree of water use.
Results revealed the following: 1) The lowest operating level of the Lake Biwa for water use (B.S.L.-1.5 m) is in the range of non-exceedance probability less than 0.1 in the present climate, and more than 0.5 in a future climate that is 4K warmer than the pre-industrial climate. 2) The probable 100-year drought annual minimum level is B.S.L.-3.04 m to -3.74 m in non-bias-corrected future experiments and B.S.L.-4.46 m to -5.92 m in bias-corrected future experiments.
