资源环境科技发展态势分析平台

Climate change significantly influences characteristics of rainfall events including rainfall depth, rainfall duration, inter-event time and temporal patterns that directly affect water resources management, flood defence and hydraulic structure design. In this study, a framework is proposed to analyse daily-scale rainfall event characteristics based on global climate model (GCM) simulations. This framework includes bias correction of raw GCM-simulated rainfall series, selection of good-performing bias-corrected GCMs based on the mean absolute percentage error (MAPE) and evaluation of selected GCMs' skills in simulating rainfall event characteristics and finally assessment of changes in rainfall event characteristics in the future. In this study, 17 GCMs, four representative concentration pathways (i.e., RCP2.6, RCP4.5, RCP6.0 and RCP8.5) and two future periods (i.e., 2041-2070 and 2071-2100) are considered. After bias correction of the GCMs using the monthly-scale double gamma distribution, 9 out of 17 GCMs with MAPE values smaller than 20% in the historical period 1971-2000 are selected. In general, these selected GCMs well capture the rainfall characteristics of different rainfall event classes. The multi-model ensembles suggest that compared to the historical period, the frequency of rainfall events with an extreme depth, short duration and long inter-event time will increase in the two future periods and the change in 2071-2100 is generally larger than that in 2041-2070, indicating that more extreme climate conditions may occur in Qu River basin in the future. Moreover, the temporal patterns of heavy rainfall events will become more non-uniform with more concentrated peak rainfall. The frequency of the delayed rainfall type (i.e., peaks occurring at the end of the rainfall event) will increase in the future, which can probably cause more severe floods and is very detrimental to flood defence in this study area.