Extreme precipitation over the Xijiang River Basin (XJRB) is closely associated with drought–flood hazards in South China and is therefore central to basin-scale climate risk assessment. This study quantifies historical model skill and characterizes spatiotemporal changes in precipitation extremes in the XJRB using a framework jointly constraining spatial pattern fidelity and interannual variability. Daily precipitation from 30 quality-controlled stations (1961–2014) was used to evaluate nine CMIP6 global climate models following statistical downscaling and bias correction based on quantile mapping with a hybrid distribution function. Eleven ETCCDI extreme precipitation indices were derived, and model performance was assessed using the Taylor Skill Score (TSS) for spatial patterns and the Interannual Variability Skill Score (IVS) for year-to-year variability; these metrics were integrated through a Combined Ranking Index (CRI) to identify key models and construct a selected-model multi-model ensemble (MME). The selected-model MME was used to project extreme precipitation for three future periods (2021–2040, 2041–2070, and 2071–2100) under four Shared Socioeconomic Pathway scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5). NorESM2-LM, NorESM2-MM, and MRI-ESM2-0 exhibit the most balanced spatiotemporal performance and constitute the selected-model MME. Projections indicate a systematic intensification of precipitation extremes toward the late 21st century, with stronger increases in intensity-related indices (R95p, R99p, RX1day, RX5day, and SDII) over the southeastern XJRB, alongside increasing consecutive dry days (CDD) in the northwestern basin, particularly under higher-emission scenarios. Frequency-related indices display pronounced scenario dependence, with increases under SSP1-2.6 and SSP2-4.5 but weakened or locally negative responses under SSP3-7.0 and SSP5-8.5. These results reveal an enhanced spatial contrast between extreme wet and dry conditions across the basin and delineate the scenario-dependent evolution of precipitation extremes in a monsoon-dominated river system.
Wang et al. (Wed,) studied this question.