Projection of future precipitation change using <scp>CMIP6</scp> multimodel ensemble based on fusion of multiple machine learning algorithms: A case in Hanjiang River Basin, China

Abstract Projecting precipitation changes is essential for researchers to understand climate change impacts on hydrological cycle. This study projected future over the Hanjiang River Basin (HRB) based multimodel ensemble (ME) of six global models from Phase 6 Coupled Model Intercomparison Project (CMIP6). An ME method using fusion four machine learning (ML) algorithms (random forest [RF], K‐nearest neighbors [KNN], extra tree [ET], and gradient boosting decision [GBDT]) was proposed in this study. The were investigated during 2023–2042 (Near‐term), 2043–2062 (Mid‐term), 2081–2100 (Long‐term) periods, with reference base period 1995–2014, under three integrated scenarios (SSP1‐2.6, SSP2‐4.5, SSP5‐8.5) Shared Socioeconomic Pathways (SSPs) representative concentration pathways (RCPs). results show that: (1) performs better than mean individual ML algorithms, a correlation coefficient value reaching 0.88 Taylor skill score 0.764. (2) SSP5‐8.5 has largest upward trend annual variation range −9.27% 112.84% 2023 2100, followed by SSP2‐4.5 −30.48% 44.67%, smallest SSP1‐2.6 −37.19% 37.78%, which significant humidification HRB future. (3) are increase time, Long‐term, Mid‐term, Near‐term. (4) northeastern parts experience large future, southeastern smaller. (5) Uncertainties still exist, can be reduced ME. findings obtained have important implications policymakers make adaptive strategies reduce risks change.