This article investigates the methodological advancements and challenges in climate modelling, focusing on hydrological projections and their application in adaptive water governance frameworks. By exploring the evolution of the Coupled Model Intercomparison Project (CMIP), the paper highlights key advancements, including improved representations of biogeochemical cycles, regional variability, and advanced parameterizations. Specific emphasis is placed on microphysical processes that are critical to precipitation formation but remain major sources of uncertainty in hydrological projections. The article emphasizes the need to address these uncertainties through refined modelling methodologies and the incorporation of robust observational data. Experimental studies and advanced tools, including numerical simulations and machine learning emulators, are essential for linking fine-scale dynamics with global models and enhancing predictions of hydrological feedbacks. The synthesis highlights practical applications, emphasizing how refined models and a precise understanding of uncertainties support adaptive and sustainable water management, particularly in regions under acute stress. This interdisciplinary approach bridges technical precision and governance, providing actionable pathways for addressing global water challenges.