Abstract

Soil water dynamics play a crucial role in agricultural productivity, ecosystem functioning, and hydrological processes. Accurate prediction of soil moisture patterns under future climate scenarios is essential for sustainable water resource management and agricultural planning. This study reviews current approaches for modelling soil water dynamics using downscaled climate predictions, examining the integration of global climate models (GCMs) with regional soil water models. We analyze various downscaling techniques including statistical and dynamical methods, their applications in soil water modelling, and associated uncertainties. The review synthesizes findings from recent studies demonstrating that ensemble approaches combining multiple downscaled climate projections with physically-based soil water models provide the most robust predictions. Key challenges include handling precipitation extremes, representing soil heterogeneity, and quantifying uncertainty cascades from climate models to local soil water predictions. Future research priorities include improving sub-daily precipitation downscaling, incorporating soil-vegetation-atmosphere feedbacks, and developing probabilistic prediction frameworks for decision support systems.