Open Collections

Abstract

Effective mine water management requires realistic precipitation modelling to assess hydrological risks and support sustainable water use. Stochastic weather generators, particularly Markov chain-gamma models, are widely used to simulate daily precipitation with seasonal patterns. However, traditional implementations often fail to capture long-term variabilities, such as multi-year droughts, prolonged wet periods, or climate trends—key factors in mine water balance modelling and environmental impact assessments. This paper introduces an enhanced stochastic precipitation model that builds upon the Markov chain- gamma framework by incorporating a random walk component to dynamically adjust input parameters— wet/dry transition probabilities (PWW, PWD) and precipitation intensity parameters (α, β). Unlike conventional methods that modify model outputs to match historical trends, this approach ensures that low- frequency climate variability emerges naturally from the model structure rather than being imposed post hoc. By continuously modulating precipitation drivers, the model better represents long-term fluctuations while maintaining daily and seasonal precipitation characteristics. By integrating a dynamic representation of long-term precipitation variability, this model provides a more robust foundation for mine water studies, risk assessments, and environmental planning. The capability to simulate potential future conditions, including climate-driven changes in precipitation, enhances decision-making for sustainable water management.