A coupled Bayesian and fault tree methodology to assess future groundwater conditions in light of climate change

Abstract. Maintaining acceptable groundwater levels, particularly in arid areas, while protecting ecosystems, are key measures against desertification. Due to complicated hydrological processes and their inherent uncertainties, investigations of groundwater recharge conditions are challenging, particularly in arid areas under climate changing conditions. To assist planning to protect against desertification, a fault tree methodology, in conjunction with fuzzy logic and Bayesian data mining, are applied to Minqin Oasis, a highly vulnerable regime in northern China. A set of risk factors is employed within the fault tree framework, with fuzzy logic translating qualitative risk data into probabilities. Bayesian data mining is used to quantify the contribution of each risk factor to the final aggregated risk. The implications of both historical and future climate trends are employed for temperature, precipitation and potential evapotranspiration (PET) to assess water table changes under various future scenarios. The findings indicate that water table levels will continue to drop at the rate of 0.6 m yr⁻¹ in the future when climatic effects alone are considered, if agricultural and industrial production capacity remain at 2004 levels.