Development of an inexact simulation-evaluation model for the joint analysis of water pricing and groundwater allocation policies.

A critical step in water management policy development is the analysis of its socio-economic and environmental implications. However, few methods could proactively and reliably predict and assess the impacts of policies while handling the inherent uncertainty. To fill such a gap, an inexact simulation-evaluation method was developed for analyzing the impacts of multiple water management policies under uncertainty. The interval positive mathematical programming (IPMP) method was proposed as the simulation tool by coupling interval programming with positive mathematical programming (PMP). The evaluation tool was developed by combining the interval TOPSIS method and the interval maximum deviation method. This simulation-evaluation method can directly communicate a policy's simulation outcomes into the evaluation process while addressing the uncertainties in both simulation and evaluation. The proposed method can also reproduce the actual situation with a calibration process, which enables accurate and smooth responses to policy changes. This approach was used for agricultural water management in arid north-west China. Seventy-five policy alternatives generated from three groundwater allocation limits and twenty-five differential water pricing levels were investigated. The impacts of these alternatives on farmer income, farmer employment, water consumption, planting areas, and fertilizer use were simulated using IPMP. Twenty-four non-inferior alternatives were selected and further evaluated with multi-dimensional criteria. The final results showed that, the water price for grain crops with traditional irrigation methods should rise by 60%, those for cash crops with drip irrigation should decrease by 60%, and the groundwater quota should be reduced by 20%. Compared with traditional models, IPMP can increase simulation accuracy by reproducing observed situations, enhance robustness by reflecting input uncertainty, and improve flexibility in decision-making by providing interval solutions. The inexact simulation-evaluation model can also be widely used to analyze other policies.