Satellite data-driven multi-objective simulation-optimization modeling for water-environment-agriculture nexus in an arid endorheic lake basin

Abstract

In the arid endorheic lake basin with highly developed agriculture, water scarcity drives the nexus of water-environment-agriculture (WEA) while potential evapotranspiration (ETp) is a critical element linking basin-scale irrigation water demand and agro-economic development. The study proposes a satellite remote sensing data-driven multi-objective simulation-optimization (SO) framework to implement the basin-scale WEA nexus management. The hydrological model (MODFLOW-NWT) combined with the MODIS (Moderate Resolution Imaging Spectroradiometer) based ETp estimation with a 500-m spatial resolution simulates the monthly surface water (SW) and groundwater (GW) irrigation practices. Then, the multi-objective evolutionary algorithm (ε-MOMA) is applied to solve management model which includes maximization of agro-economic benefit and environmental flow and minimization of environmental risk of total nitrogen loading subject to irrigation water demand, land availability and food supply. The SO framework is validated in an arid endorheic lake basin with intensive agricultural irrigation in northwestern China. The four-dimensional Pareto-optimal solutions are achieved to optimize crop planting structure and irrigation water allocation in order to avoid cognitive myopia triggered by low-dimensional or single objective optimization. The results highlight the trade-off among the high value crop planting, SW-GW irrigation and environmental flows, and indicating that lowering nitrogen loading is more sensitive to agro-economic benefit. Historical maximum and minimum ETp scenarios are tailored to quantify the impact of ETp change on WEA nexus management. Results show the increasing ETp has a great negative effect on the environmental flow objectives. Therefore, how multi-sectoral interests change depending on the basin-wide development goals may be difficult to regulate due to complex WEA nexus and associated deep uncertainty in the future.