Assessing the trade-off between shallow groundwater conservation and crop production under limited exploitation in a well-irrigated plain of the Haihe River basin using the SWAT model

Abstract

Groundwater overdraft due to extensive irrigation has led to a shallow aquifer depletion crisis in the Haihe River basin. Quantitative simulation of the variations in shallow groundwater and crop yield under different conditions of limited exploitation using a distributed hydrological model is important in the well-irrigated plain of this basin. Based on multiple modeling experiments by a modified Soil and Water Assessment Tool (SWAT) model, three limited irrigation schemes for winter wheat (Triticum aestivum L.) were selected as simulation scenarios. The simulated results using the SWAT model showed that under scenario 1 (applying two rounds of irrigation corresponding to the jointing and heading stages of winter wheat), the average rate of decline in the shallow groundwater table was approximately 2/3 of that under the basic scenario (current irrigation schedule), but the winter wheat yield decreased by 13% compared with the basic scenario. Under scenario 2 (applying one round of irrigation at the jointing stage of winter wheat), the average rate of decline in the shallow groundwater table was approximately 1/4 of that under the basic scenario, but the reduction in the winter wheat yield compared with the basic scenario increased to 28%. The amount of overexploited shallow groundwater in the cropland area decreased from 17.5 × 108 m3 a−1 (under the basic scenario) to 11.0 × 108 m3 a−1 under scenario 1 and 4.5 × 108 m3 a−1 under scenario 2. Under scenario 3 (rain-fed conditions during the winter wheat season), the regional variation in the shallow groundwater table shifted to a recovery trend with an average rate of 0.22 m a−1, which was equivalent to restoring 3.5 × 108 m3 a−1 of shallow aquifer storage in the cropland area. However, the reduction in the winter wheat yield compared with the basic scenario reached 54% under the rain-fed scheme. Considering the trade-off between groundwater conservation and crop production under limited exploitation, linear programming was used to optimize the irrigation schedule at the subbasin scale. As a result, to satisfy the constraint of stopping groundwater drawdown, the average minimal reduction in the winter wheat yield would be 42% under the optimal irrigation schedule. By contrast, to satisfy the constraint of restricting the reduction in the crop yield to within the threshold required to maintain winter wheat self-sufficiency, the minimal rate of decline in the shallow groundwater table would be 0.26–0.52 m a−1 under the optimal irrigation schedule. In addition, the uncertainty in the simulated shallow groundwater variation in this study was acceptable, indicating that the above assessments could provide reasonable references for regional groundwater management.