A novel uncertainty analysis method to improve the accuracy of agricultural grey water footprint evaluation considering the influence of production conditions

Abstract

Agricultural grey water footprint (GWF) is an important index for quantitatively describing the impact of agricultural production on water environment, which can help evaluate regional water pollution risk (WPR). However, changes in agricultural production conditions lead to strong uncertainties in agricultural GWF. Therefore, advanced methods to deal with these uncertainties for improving the accuracy of regional WPR assessment are required. By combining min–max/max-max programming models with a GWF assessment framework, an integrated uncertainty analysis method for agricultural GWF affected by production conditions was proposed in this study. Based on the method, extreme scenarios of GWF evaluation were established to obtain lower and upper bounds of agricultural GWF. The proposed method was applied in the Guangdong Province in South China, and the following results were obtained. Agricultural GWFs in the Guangdong Province were [124.67, 237.37], [110.63, 204.90], and [104.04, 178.01] Gm3 in 2005, 2010, and 2015, respectively. Different contributing factors of agricultural GWF were shown in the minimum and maximum GWF scenarios. For crop planting, the contribution of dryland and paddy fields to GWF was larger than that of garden land in the minimum GWF scenario, whereas the contribution of garden land was greatest in the maximum GWF scenario. For livestock and poultry breeding, the contribution of livestock raised under dry cleaning condition to the GWF was larger than that under flushing condition in the minimum GWF scenario, whereas the opposite situation was observed in the maximum GWF scenario. Based on the interval values of agricultural GWF, the WPR of the cities of Zhanjiang, Maoming, Yunfu, Chaozhou, Foshan, and Dongguan were identified as unacceptable. By identifying the uncertainties within agricultural GWF evaluation induced by the variability of production conditions, the proposed method can provide helpful information for agricultural planning in the context of sustainable agricultural development.