An inexact simulation-based stochastic optimization method for identifying effluent trading strategies of agricultural nonpoint sources

摘要

Agricultural nonpoint sources (NPS) pollution has long been regarded as the leading challenge in maintaining water quality of receiving water bodies. Effluent trading can serve as a cost-effective way to achieve optimal configuration of discharge permits in agricultural NPS pollution control. However, great difficulties exist in practical effluent trading planning, including uncertainties related to randomness and imprecision, system risk of nutrient loadings being unacceptably high, and factors with spatiotemporal heterogeneity within the watershed. In this study, an inexact simulation-based stochastic optimization method (ISSOM) is developed for identifying effluent trading strategies in response to the above challenges. With the aid of technique of interval analysis, uncertain parameters related to nutrient yields can be handled and dynamic variation of NPS contaminant loadings can be reasonably addressed. Besides, ISSOM can tackle uncertainties expressed as fuzzy, stochastic and interval formats and capture the notion of risk under high-variability situation in NPS pollution control. The ISSOM is applied to a real case of agricultural NPS pollution mitigation through effluent trading in Xiangxihe Watershed. Results show that the uncertainties play a major role in successfully launching an effluent trading program, and trading scheme can mitigate agricultural NPS pollution with an increased system benefit. Results also reveal that the agricultural zones of Xiakou and Gufu are the major pollution sources and main purchasers in effluent trading, and Guizhou zone is the main vendor which contributes least to nutrient discharge. These findings can not only facilitate identification of the main pollution sources and optimal effluent trading schemes, but also gain insight into the tradeoff among the agricultural benefit, system risk, and satisfaction degree. (C) 2015 Elsevier B.V. All rights reserved.