Fuzzy optimization model for integrated management of total nitrogen loads from distributed point and nonpoint sources in watershed

Abstract

A fuzzy optimization model is developed to allocate allowable total nitrogen (T-N) loads to distributed nonpoint sources (NPSs) and point sources (PSs) in a watershed for river water quality management using the linear programing technique. The watershed is divided into uniform grid cells on which T-N loads issuing from NPSs such as paddy fields, upland crop fields and cities are controlled. A geographic information system integrated with the digital elevation model facilitates computation of route lengths of surface and subsurface flows from cells to a river running through the watershed. The T-N loads discharged from their sources are assumed to decay, subject to distance-related first-order kinetics. As management goals, maximizations of total allowable NPS loads, total allowable PS loads and total yield of rice are considered from environmental and economic viewpoints. A prime constraint is an effluent limitation standard for the aggregate amount of loads that arrive at the downstream end of the river. The fuzzy sets theory helps appropriately describe vague attitudes of decision-makers (i.e., stakeholders and management authorities) in terms of constraints and conflicting goals. An application of the fuzzy optimization model, developed as an improvement over our last nonfuzzy model, to a real watershed in Shiga prefecture, Japan, demonstrates that the fuzzy model embodies our last model, and is capable of creating management alternatives for T-N load allocation in a more practical and flexible manner.