Land-use optimization for controlling peak flow discharge and nonpoint source water pollution

摘要

In this paper we present and apply a modeling methodology to reduce stormwater runoff through land-use planning, in order to mitigate the impact of nonpoint source pollution. A storm-water-runoff simulation model is used to generate peak discharge pseudodata, that are inputted into a regression analysis, in which the functional relationship between peak discharge and land-use variables is approximated as a quadratic function (R~2 = 0.98). This function is then used in an optimization model to allocate future land uses (urban, conservation, and agriculture) at the sub-watershed level, in a way that minimizes the resulting peak discharge at the watershed outlet. A discharge reduction of 44% is achieved, compared with the discharge resulting from the present (actual) land-use pattern. The most downstream and upstream areas are to be protected with more conservation, and most urban activities are allocated to the minimal impact subwatershed. Sensitivity analyses are performed and suggest maintaining at least 25% of the land in a conservation state, developing no more than 12% for urban purposes, and no more than 70% for agricultural purposes.