Minimizing impacts of urbanization on surface runoff: Development and application of optimization techniques

Zhenxu Tang, Purdue University

Abstract

The overall goal of this research was to develop technologies that can be used in efforts to minimize the impact of urbanization on surface runoff. The goal was pursued in two steps. First, the runoff and non-point source pollution impacts due to urbanization were identified and assessed using the Long-Term Hydrologic Impact Analysis model, L-THIA, and secondly, runoff impact minimums were identified by developing and applying optimization techniques integrated with the L-THIA model. To quantify the environmental impact of urbanization on runoff and non-point source pollution, L-THIA was employed in a watershed in Michigan. The outcomes indicated urbanization will slightly or considerably increase runoff volume depending on the development rate, slightly increase nutrient losses in runoff, but significantly increase losses of oil and grease and certain heavy metals in runoff. The web version of the L-THIA model was enhanced by developing a runoff optimization component and a solution heuristic algorithm, which allow selection of land use change placements that minimize runoff increase. The enhanced L-THIA model was utilized to evaluate runoff impacts of historical and projected urbanization for two watersheds in Indiana and Michigan. Optimizing land use change placement would have reduced runoff increase by as much as 4.9% from 1973 to 1997 in the Indiana study watershed. For non-sprawl and sprawl scenarios in the Michigan watershed from 1978 to 2040, optimizing land use change placement would have reduced runoff increase by 12.3% and 20.5%, respectively. This research also proposed a land allocation approach, by integrating the GIS-based L-THIA model with a spatial optimization model, to identify locations for urbanization with minimum runoff impact while satisfying area and shape requirements. The land allocation approach was applied to a watershed and showed that without shape constraints, the runoff increase from the regions identified by the land allocation approach range from 0.5% to 4.6%, depending on the magnitude of the decision weight factor settings. In comparison, the pattern of development projected by a land use transformation model would result in a runoff increase of 4.5%, which are 9 times more than development in the optimal locations with minimum runoff. The land allocation approach proposed in this study has significant potential implications for urban planning in developing and identifying land use plans that reduce the impacts of urbanization on runoff while reducing sprawl.

Degree

Ph.D.

Advisors

Engel, Purdue University.

Subject Area

Agricultural engineering|Environmental engineering|Civil engineering

COinS