Development of methods for modeling and evaluation of low impact development practices at the watershed scale
Abstract
Effective planning and management of water resources at the watershed scale using strategies such as low impact development (LID), requires an understanding of variations in runoff and baseflow processes. The scientific literature, however, provides limited quantitative information describing potential impacts of LID practices at the watershed scale. Further, little information is available for exploring the impacts of LID practice adoption on baseflow processes. Computational methods for baseflow estimation for modeling LID practices at the watershed scale were developed and evaluated with data from Indiana watersheds. The method consists of techniques to develop baseflow equations, determine baseflow threshold area, and estimate baseflow pollutant coefficients for individual land use types. Estimated baseflow in the study watersheds indicated that baseflow represents an important proportion (approximately 60%) of total streamflow, suggesting that a better understanding of variations in baseflow quantity and quality can help improve watershed planning and management strategies. Differentiation of baseflow pollutant coefficients for individual land uses revealed that the proposed method is a viable option to isolate constituent mean concentrations for land use types in a watershed. To evaluate potential impacts of LID adoption on runoff, baseflow, total flow, and pollutant loading, a framework was developed to represent, evaluate, and report the effectiveness of LID practices using the Long-Term Hydrologic Impact Assessment-Low Impact Development (L-THIA-LID) model. The LID practices were represented with modified curve number (CN) values in the L-THIA-LID model. The proposed methods (i.e., baseflow estimation and LID representation) were utilized to enhance the capabilities of the L-THIA-LID model, which was applied in two urbanized watersheds to show that various levels of LID adoption can be used as retrofitting technologies to reduce hydrology and water quality impacts of urban development. Finally, a numerical procedure to quantify uncertainty associated with the L-THIA-LID model output was developed. Most of total variance in runoff and baseflow estimates originated respectively from uncertainty in precipitation data and watershed drainage area. Uncertainty in total streamflow was largely due to uncertainty in baseflow estimates. Minimization of uncertainty sources of the L-THIA-LID model would likely result in more accurate model predictions as an easy-to-use decision support tool.
Degree
Ph.D.
Advisors
Engel, Purdue University.
Subject Area
Water Resources Management|Environmental engineering|Urban planning
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.