A distributed modeling approach for evaluating hydrological effects of rain gardens in urban watersheds

Best management practices (BMPs) and low impact development (LID) are sustainable stormwater management practices used to mitigate the effects of urbanization such as excess runoff and water quality issues. Implementation of BMPs and LID have been limited and sometimes restricted because of the lack of recognized methodologies to estimate their hydrologic effects in urban watersheds under a continuous rainfall period. It is expected that rain gardens will have a significant effect in the reduction of peak discharge and volume for a range of different storms magnitudes including less frequent events. Rain gardens are small depressions covered by native vegetation, which receive the runoff coming from impervious areas. These practices are part of the sustainable LID and BMPs approach with the goal of reducing runoff coming from urban areas, promoting evapotranspiration and restoring some of the infiltration capability of the predevelopment site. These distributed stormwater management practices modifies the urban watershed???s hydrologic response by varying the size and quantity of these distributed stormwater practices. Hydrologic processes of BMPs can be complex and non linear. Uncertainty could arise when commonly simplified models are use to simulates the effects of BMPs on the hydrologic response of the watershed. This research used a methodology developed to understand the hydrologic effects of rain gardens at different quantities distributed in an urban watershed for a continuous rainfall period. The methodology used in this research tries to improve the estimation of hydrologic process of rain gardens by using a physically distributed model, Mike SHE. Mike SHE, distributed by DHI, Inc. is a fully distributed model that is able to estimate a range of hydrological processes occurring in a rain garden. This model provides an improvement over simplified models, which cannot estimates relevant hydrologic processes. The Mike SHE model simulates evapotranspiration, subsurface flow and overland flow by coupling a finite difference method in two dimensions and the Richard???s equation for the unsaturated zone calculations. As part of the methodology used in this research, two rain garden scenarios with different quantities of rain gardens simulated are implemented in an urban watershed. Data from rain garden sites monitored by the U.S. Geological Survey Wisconsin Water Science Center were used to build and calibrate single rain garden models. The calibrated rain gardens were incorporated to an urban watershed with an area of 13 acres and 86 houses. The urban watershed model was calibrated by using observed data monitored in the 1960s without rain gardens. Rain garden scenarios were simulated under a continuous rainfall period. Results from this research showed that simulated rain gardens are able to reduce the peak discharge and volume among different return periods. The reduction of peak discharge and volume increased when the quantity of rain gardens increased. The hydrologic effects of rain gardens decreased when the magnitude of the storm increased. The reduction of peak discharge and volume ranged from 5% to 80% depending on the magnitude of the storm. It was found that the antecedent moisture conditions of rain gardens affected their capacity for runoff retention. The results found in this research show that physically distributed models are able to estimate hydrologic effects of rain gardens inside urban watersheds. This modeling approach provides the flexibility to estimate hydrologic effects of different rain gardens layouts under continuous rainfall periods. This modeling approach could be used by engineers and planners to examine hydrologic effects in urban watershed for design purposes