Analyzing Low Impact Development Strategies Using Continuous Fully Distributed Coupled Groundwater and Surface Water Models
Abstract
In recent years, Low-Impact Development (LID) strategies have received more attention due to the requirement to reduce the adverse hydrologic and water quality effects of urbanization. The obvious benefits of reduced peak flows and smaller storm water ponds can be simulated in models such as SWMM; however, the analysis of impacts to the groundwater system is challenging. Addressing questions related to local infiltration capacity, feedback from the groundwater system (i.e., rejected recharge and saturation-excess runoff), and ecological benefits including preservation of baseflow and wetland hydroperiod requires a spatially distributed and integrated analysis of the flow system. The purpose of this paper is to illustrate the challenges and insights that Earthfx has encountered simulating and comparing the effectiveness and ecological benefits of different LIDS scenarios using an integrated groundwater and surface water model and how this analysis can be combined with SWMM-based hydrological analyses.
Identifying the optimum type and placement of certain LID strategies requires consideration of local terrain, existing natural heritage features, soil and subsurface geology, and depth to water table. Working in a collaborative manner with storm water engineers, Earthfx has addressed these site-specific features and land use specific LID design proposals with the USGS integrated GSFLOW model. GSFLOW is unique in that the surface water and overland runoff processes can be simulated on a finer distributed grid resolution than the groundwater system. The groundwater component of GSFLOW is the widely accepted MODFLOW model, which also allows variable cell size grids to better represent key subsurface features. This ability to independently refine components of the distributed model, and yet simulate the processes and LID options in an integrated manner, is essential for design scenario analysis and comparison.
Examples presented in this paper show how an existing multi-aquifer MODFLOW model was upgraded to GSFLOW in order to assess various LID designs. Comparisons between the simulated pre- and post-development conditions helped quantify the direct impacts of urbanization on groundwater levels, groundwater discharge to streams (at the reach scale), aquatic systems, and habitat. Overland flow routing was used to quantify changes to feature-based water balances by assessing relative contributions of surface water and groundwater entering wetlands and riparian areas.
Next, the model was applied to predict the effectiveness of LIDs in mitigating the predicted impacts on the surface water and groundwater regimes. Using a detailed distribution of land use types, multiple LID scenarios were applied to fit developer needs and land use function by: (a) increasing evaporative loss and reducing runoff volumes through mechanisms such as green roofs, bio-swales, increased soil depth, increased vegetation density; (b) increasing groundwater recharge through permeable/pervious/porous surfaces or by capturing runoff and routing it to infiltration galleries constructed under impervious surfaces; and © combining (a) and (b) in the form of leaky detention and/or retention ponds and roof runoff captured though downspout disconnects. The models enabled a direct, distributed comparison of the alternatives and multiple runs were conducted to optimize the effectiveness of the various designs. For example, the simulation of LID measures helped to reduce groundwater drawdowns by 86%, regain groundwater discharge to streams by 42%, and reduce the increased runoff generated by 80% from development without LID implementation. For each scenario, subbasin-scale predictions of runoff where then re-integrated into the storm water/erosion model analysis and design.
In summary, the GSFLOW analysis complements storm water modelling and thus provides for an integrated LID analysis. It will provide SWMM users with the necessary information to assess distributed LID effectiveness from multiple design scenarios. The work demonstrates that scenario comparison is complex, and that site specific and integrated LID design analysis is helps to achieve the benefits of this important new water management approach.
This paper is only available in PDF Format:
View full text PDF