HIGH-RESOLUTION DISTRIBUTED HYDROLOGIC MODELING ACROSS A GRADIENT OF DEVELOPMENT TYPES

We are using three-dimensional distributed groundwater surface water modeling to quantify how patterns and ages of urban development affect the hydrologic cycle. We have conducted distributed watershed modeling simulations with the ParFlow.CLM model in small urban watersheds (0.2-0.6 sq km in area) in the Gwynns Falls watershed in Baltimore County, Maryland, across a spectrum of development ages and types. Horizontal model resolution for each domain is 10m by 10m, with a 1m vertical discretization. We modeled the development footprint in each subbasin, setting the hydrologic properties of the surface layer and the land cover parameters for the coupled CLM land surface model according to high-resolution land cover data. Impervious cover fraction varies from 22% to 70% among the selected domains. Kevsway subwatershed is a highly developed end member (70% impervious), with a mix of residential and commercial types. Runnymeade subwatershed has suburban residential development, more recent stormwater management, and development has been restricted from the valley bottom. In each model, the distribution and extent of developed area and impervious surface influence the location of recharge and discharge from groundwater storage. Model water table and saturation results show the influence of artificial topographic lows, such as detention basins and ditches. For example, in the Kevsway subwatershed model, the stream headwaters form in a constructed topographic low associated with a large stormwater feature. Steeper topography in the Runnymeade subwatershed model, some of which is constructed, result in a large vegetated bottomland area that intersects the modeled water table elevation. This results in shallower average depth to water and a greater saturated surface extent. In both models, recharge and withdrawal from storage are spatially focused by land cover and development footprint. On the watershed scale, unsaturated zone storage for Kevsway is less sensitive to changes in climatic forcing than Runnymeade, with these differences on the order of the interseasonal variability.