As watersheds are urbanized, their surfaces are made less pervious and more channelized, which reduces infiltration and speeds up the removal of excess runoff. Traditional storm water management seeks to remove runoff as quickly as possible, gathering excess runoff in detention basins for peak reduction where necessary. In contrast, more recently developed “low impact” alternatives manage rainfall where it falls, through a combination of enhancing infiltration properties of pervious areas and rerouting impervious runoff across pervious areas to allow an opportunity for infiltration. In this paper, we investigate the potential for reducing the hydrologic impacts of urbanization by using infiltration based, low impact storm water management. We describe a group of preliminary experiments using relatively simple engineering tools to compare three basic scenarios of development: an undeveloped landscape; a fully developed landscape using traditional, high impact storm water management; and a fully developed landscape using infiltration based, low impact design. Based on these experiments, it appears that by manipulating the layout of urbanized landscapes, it is possible to reduce impacts on hydrology relative to traditional, fully connected storm water systems. However, the amount of reduction in impact is sensitive to both rainfall event size and soil texture, with greatest reductions being possible for small, relatively frequent rainfall events and more pervious soil textures. Thus, low impact techniques appear to provide a valuable tool for reducing runoff for the events that see the greatest relative increases from urbanization: those generated by the small, relatively frequent rainfall events that are small enough to produce little or no runoff from pervious surfaces, but produce runoff from impervious areas. However, it is clear that there still needs to be measures in place for flood management for larger, more intense, and relatively rarer storm events, which are capable of producing significant runoff even for undeveloped basins.

Abstract. Although the effects of spatial and temporal variability of precipitation on hydrologic modeling results have been well established, there has been very little attention given to sampling impacts on model calibration. To examine the effects of temporal sampling of rainfall, a combination of numerical simulations and calibration experiments have been carried out. A "true" runoff scenario, which is used for calibration of models using longer sampling periods, is established using very high resolution precipitation data and parameters derived from the physical properties of the soils. Three point infiltration models were chosen to isolate the effects of temporal sampling. A simple empirical model, a more complex empirical model, and a physics-based model were tested and compared. Numerical simulations using parameters based directly on soil properties demonstrated an undersimulation of runoff as the precipitation signal is smoothed at lower sampling frequencies. Calibration of selected parameters may compensate for this undersimulation effect. This impact becomes more pronounced for higher infiltration capacities, as calibration is required to compensate for greater underestimation of runoff. The adjustments that are necessary to compensate for the sampling effect illustrate how some of the physical meaning of the parameters is lost in the calibration process. Since the calibrated model parameters correspond to physical properties of the system, these results demonstrate that it may not be possible to make a priori parameter estimates based solely on physical properties of the system or to use parameters calibrated using data with different temporal or spatial sampling.

Removal of a NuisanceWith increasing population density and the replacement of native vegetation with other ground covers, surface water begins to play a much more substantial role in local hydrology, creating a need for "stormwater management." Early development relied on moving water away from a site as quickly as possible to limit nuisance and flooding potential as increasing volumes of stormwater were generated (Strassler et al. 1999). This approach was feasible prior to 1945, as development generally occurred on a lot-by-lot basis on small parcels of land. Even in the residential boom after World War II, when large tracts of land were completely subdivided, the natural sites would be stripped and replaced by a hydraulically efficient design as a logical extension of the previous approach (Urban Land Institute et al. 1975).To achieve the primary goal of limiting nuisance and flooding risk, the creation of a highly efficient conveyance system was needed. "Every feature of a conventionally developed site is carefully planned to quickly convey runoff to a centrally located management device, usually at the end of a pipe system. Roadways, roofs, gutters, downspouts, driveways, curbs, pipes, drainage swales, parking, and grading are all typically designed to dispose of the runoff in a rapid fashion" (Prince George's County 1999a, pp. 1-4 to 1-5). A key element of this conveyance system was development of stormwater sewerage, either as separate stormwater systems or combined sewer systems, which direct stormwater into the existing sanitary sewer system (Strassler et al. 1999).However, as development progressed, it became clear that this kind of practice could not continue indefinitely. "The cumulative effects of such approaches have been a major cause of increased frequency of downstream flooding, often accompanied by diminishing groundwater supplies, as a direct result of urbanization; or have necessitated development of massive downstream engineering works to prevent flood damage" (Urban Land Institute et al. 1975, p. 7). As a result, new strategies were clearly needed to manage stormwater. Slowing Down Storm FlowsAs development became more rapid, it quickly became obvious that the cumulative impact of removing stormwater from a large number of sites in a

No abstract