Abstract

The overland flow runoff algorithm used in the USEPA SWMM model (Huber and Dickinson, 1988) has been a leading method for dynamic runoff simulation for over 30 years. The Runoff module in SWMM divides drainage catchments into two principal compartments, one each for impervious and pervious surfaces. Runoff discharges via a non-linear reservoir discharge equation to a drainage inlet, from where it can be routed through a collection system or downstream drainage subcatchments (Figure 18.1). SWMM44H, developed in 2002, allowed routing onto another drainage subcatchment (Huber, 2001). SWMM5, released in 2004, introduced new parameters that improve representation of typical urban runoff. The new parameters partition directly-connected impervious area (DCIA) and non-directly-connected impervious area (NDCIA) within a single catchment (Figure 18.2).

The traditional representation of an urban watershed can be quite effective in environments where impervious area dominates and runoff from pervious area is of minor importance. However, in other settings, this approach suffers from its failure to directly represent impervious areas that drain onto pervious areas, such as the cases that roofs drain onto lawns, or high impervious lands drain onto the low impact development areas or other BMP infrastructures. Explicit representation of DCIA and NDCIA better represents the physical configuration of a catchment and enables more direct parameterization of physical measurements such as soil infiltration rates and imperviousness. Those parameters were traditionally first estimated and then adjusted to represent DCIA and NDCIA in the model. For example, many modelers first estimate imperviousness of a catchment and then scale that fraction downwards to represent DCIA. Some modelers scale estimated soil infiltration rates downward by the ratio of the pervious area to the sum of the pervious area and the NDCIA. When NDCIA is explicitly represented, the model’s parameter for soils can instead be used to model infiltration rates that correspond with the physical soil parameters, and total imperviousness becomes a directly specified parameter, which eliminates the intermediate steps for these input data adjustments.

In SWMM5, NDCIA can be controlled by the Subarea Routing and Percentage Routed parameters (Figure 18.2). Subarea Routing directs surface flow from pervious land to impervious land or vice versa. The Percentage Routed parameter controls how much flow is transferred between compartments. To use internal routing in a catchment, the Subarea Routing parameter should be changed to either PERVIOUS or IMPERVIOUS from the default value OUTLET. The Percentage Routed value then controls the transfer fraction. The PERVIOUS value directs SWMM to route run-on from impervious lands to pervious lands and the IMPERVIOUS value directs SWMM to route run-on from pervious lands to impervious lands.

DHI’s MOUSE software and Wallingford Software’s InfoWorks, which are widely used for applications comparable to those modeled in SWMM, have alternative model paradigms that are more complex than the SWMM algorithms. Runoff Model B in MOUSE divides each model catchment into five compartments (versus SWMM’s two), but does not route flow from one compartment to the other. The New UK Runoff Model in InfoWorks does route flow from impervious onto pervious surfaces. However, the variable percentage runoff model employed in InfoWorks possesses its own limitations (Allitt, WaPUG Spring Meeting, 2002).

To evaluate the influence and merits of the DCIA/NDCIA parameters on model simulations, experiments of application of DCIA/NDCIA to hypothetical catchments and to a real study area near Philadelphia, Pennsylvania were conducted.

This paper is only available in PDF Format:

  View full text PDF