P8 Runoff Parameters Consistent with SLAMM Simulations

Impervious Areas

P8 and the Source Area Loading & Management Model (SLAMM, PV & Assoc., 2007, http://www.winslamm.com) are used widely in Wisconsin for designing urban stormwater BMP's.   SLAMM's runoff algorithm and input requirements are more complex than P8's.  SLAMM permits detailed specification of individual source area types (roofs, streets, etc). This allows for evaluation of BMP's targeting individual source areas.  SLAMM has been calibrated to extensive runoff data from Wisconsin watersheds and tested elsewhere (R. Pitt, Small Storm Hydrology, SLAMM Version 9.2.5 documentation). P8's runoff algorithm is simpler and requires specification of two fundamental parameters driving simulation of runoff from impervious areas (depression storage & runoff coefficient), based upon professional judgment with general guidance from the literature. 

P8 watershed parameters yielding predictions of impervious-area runoff from Wisconsin watersheds similar to SLAMM's are described below.  Since each model has its unique features and purposes, it is not implied that the parameters allow P8 to be used in place of SLAMM.  Rather, the values provide indirect calibrations to Wisconsin runoff data upon which SLAMM is based.  It also provides reasonable consistency of P8 and SLAMM results for users applying both models simultaneously to specific design projects.  SLAMM Version 9.2.5 has been used to develop the calibrations. The analysis is based on simulations of May 15 - October 31, 1981 rainfall in Madison.  A previous study by the USGS (Corsi, 1996) determined this to be representative of an average year. The applicability of the calibrations outside of Wisconsin has not been evaluated.

The calibrations apply specifically to runoff from directly-connected impervious areas, which typically drives BMP designs. 

Walker (2007) shows P8 calibrations (DS & R1) to 7 impervious-area categories considered in SLAMM.  The calibrated runoff coefficients (R1) apply while the cumulative rainfall/snowmelt during a given event is less than 0.8 inches ("breakpoint" specified on the Edit General screen).  The runoff coefficient is constrained to 1.0 above 0.8 inches (see runoff model description). Calibrations are based upon runoff coefficient vs. cumulative rainfall tables for Wisconsin watersheds specified in the SLAMM input file "WI_SL06 DEC06.rsv". 

The table below lists results for SLAMM impervious area types and standard land use categories.  Each category has a specific distribution of impervious area types, as listed in the table.  There are essentially three ways to simulate a SLAMM standard land use category in P8:
 

1. Specify a separate P8 watershed for each impervious area type.  Use the calibrated depression storage and runoff coefficient values listed below for each impervious area type.  Add another subwatershed to reflect the pervious and indirectly connected area. Link each watershed to a single P8 Pipe device, the outflow from which would reflect the combined runoff from the entire watershed.
2. A simpler method uses a single P8 watershed with depression storage and runoff coefficient based upon the weighted average across individual impervious area types, as listed in the table below.  This method is approximate because a single value for depression storage would not reflect the distribution of values across impervious area types.  The directly-connected impervious fraction for each category can be derived from the table below or adjusted to site-specific conditions.  Runoff predictions using this method are compared in the figure below. 
3. An even simpler method uses a single P8 watershed with parameters for impervious runoff (DS =0.022, R1=0.8). Agreement between the models is similar to or better than that based upon weighted-average coefficients (Method 2 above).  As far as P8 is concerned, runoff from impervious areas in typical urban watersheds is primarily driven by the overall impervious fraction and the distribution of impervious area types is of secondary importance.  That distribution is more important in SLAMM, which allows evaluation of BMP's targeting specific source areas.   Runoff predictions using this method are compared in the figure below.

Calibration Results:

Predicted runoff from directly connected impervious surfaces for each standard land use and model, using impervious fractions and weighted-average P8 parameters (depression storage and runoff coefficient) listed in the above table :

Predicted runoff from directly connected impervious surfaces for each standard land use and model, using impervious fractions listed in the above table and constant P8 parameters (DS = 0.022 inches, R1=0.8):

Pervious Areas

The figure below compares pervious area runoff volumes predicted by P8 and SLAMM.  P8 results are shown as a function of Curve Number with and without seasonal factors (snowfall/snowmelt, frozen soil, not considered in SLAMM).  SLAMM results are shown for three soil textures and the "Undeveloped" source area category.

These results are provided for general information.  P8 and SLAMM use fundamentally different algorithms for pervious area runoff, so differences in the results could be due to several factors.