The U.S. Department of Agriculture's National Resources Conservation Service (NRCS) runoff curve number method is the most widely used method in the U.S. for predicting stormwater runoff in a watershed, with the runoff curve number (CN) being a key factor in estimated the hydrologic loss.  The traditional method of calculating the composite CN is time consuming and tedious, requiring areally weighting multiple land uses and soil types within a subbasin.  Now, however, the GIS formatted soil and land use data that are available more and more readily from the public domain make it possible to automate calculation of the CN. 

Methods for Calculating Composite CN

Determination of a composite CN depends on the soil hydrologic characteristics and ground cover conditions within the watershed.  The factors determining CN include the hydrologic soil group (HSG), cover type, treatment, hydrologic condition, and soil antecedent moisture condition (AMC).  The HSG is one of four types-A, B, C or D-based on the infiltration rate of the soil.  Group A represents sands and gravels, which demonstrate the highest infiltration rate, or the lowest CN; whereas Group D characterizes silts and clays, which have the lowest infiltration rate, or highest CN.

Runoff curve numbers under the different ground cover and soil features are summarized in Table 1.  This simplified CN look-up table is based on information obtained from NRCS staff and my own experience.  A more detailed list is published in Technical Report 55 (NRCS, 1986). 

When the watershed is composed of several CNs, its representative CN, or the composite CN, can be described as:

CN = 98 * Imp% + ΣA_i CN_i * (1-Imp%)/ ΣA_i                                                       Eq. 1

Where:
             A_i is the area associated with CN_i,
             Imp % is the soil's imperviousness percentage in the area, A_i
             CN = 98 is used for the connected impervious area, where runoff              eventually flows into the drainage system (NCRS 1986).

From Equation 1 it can be seen that the calculation of composite CN includes the summation of each intersection of every HSG and land use polygon within the watershed; however, this calculation process can be done much more efficiently by using GIS data instead.

Retrieving GIS Data 

The procedures for retrieving the GIS data for use with Equation 1 are as follows:

1.    Drainage Area Delineation.  Automation of drainage area       delineation has been used widely.  Many hydrologic programs, such as ArcHydro, HEC-GeoHMS, Pre-pro2000, WMS and RiverCAD, have this capability. 

Instead of the delineation of the watershed itself, the availability of the 3-D data is key to automating the calculation of the composite CN.  Currently, the most readily available 3-D data is the U.S.Geological Survey (USGS) Quadrangle Topographic Map.  Although its accuracy is limited to 10 feet (3 m) vertical and 100 feet (30 m) horizontal, it should be sufficient for the basic requirements for square-mile-sized watersheds.  (The equation works for areas of 1 to 100 square miles.)  

2.   Soil Data Processing.  Originally, soil data was obtained from the U.S. Department of Agriculture NRCS County Soil Survey Book.  Later, maps could be digitized into a GIS platform. Now, with the Soil Survey Geographic (SSURGO) Database available on-line [NRCS, 2006], GIS-formatted soil survey maps can be used directly as the soil data base. Using Brazos County in Texas as an illustration, the attributes of the SSURGO data are shown in Figure 1. Unfortunately, Figure 1 does not list the HSGs for the watershed.  However, the soil name and map symbol (MUSYM in SSURGO data from the county soil survey map (Figure 2) can be related to HSG.

The next step is to assign a new column as HSG using the above-mentioned classifications.  In the case of Brazos County, I then wrote a Visual Basic program to sort MUSYM and generate the HSG column (see raw land uses A, B, C and D in Table 2 on the following page).  This could also be done on the ARC GIS platform.

3.    Land Use Map Processing.  The GIS formatted land use or zoning maps are becoming more and more readily available in the public domain.  Most of the time, these maps can be obtained from the local government agencies.  They might not be up to date; however, so could need to be updated based on surveys or the most current aerial ortho-photograph.

Figure 3 shows an example of land use map attributes for Brazos County.  The attribute, SPTBCODE (State Property Tax Board) lists each lot's land use information.  Each government entity had its own definitions on land use categories, so these attributes needed to be re-classified to match the definitions in Table 1.

4.    Intersecting Three Base Maps.  After obtaining the three fundamental maps/ shape files-drainage basin, land use and soil data-Overlay in Arc Toolbox is used to  intersect the files as a new map/shape file that contains each individual polygon with its unique HSG and land use features.  Figure 4 shows the final intersected map for the Brazos County project.       

Figure 5 shows the list of attributes of this intersect map where the key elements are BasinID, LUCODE (SPTBCODE), MUSYM and Area.

Using the GIS Data to Calculate the Composite CN

The final step in calculating the composite CN is based on data in Figure 5.  With this and Equation 1, data queries can be performed in the ArcGIS platform.  In this study, another method is introduced in which the pivot table from a spreadsheet application is used.  In an MS Excel spreadsheet using Pivot Table and Pivot Chart Wizard:

• Soil data is assigned to attribute or column field
• Land use data is assigned a feature or raw field
• Polygon area is assigned a data field
• Each subbasin is assigned to a page field.

In this way, automatic sorting can be performed in the pivot table wizard and shown in Table 2 (top).  Based on Equation 1, the calculation portion is illustrated in the bottom of Table 2.

The result, as shown in Table 2, is the final composite CN.

Final Notes

The method described in this article can quickly and efficiently calculate the composite CN.  For instance, a task taking several weeks using the conventional approach can be completed within a day.  This approach will also lower the possibility of human error during data management.  Therefore, it is believed that this approach will increase productivity and can also significantly improve the accuracy of calculations.