Impervious Cover Variability in Urban Watersheds

Celina Bochis and Robert Pitt


Data from about 160 neighborhoods in six Jefferson County, AL drainage areas were intensively investigated to determine the surface covers for each land use type. The data shows that the watersheds are highly impervious, with three of them having more than 50% of the watershed area composed of impervious cover. However, TR-55 (USDA 1986) guidance still shows that the impervious cover for all land uses investigated to be much greater than we observed for our area. It was also concluded that the variabilities of the surface covers within the different land uses for the investigated areas was small, especially for the impervious covers.

The percentage of directly connected impervious cover (DCIA) was determined by direct field observations and this data was also compared to commonly used empirical equations. Equations for determining DCIA developed during prior studies were fitted with our observed DCIA for our data, but they did not result in good relationships, especially when analyzed at the land use level. There was a good comparison between Sutherland’s “highly connected basins” equation and the fitted equation for our overall data. However, the residual analysis for the regression model was poor, suggesting that the power equation may not be applicable to our data in our region. Suitable equations were developed for each land use separately, showing that a single equation could not be accurately used to estimate DCIA for all regions and land uses.

A Pearson correlation matrix was created to show the magnitude of the relationships between the field measurements of the major land covers and the calculated stormwater concentrations for those drainage areas. The matrix illustrated that runoff volume can be mostly predicted by using DCIA, TIA, and parking lot areas. Also, there were high correlations between all of the runoff constituents and the particulate solids concentrations. It is expected that the variability of the land covers (especially the DCIA) would have a similar effect on the variabilities of the runoff volumes.

Permanent link: