Estimating the presence of paved surface parking lots in the conterminous U.S. from land use coefficients for 1974, 1982, 1992, 2002, and 2012

Parking lots may be a significant source of pollution. Oil, sediments, and heavy metals may accumulate on their surface, then be flushed into rivers, streams, and lakes via rainfall. At present no dataset provides a mapping or estimation of parking lot area or locations nationwide. This product consists of a time series of five national 60-meter raster datasets which estimate the proportion of each pixel represented by parking lots, based on land-use coefficients. The rasters span the conterminous United States, for the years 1974, 1982, 1992, 2002, and 2012. The dataset was derived by calculating coefficients for 18 land-use types (Commercial, Industrial, Residential, Recreation, and so on) from the 2012 U.S. Geological Survey NAWQA Wall-to-wall Anthropogenic land-use Trends (NWALT) product. The coefficients were calculated by comparing NWALT land-uses to 1-meter rasters representing detailed paved surface parking lot polygons available from six cities: Bloomington, IN; Chattanooga, TN; Denver, CO; Hartford, CT; Raleigh, NC; and St. Paul, MN. The land-use classification overlying the largest amount of parking lot land areas was Commercial land (20.1% of land area), followed by Industrial land (19.6%), and Major Transportation (7.4%). The results were cross-validated against ground-truth data withheld from the calculations. The coefficients derived from the year 2012 data were then applied to the prior four years to create a time series. The rasters provide a way to estimate percent parking lot area by watershed or other area of interest, over the last four decades.