Urban Growth, DRECP

This raster dataset is the result of a subtraction of the classification of raster datasets bhd2030us and bhd2000us. Both are produced using the SERGoM v3 model, they depict housing density for the coterminous US in 2030 and 2000, resp., and are based on 2000 US Census Bureau block (SF1) datasets. (The classication of the subtraction is shown below.) A raster layer that portrays not-developed (and assumed to be undevelopable) called CUS_UPP_100 depicts protected/unprotected lands and Census water polygons. First, Land was removed from blocks (reducing the area of a block, but not its number of housing units -- assuming that private housing units must be on private land) that was public and/or protected in the UNPROTPRIV100 dataset produced by Forest Inventory and Analysis, North Central Research Station, USDA Forest Service -- which is based on the second edition (2001) of the Protected Areas Database, produced by the Conservation Biology Institute. Second, locations of blocks that are identified as water polygons (rivers, lakes, reservoirs, marshes -- basically unbuildable areas) in the US Census Bureau TIGER files were also removed from the blocks. Third, blocks that had no housing units in them AND were dominated (>25% of their area) by developed land use were also not allowed to have any housing density in subsequent years (by removing those cells from the developable layer). The value of 22, 23, and 24 from the USGSs NLCD 2001 was used, then aggregated by a factor of 4 (from 30 m to 120 m cells). Locations where >25% comm/ind/trans were found to create NLCD_COMMIND dataset. Classification of Housing Density Value Acres per Unit Name Definition 0-10 scale 0 Private, no units Private, no units No development risk 0 1 >40 Rural Low development risk 3 2 10 - 40 Exurban Moderate development risk 5 3 1.7-10 Suburban High development risk 8 4 <1.7 Urban Very high development risk 10