Figure 1.
A flow chart depicting the process to map the residential land uses.
Locations mapped as high-density urban residential were mapped in the National Land Cover Dataset 2006 (NLCD) as a built-up cover type and had a census block housing density in 2010 (BHD) of at least 10 dwelling units per acre (dua). Low-density urban residential was mapped as within PLC, on land but not water cover types (NLCD water, barren, or wetlands), and had a housing density between 1.6 and 10 dua. Suburban residential was mapped similar to low-density urban but at a density between 0.4 and 1.6 dua. Low-density residential areas were required to be outside of cities/towns (and were defined using Census places; PLC), and exurban residential had a housing density between 0.1 and 0.4 dua, while rural residential had a housing density between 0.025 and 0.1 dua.
Figure 2.
A flow chart depicting the process to map commercial land uses.
Locations mapped as office commercial use were defined as an office building or office park by a census landmark polygon (LAND) or by North American Industry Classification System 2012 (NAICS) 2-digit code 51–56, which includes information, finance and insurance, real estate, professional, scientific and technical services, management of companies, and administrative and support services. Because commercial and residential uses are often mixed (e.g. particularly vertically, with commercial on the ground floor and residential above), and because the Longitudinal Employer-Household Dynamics (LEHD) data are mapped at the Census block-level, the residential housing density also had to be low (<0.1 dua). Consequently, areas that are mapped as urban residential land use could also contain some commercial land uses as well.
Figure 3.
A flow chart depicting the process to map institutional land uses.
Educational uses are mapped through Census landmark polygons (LAND) with values of D43 (Educational institution, including academy, school, college, and university) or if designated as a school in the Protected Areas Database US v1.3 (PAD). Educational institutions are further classified as developed (including buildings and playgrounds) using the National Land Cover Dataset 2006 (NLCD) with built-up values 22–24, or undeveloped (that includes grass fields and other vegetation types). Medical uses and government/public uses are mapped similarly, using LAND values = D31 (hospitals) or PAD designation as a hospital or other medical use; or D65 (government complex) or government designation from PAD. Military uses are depicted using LAND with values of D10 or as mapped in PAD or in the military polygons from the National Transportation Atlas Database 2010 (MILI). Developed portions of military lands, which can range from airports to storage facilities to residential uses, are differentiated using NLCD built-up values 22–24 (or if not, then are mapped as military undeveloped).
Table 1.
A multi-purpose, hierarchical land use classification system.
Table 2.
Datasets used to construct the national land use database.
Figure 4.
The National Land Use dataset for 2010, depicting 59 land use classes for the conterminous US, at a resolution of 30(0.09 hectares).
Table 3.
Area and proportion of the US for sub-group level of the National Land Use dataset 2010.
Figure 5.
Regional variation of the proportion of major land uses in the conterminous US.
The relative extent of 7 regions is depicted by the size of the pie-charts: Pacific (CA, OR, WA); Mountain (AZ, CO, ID, MT, NM, NV, UT, WY); North central (IA, KS, MN, MO, ND, NE, SD); South central (AR, LA, OK, TX), Great Lakes (IL, IN, MI, OH, WI), Northeast (PA and northeastern states), Southeast (DE, KY, WV and southern states). Land in productive uses dominated the major land uses for all regions but the Northeast, which was narrowly dominated by built-up uses. The proportion of land in recreation and conservation land uses exceeded 10% for the Pacific, Mountain, and Northeast regions. Note: “water” is not a land use per se, and includes both water and wetland cover types.
Table 4.
Results of the accuracy assessment testing the National Land Use dataset against a sample-based, detailed land use dataset.
Figure 6.
The National Land Use dataset for 2010 centered on Denver, Colorado (I-25 and I-70), showing NLUD at 4 different scales.
Detailed industrial (light grey), commercial (dark grey), and transportation (black) land uses can be readily scene in the detailed map (3 km scale), while regional land use patterns surrounding the Denver metro area, including “open space” lands composed of recreational, conservation, and production lands, are shown in the coarse map (24 km scale). The full legend is provided in Figure 4.
Table 5.
A comparison of the National Land Use database (2010) against the National Land Cover Dataset (2006) for built-up uses.
Figure 7.
The diversity of land uses shown for Fort Collins, CO (left) and Ashville, NC (right).
Higher diversity is depicted in yellow/red, with lower diversity shown in blue and dark blue. Major highways and interstates are labeled for reference.
Figure 8.
The population of US metropolitan areas follows a power law-scaling to the developed “footprint”, which includes built-up land uses except for urban parks and golf courses.
Computing the ratio of the raw population to the footprint (scale) adjusted population provides a relative measure of land use efficiency for metropolitan areas.