Resource information
Accurate identification of urban land use is essential for many applications. However, as physical surfaces of land-cover types are not necessarily related to their use and economic function, differentiating among thematically detailed urban land uses (single-family residential, multi-family residential, commercial, industrial, etc.) using remotely sensed imagery is a challenging task, particularly over large areas. Few thematically detailed urban land-use mapping products have successfully been created via automated means at regional scales, and no such product currently exists for the entire United States. This study demonstrates an approach that uses US census block groups (BGs) as the basic unit of geography and predicts the percentages of 10 urban land uses for each BG. Predictors are based on a number of national-scale and publicly available data sources, including census data, point locations of land-use features, and historical land cover. Results indicate the importance of accounting for contextual/proximity effects when identifying land uses. The method, demonstrated over a four-county area surrounding the city of Boston, provides a potential model for broad-scale urban land-use mapping.