Monitoring and Predicting Urban Land Use Change: Applications of Multi-Resolution Multi-Temporal Satellite Data

The ability to map and monitor the spatial extent of the built environment, and associated temporal changes, has important societal and economic relevance. Multitemporal satellite data now provide the potential for mapping and monitoring urban land use change, but require the development of accurate and repeatable techniques that can be extended to a broad range of conditions and environments. We have developed an approach using Landsat imagery, trained with the high resolution data sets, that identifies impervious surface areas (buildings, roads, etc) at subpixel resolution. We report on application of the approach over a range of scales, from the local to the entire Chesapeake Bay Watershed (168,000 km²). We also developed maps of past changes in the built environment, used them to calibrate a spatial predictive model, and generated maps of expected future change under various policy scenarios out to year 2030. We believe these techniques have applicability to a wide range of applications.