Spatial-temporal-thematic assimilation of Landsat-based and archived historical information for measuring urbanization processes

Abstract

Measuring urbanization and assessing its impacts require long-term records of land changes, which cannot typically be provided from a single information source. A prerequisite to the creation of a reasonably consistent national information database on urban growth was the development of a robust methodology to assimilate land information from diverse sources. This method was applied to assimilation of two information sources, the Canadian Urban Land Use Survey (CUrLUS) and the Canada Land Use Monitoring Program (CLUMP). CUrLUS consists of a suite of contemporary thematic maps derived from satellite images while CLUMP information was extracted through conventional visual interpretation of aerial photography. In the process of generating integrated temporal series, the compatibility between the two information sets was assessed. The application of the assimilation methodology has led to generation of reasonably consistent urban land-cover and land-use change information for major Canadian urbanized areas spanning a 35-year period.