Land-Cover Classification and Change Detection

Various multispectral data sets and tools are available for mapping land cover and land-cover change for inputs to hydrologic applications. The best choice is often a function of the specific application. New sensors offer increased capability in mapping spatial detail and temporal variation of land-cover categories. Algorithms have advanced in robustness and include distribution-free methods that are superior to traditional approaches in terms of modeling the multispectral distributions of reference data. Advanced subpixel methods of mapping land cover offer greater thematic coherency and the possibility of using consecutive land-cover characterizations to map change. The success of deriving a meaningful result, such as deriving a relationship between aquatic biotic indices and land-cover change, relies on a high-quality land-cover reference map. By choosing the most appropriate data sources, constructing a defensible land-cover definition set, and employing robust algorithms, analysts allow for the meaningful incorporation of land-cover map information into hydrological studies.