Evaluating seasonal variability as an aid to cover-type mapping from Landsat Thematic Mapper data in the Northeast

Classification of forest cover types in the Northeast is a difficult task. The complexity and variability in species composition makes various cover types arduous to define and identify. This project employed recent advances in spatial and spectral properties of satellite data, and the speed and power of computers to evaluate seasonal variability as an aid to cover-type mapping from Landsat Thematic Mapper (TM) data in New Hampshire. Data from May (bud break). September (leaf on), and October (senescene) were used to explore whether different leaf phenology would improve our ability to generate forest-cover-type maps. The study area covers three counties in the southeastern corner of New Hampshire. A modified supervised/unsupervised approach was used to classify the cover types. A detailed accuracy assessment was performed to evaluate the classification. The results indicate that specific northeast hardwood species can be identified and that time of the year can significantly affect the cover-type classification accuracy