Weighting Landsat Digital Data According to Land Cover Emissivity for Surface Temperature Mapping

Publication Year:
Usage 109
Downloads 69
Abstract Views 40
Repository URL:
Polanski, Thomas
thesis / dissertation description
Regional urban planning and natural resource management problems require efficient and accurate data concerning land use/land cover and temperature gradients for informed decision making. Remotely-sensed data provide a method for acquiring such information in a dependable and efficient manner. Regular data acquisition and a synoptic view make the Landsat Thematic Mapper (TM) an excellent resource for entities needing land cover and surface temperature information. Landsat 5 TM digital data (1985) are used to classify land cover in the vicinity of New Orleans, the study area encompassing approximately 185 square kilometers. The maximum likelihood, minimum distance to means, and the parallelepiped classifiers produce land cover classified images with highly significant differences and the maximum likelihood rule outperforms the other methods. The maximum likelihood land cover classification is used as ancillary data for the surface temperature conversions and meets the standard of 85% thematic accuracy set by the United States Geological Survey (USGS). The Landsat 5 TM thermal channel (band 6) provides exceptional spatial resolution and is an excellent tool for mapping surface temperatures. Variable emissivities of land cover types and atmospheric conditions often need to be incorporated into surface temperature calculations from TM data. The thermal channel digital counts are weighted according to land cover emissivity and converted into kinetic temperatures (atmospheric conditions are deemed negligible for the TM data) . Statistics generated and qualitative analyses demonstrate a strong relationship between surface temperatures and land cover types, allowing for the prediction of the surface temperature change that a change in land use/land cover will incur. Applications of the research include modeling and monitoring of land use/land cover in a region, urban planning, urban heat island mapping, and natural resource management/conservation.