Developed and vacant land (Essex Fells) derived from TIGER and LANDSAT TM: Map

Comparison map (Essex Fells) of developed and vacant land derived from TIGER and LANDSAT TM; the determination of the amounts of developed land can vary when LANDSAT TM data and the TIGER data are combined, and/or when different housing unit density thresholds are used.Although the LANDSAT TM data and the TIGER data are highly independent of each other, each type of data can be used, either alone or in combination, to map and compute areas of vacant and developed land. The determination of the amounts of developed land can vary when different data types are combined, and/or when different housing unit density thresholds are used. The increase in the estimated amount of developed land, when different housing unit densities are used in combination with the satellite data, is shown for the four municipalities in this map (gaspric1.gif) and gaspric2.gif, gaspric3.gif and gaspric4.gif. As a lower housing unit density threshold is used in conjunction with the LANDSAT TM data, the estimated amount of developed land within a municipality increases. The amount of developed land as a percentage of total area in Essex Fells increases dramatically as housing unit density decreases. Essex Fells is a prime example of the difficulty encountered in detecting certain types of residential areas using LANDSAT TM data exclusively. This municipality consists of residential homes on lots with well maintained lawns and large trees. The well established tree vegetation or "suburban savanna" tends to obscure or mix with the spectral reflectivity of the impervious surfaces, such as streets and rooftops. During the classification process, the algorithm classifies areas such as Essex Fells as being most similar to forest cover or nonforest vegetation cover. This misclassification can be altered or eliminated using different housing unit density thresholds