Evaluation of landsat TM generated forest-type map for tax appraisal purposes

Abstract

A tax appraisal system of private timberland based on the productivity value of the timber relies on current, accurate information about the land being assessed. Historically, such information has been supplied by aerial photos and field surveys. These methods are time- consuming, expensive, often out-of-date, and must be converted to digital data in order to be utilized in a Geographic Information System (GIS). This study hypothesized that satellite-based products and technology would generate a forest-type map that would provide accurate acreage estimates of forest types for tax appraisal purposes. Accuracy of over 91 percent was achieved for a forest-type map generated from leaf-off Landsat TM imagery relative to aerial photo reference data. Forest-type data obtained through traditional methods were compared to TM data of various minimum mapping units and determined to be less precise than the satellite forest-type data. Elimination of the pine-hardwood mixed forest type for individual tracts and making each tract have percentage estimates of the pine and hardwood components may be a more precise evaluation of land cover due to the finer resolution of satellite imagery.