Comparison of Landsat, Resourcesat-1, and modis to classify and determine potato area in southern Idaho.

摘要

Landscape classification is one method of exploiting satellite data for crop area estimation over large areas that would otherwise take months or years to accomplish. Organizations use this, information to make marketing, planting, or harvesting decisions.; Recent developments have made the Landsat data continuity unstable. One objective of this study is to compare other satellite sensors that might function as replacements for Landsat data should it become unavailable.; Three sensors acquired imagery for the 2005 growing season for Bingham County in southern Idaho. We evaluated Landsat 5, Resourcesat---1 advanced wide field sensor, and the moderate resolution imaging spectroradiometer for their ability to make an accurate classification and area estimation. We found that Landsat and the advanced wide field sensor were comparable at 90% or better user's accuracy and the moderate resolution imaging spectroradiometer was around 55% user's accuracy.