Search

A process for integrating remote sensing and spatial data analysis to accurately map and monitor agricultural crops and other land cover in the Lower Colorado River Basin is described. These maps were then used as input into a model that accounts for consumptive use throughout the basin. Water is an important and incredibly valuable resource in this area. International treaties and court decrees dictate water allocation to the states of Arizona, Nevada, and California, and to Mexico. Maps of the agricultural crops with a requ ired overall accuracy of 93 percent for use in the water model were generated from Landsat Thematic Mapper data four times per year. An automated signature extraction process and data exploration techniques were developed to aid in achieving these required accuracies. All maps were subjected toquantitative accuracy assessment, and error matrices were produced to evaluate overall and per-class accuracies. [ABSTRACT FROM AUTHOR]