This Land Use and Land Cover classification was produced using the expert system model that was originally developed for the set of Landsat images acquired in May 24 and June 18, 1998 (Stefanov et al. 2001). The model performs a posteriori sorting of classes initially derived using Maximum Likelihood Supervised Classification by integrating with geographically co-registered auxiliary data layers such as land-use maps, variance texture computed for the 3X3 or 5X5 pixel neighborhood, water rights database, city boundaries, and Native American reservation boundaries. These additional GIS layers are updated accordingly to reflect changes in land use. Our final classification consists of 12 classes and has a reported overall accuracy of 83 %. 1057 randomly generated reference points stratified by land cover classes were used in accurayUser’s accuracy for individual classes varied from 71% to 100% with the exception of commercial/industrial class (51%). Overall Kappa statistic is 0.8127. Erdas Imagine software was used to run all analyses.