WATER IN THE WEST: UTILIZING LANDSAT IMAGERY, GIS, AND STATISTICAL MODELS FOR MAPPING CRITICAL WETLANDS IN NORTHERN COLORADO’S CACHE LA POUDRE WATERSHED
This study utilized Landsat imagery, ancillary GIS layers, and boosted regression trees modeling for predicting riparian and palustrine wetlands across the Cache la Poudre watershed. Because wetlands and riparian flora vary along an elevation gradient, the study also tested the effectiveness of modeling the watershed within three elevation zones. This proved successful, generating model accuracies of up to 91.3% and Kappa statistics of up to 0.8261. The models successfully differentiated between open water and palustrine areas, and also identified many previously unmapped wetlands, including those in alpine zones. These results will be field-validated and serve as a baseline dataset for local researchers and land managers. Furthermore, the study developed an efficient and reproducible methodology for modeling wetlands that can be applied to other watersheds across the Intermountain West.