Traditional land cover data sets such as the National Land Cover Database (NLCD) provide about fifteen land cover classes for Texas at a spatial resolution of 30 m. Ecological systems maps leverage LandFire environmental models and the NLCD to map ecological systems and improve thematic resolution, but workers lacked time to make all of the improvements desired. To enhance outcomes, a team of more than a dozen workers have developed methods to improve both the number of types mapped (thematic resolution) and the spatial resolution of results. We used three-date mosaics of satellite imagery together with environmental data to classify land cover. Classification was improved by a coordinated effort to collect ground data and provide successive refinements. Ancillary data including digital county soils, which were extensively manipulated, and variables derived from digital elevation models were used to define abiotic site types. Image objects generated from NAIP were used to improve spatial resolution.
We improved the number of mapped types in a given area by more than 10X versus the NLCD, and improved the spatial resolution of maps by 9X. Improvements include mapping of communities such as flatwoods, live oak woodlands, dense and sparse shrubland types, and a variety of steep slope, bottomland, and disturbance types. We have also developed information on the general composition of each mapped type, and present results in GIS accompanied by easy to understand interpretive materials. As a result, landscape scale conservation strategies and decisions by individual land managers can be made with better information on the extent and spatial distribution of communities. More than half of Texas is mapped to date. Plans are underway to see how these methods may apply to work in the Northeast region.