Rangeland monitoring and prediction at range to regional scales

TitleRangeland monitoring and prediction at range to regional scales
Publication TypeConference Paper
Year of Publication2004
AuthorsMoran M.S., Thoma D., Hernandez M., Heilman P., Stone J.J., Starks P.J., Arnold J.G., Kiniry J.R., Richardson C.W., Rango A., Herrick JE, Peters DC, Havstad K
Conference Name57th Annual Meeting, Society for Range Management
Date PublishedJanuary 25, 2004
Conference LocationSalt Lake City, UT
ARIS Log Number163889
AbstractThe USFS, BLM and NRCS have responsibility to inventory and monitor rangelands in connection with their land use planning and management duties. Despite these efforts, a recent National Research Council study concluded that rangeland monitoring is inconsistent, inadequate, infrequent and does 'not provide the data needed to support national assessments of rangeland health.' In preliminary studies, ARS scientists have shown that a combination of ground-based inventory with ARS models and available satellite images provides accurate and consistent information. To assimilate these three methods - conventional ground-based inventory, remote sensing and simulation modeling ' a multi-stage research thrust is necessary. First steps would be to refine current inventory and modeling approaches to accept image data and to improve extraction of range-specific information from satellite images. Second, an error assessment mechanism should be developed to determine information accuracy for any combination of the three methods. Third, all 1-, 2- and 3-method combinations should be tested for monitoring and predicting rangeland information at ARS experimental watersheds and rangelands in key locations in the U.S. Finally, a decision tool should be developed to assist users in selecting the most appropriate approach for their location. At all stages, an adaptive research approach must be used to improve all methods in response to error assessment and user feedback. The ultimate goal of this initiative would be to develop the best possible methods for rangeland monitoring and prediction. This synergistic approach will produce better information for a variety of rangeland types than could be obtained by any one method alone.