Synthetic Satellite Data for Rangeland Management

Land managers often lack vegetation data of the proper spatial and temporal resolution and extent needed to assess range conditions and monitor trends in the arid and semiarid western United States.  We combined Landsat TM and Terra MODIS imagery to develop time series of predicted surface reflectance and NDVI products of higher spatiotemporal resolution than possible with either platform alone.  The quality of the synthetic satellite data was influenced by the length of time between the start and prediction dates and the type and timing of management activities.  NDVI predictions, stratified by land cover and land use, were accurate through the growing season; however, inter-class performance varied.  Our findings demonstrate a possible alternative to traditional field-based methods when it comes to the assessment and monitoring of the phytomass (living organic plant matter, specifically the aboveground portion in this case) component of semiarid rangelands.