Estimation of green herbaceous phytomass from Landsat MSS data in Yellowstone National Park
MetadataShow full item record
CitationMerrill, E. H., Bramble-Brodahl, M. K., Marrs, R. W., & Boyce, M. S. (1996). Estimation of green herbaceous phytomass from Landsat MSS data in Yellowstone National Park. Journal of Range Management, 46(2), 151-157.
PublisherSociety for Range Management
JournalJournal of Range Management
AbstractGreen herbaceous phytomass was measured in August 1987 in grassland and sagebrush-grassland communities of Yellowstone National Park and related to August 1987 Landsat MSS data. A linear model using MSS band 7 and the ratio of MSS bands 6 to 4 accounted for 63% of the variance in green herbaceous phytomass on ground-truth plots (n = 25). Error in estimates of green herbaceous phytomass was influenced by the relative amount of bare ground and the proportion of green to green plus dead herbaceous vegetation present at a site. The model was used to predict average green herbaceous phytomass in grassland and sagebrush-grassland communities across a 600 km2 portion of ungulate summer range in Yellowstone National Park for 11 years during 1972-1987 using additional Landsat MSS imagery. Green herbaceous phytomass declined seasonally from late July to early September. Annual deviations in green herbaceous phytomass from the 11-year average, corrected for date of satellite overpass, were not significantly related to precipitation or temperatures during the growing season but were related quadratically to December-March precipitation. Below-average green herbaceous phytomass in years of low and high winter precipitation may be related to the effects of snow accumulation and melt on phenological development (green wave) of plants across the summer range. Models based on MSS spectral data can provide useful descriptions of broadscale patterns of plant biomass in Yellowstone National Park but may not suffice when precise estimates are required. Climatic influences on plant phenology may confound the interpretation of results when spectral models are used to compare vegetation yield of forage availability among years.