Reeves, MC, Zhao, MS, Running, SW (2006). Applying improved estimates of MODIS productivity to characterize grassland vegetation dynamics. RANGELAND ECOLOGY & MANAGEMENT, 59(1), 1-10.
The efficacy of Moderate Resolution Imaging Spectroradiometer (MODIS)-derived vegetation productivity was tested in the grasslands of western North Dakota in terms of its ability to characterize fluctuations in aboveground green biomass and provide regional perspectives of interannual vegetation dynamics. To accomplish this task, we scaled plot-level observations of grassland biomass throughout the growing seasons of 2001 and 2002 using high-resolution satellite imagery and meteorological observations. Regionally scaled biomass measurements were compared with MODIS net photosynthesis (PSNnet) estimates at 3 times during the growing seasons of 2001 and 2002. The relationships between MODIS PSNnet estimates and scaled aboveground green biomass improved steadily during the progression of each growing season, and reached a maximum (r(2) = 0.77 and 0.57 in 2001 and 2002, respectively) near peak greenness. Aboveground green biomass was more tightly coupled in 2001 because of the relative abundance of green biomass compared with standing dead from the previous year. We characterized interannual variability in grassland vegetation through analysis of MODIS-derived net primary productivity(NPP) for the years 2001-2003. MODIS NPP estimates showed a significant decline (P <= 0.05) from 2001 to 2003, partly induced by a significant decline (P <= 0.05) in growing-season precipitation. The results indicate the reliability of productivity estimates for monitoring grassland biomass fluctuations is improved during years where plant growth conditions are more favorable. Additionally, MODIS data may be more useful for addressing administrative, rather than managerial, needs given the coarse resolution and regional perspective of the vegetation products.