Surface Carbon Dioxide Fluxes Predictable from Remotely Sensed Vegetation Indices for Wyoming Landscapes

Robert D. Kelly,

Dept. of Atmospheric Science, University of Wyoming

The Wyoming King Air research aircraft was used to measure sensible heat, latent heat, and CO₂ fluxes and a remotely-sensed vegetation index during February - September 1999 over four different ecosystems in southeastern Wyoming. Three of the experiment sites (mixed conifer forest, mixed short-grass prairie, and sagebrush shrubland) had native plant cover, while the fourth was a mixture of pasture and irrigated/non-irrigated cropland. The CO₂ fluxes at each site followed seasonal trends, with the peak, summer fluxes occurring later over the forest than over the other systems. In spite of differences in species composition, plant architecture, above-ground biomass, and other characteristics, plots of CO₂ flux vs. vegetation index over the grassland, shrubland, and forest show nearly identical linear regression slopes. Both annual and seasonal changes in vegetation index accurately predicted corresponding shifts in net surface CO₂ flux for these quite different vegetative landscapes. The regression equations from the 1999 data also accurately predicted CO₂ fluxes over the same sites in 2000, even though the period of peak growth in 2000 ended earlier due to lower precipitation. Because these ecosystems are representative of large fractions of the total North American and global land cover, the CO₂ flux vs. vegetation index relationships should be very useful in satellite-based monitoring of surface processes at regional or even global scales.