SATELLITE REMOTE-SENSING OF SURFACE-ENERGY AND MASS BALANCE - RESULTS FROM FIFE

In FIFE a diverse, interdisciplinary group of scientists are cooperating to understand how the earth's land-surface vegetation and atmospheric boundary layer interact to affect weather and climate. The focus of this research is the interception of solar and long-wave radiation by a vegetated surface, and the subsequent thermodynamic and biological control of sensible and latent heat release. In addition, the science teams are investigating the use of satellite remote sensing to monitor the components of the surface energy balance at point, regional and global scales. Field experiments were conducted in the summer of 1987 and again in the summer of 1989. Analyses of the 1987 data are reasonably mature, whereas the 1989 data analyses are in the early stages. In this paper we summarize the results of the analyses to date. The analyses have shown that the hypotheses linking energy balance components to surface biology and remote sensing are reasonable at a point level, and that satellite remote sensing can potentially provide useful estimates of the surface energy budget. GOES data have been used to estimate solar insolation at the earth's surface to an accuracy of about 21.6 W m-2 and about 8.2 W m-2 for photosynthetically active radiation. The ratio of near-infrared to red reflectance has also shown to be linearly related to measured CO2 flux. Radiometric temperatures estimated from remote sensing are highly correlated to the canopy aerodynamic temperatures; however, the remote sensing estimation of H to useful accuracies (+/- 100 W m-2) appears to be a more difficult problem than the accurate estimation of LE. The investigation of atmospheric scattering and absorption effects on satellite remote sensing of surface radiance shows that the magnitude of atmospheric opacity variations within the FIFE site and with season can have a large effect on satellite measured values of surface radiances. However, comparisons of atmospherically corrected TM radiances with surface measured radiances agreed to within about 2% in the visible and near-infrared wavelengths and to 6% in the mid-infrared. Looking at the frequency of cloud-free (< 10%) acquisitions actually achieved during FIFE shows that for land-surface climatology, multiple polar orbiters may be required to achieve the minimum desired acquisition frequency: once each 5 days.