Remote Sensing Science
Biospheric environmental monitoring at BOREAS with AVHRR observations
Article first published online: 21 SEP 2012
Copyright 1997 by the American Geophysical Union.
Journal of Geophysical Research: Atmospheres (1984–2012)
Volume 102, Issue D24, pages 29651–29662, 26 December 1997
How to Cite
1997), Biospheric environmental monitoring at BOREAS with AVHRR observations, J. Geophys. Res., 102(D24), 29651–29662, doi:10.1029/97JD01327., , , , , and (
- Issue published online: 21 SEP 2012
- Article first published online: 21 SEP 2012
- Manuscript Accepted: 2 MAY 1997
- Manuscript Received: 18 JUN 1996
Global models of biospheric processes such as production efficiency models need environmental input and validation data sets at high temporal and spatial resolution. Methods developed to assess biospheric environmental conditions from advanced very high resolution radiometer (AVHRR) observations, specifically air temperature and surface moisture, are explored through exploitation of field measurements collected at the Boreal Ecosystem-Atmosphere Study site. The surface temperature/spectral vegetation index (TVX) concept provides the potential for estimating air temperature and surface moisture from AVHRR-type satellite observations. Initial results show that the slope of TVX over boreal landscapes is related to near-surface soil moisture in addition to vegetation type and solar irradiance. Also, the TVX air temperature estimation correlates well with shelter height observations. However, our analysis shows that challenges remain in using the TVX approach, largely because it is difficult to isolate the effects of variations in atmospheric conditions from physical environmental conditions at the Earth's surface. Clouds significantly limit the application of the TVX technique. It is also limited when the variability of the normalized difference vegetation index is low or a sharp vegetation boundary is contained within the TVX contextual array. The TVX air temperature can be used to approximate shelter height temperature to within ±5 K under most conditions; however, there are outliers with differences as large as 15 K which, at this time, cannot be explained. The air temperature estimation also exhibits a 3.2 K warm bias. Part of this bias is a result of errors in the split window surface temperature estimate used in the TVX regression.