EarthRef.org Reference Database (ERR)
Development and Maintenance by the EarthRef.org Database Team

Detailed Reference Information
Caselles et al. 1997
Caselles, V., Coll, C., Valor, E. and Rubio, E. (1997). Thermal band selection for the PRISM instrument: 2. Analysis and comparison of existing atmospheric and emissivity correction methods for land surface temperature recovery. Journal of Geophysical Research 102: doi: 10.1029/97JD01446. issn: 0148-0227.

This is the second paper in a series which discusses the thermal band selection for the Processes Research by Imaging Space Mission (PRISM) planned by the European Space Agency (ESA). In part 1, algorithms for emissivity-temperature decoupling were analyzed with the aim of identifying optimum methods and attainable accuracies for extracting emissivity values in the PRISM thermal bands. In this paper, we have focused on studying the existing methods for land surface temperature (LST) recovery. The preliminary configuration of the PRISM instrument includes four thermal channels: A (~3.7 μm,) B (~8.8 μm), C (~10.8 μm), and D (~11.9 μm) and along-track multipointing capability. Single-channel, split-window, and dual-angle algorithms have been tested and compared using a database of simulated brightness temperatures. Although focused on PRISM, the study is relevant to present instruments with similar features, such as the advanced very high resolution radiometer (AVHRR) and the along-track scanning radiometer (ATSR). In particular, we have shown that (1) atmospheric effects dominate errors in single-channel LST recovery methods, (2) spectral emissivity dominates errors in the dual-angle and split-window methods, and (3) current AVHRR split-window channel combination (channels C and D) provide the best results. It does not need auxiliary atmospheric information and therefore it can work in an operational scheme using only satellite data. It can be applied to any type of surface, and accuracies around ¿1 K can be obtained for most natural surfaces. For the single-channel method the use of coincident radiosoundings is necessary, because of the high sensitivity to the atmospheric variability. Dual-angle methods are constrained to homogeneous surfaces, since in heterogeneous, rough surfaces, the temperatures measured at the two observation angles may present directional effects caused by the surface structure and the differences between vegetation and background temperatures. We have also shown that the utility of channel B to LST recovery is marginal. Combinations including channel A are only applicable in nighttime conditions owing to reflected radiation. For these reasons the split-window algorithm, using channels within the 10-12.5 μm window, has been suggested for operational determination of LST.¿ 1997 American Geophysical Union

BACKGROUND DATA FILES

Abstract

Keywords
Meteorology and Atmospheric Dynamics, Remote sensing
Journal
Journal of Geophysical Research
http://www.agu.org/journals/jb/
Publisher
American Geophysical Union
2000 Florida Avenue N.W.
Washington, D.C. 20009-1277
USA
1-202-462-6900
1-202-328-0566
service@agu.org
Click to clear formClick to return to previous pageClick to submit