Publications

Kato, Soushi; Matsunaga, Tsuneo; Tonooka, Hideyuki (2014). Statistical and in-situ validations of the ASTER spectral emissivity product at Railroad Valley, Nevada, USA. REMOTE SENSING OF ENVIRONMENT, 145, 81-92.

Abstract
This study investigated the accuracy of the surface emissivity generated as an Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) standard product applying the Temperature Emissivity Separation (TES) algorithm. We statistically compared the emissivity values using 61 daytime and 25 nighttime ASTER scenes at the Railroad Valley playa, USA from 2000 to 2012. The spectral emissivity was consistent within +/- 0.007 at bands 13 (10.60 mu m) and 14 (11.30 mu m), although the emissivity at bands 10 to 12 (8.30, 8.65, and 9.10 mu m) varied temporally, probably due to surface condition changes and errors in atmospheric correction. The atmospheric correction based on the climatology data resulted in poor accuracy in TES-retrieved emissivity for nighttime data because the contribution of the atmospheric effect on the at-sensor radiance increases due to decreased surface radiance in nighttime. For the National Centers for Environmental Prediction (NCEP) data, the agreement, which is comparable to the nominal accuracy (= 0.015), between the emissivity based on the day and night ASTER data acquired on the same day suggested that NCEP data is sufficiently accurate to correct for the atmospheric effect on the ASTER data under nominally dry atmospheric condition. Although the spatial scales differ between ASTER data and laboratory data, the satellite and ground emissivity agreed within +/- 0.013. Despite the similarity of the spectral emissivity patterns, the ASTER emissivity was as much as 0.066 lower than the field emissivity at 8.30 mu m due perhaps to the roughness effect on the field emissivity. (C) 2014 Elsevier Inc. All rights reserved.

DOI:
10.1016/j.rse.2014.02.002

ISSN:
0034-4257; 1879-0704