The Remote Sensing Group at the University of Arizona has been using reflectance-based vicarious calibration of earth-observing satellites since the 1980s. Among the sensors characterized by the group are the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and the MODerate Resolution Imaging Spectroradiometer (MODIS) that are both on NASA's Terra platform. The spatial resolution of MODIS requires that the group use a large-sized site such as Railroad Valley Playa, Nevada as a test site. In addition, the large footprint size of MODIS forced a modification to the ground-sampling scheme for the surface reflectance retrieval. This work examines the impact of the new sampling scheme through coincident ASTER and MODIS imagery making use of the higher resolution spatial resolution of ASTER. ASTER and MODIS imagery were obtained for dates on which both sensors imaged the Railroad Valley test site and ground-based data were collected at the site. The results of the comparison between the sensors shows differences in the radiometric calibration that exceed the accuracy requirements of the sensors, but that the sampling strategy for large-footprint sensors produces reflectance-based results at the same 3% level of accuracy as that for small-footprint sensors.
The Advanced Spaceborne Thermal Emission and Reflection radiometer (ASTER) sensor on the Terra spacecraft has been providing remote sensing data for the past five years. ASTER has three separate sensor sections including a sensor with six bands in the shortwave infrared section of the spectrum. The radiometric calibration of the SWIR sensor has been updated from preflight values based on the on-board calibration sources. The SWIR sensor shows evidence of crosstalk between SWIR bands which is probably optical in origin. The crosstalk was present during preflight calibration and is present in all data collected in-flight including calibration data. The effects of crosstalk can be partially removed by applying a crosstalk correction program. This correction changes the calibration of the system. In this paper we apply a vicarious calibration to crosstalk corrected ASTER imagery over high reflectance desert test sites using a reflectance based method. The updated calibration provides for better retrieval of spectral reflectance or radiance of ground targets
in ASTER SWIR imagery.
Ground-reference techniques for the Enhanced Thematic Mapper Plus (ETM+) on Landsat 7 are described. The reflectance-based approach for vicarious calibration has been applied by the Remote Sensing Group of the Optical Sciences Center to 59 sets of ground-based data collected at large uniform test sites imaged by Landsat-7 ETM+. The results of this work coupled with the apparent stability of the ETM+ sensor shows that a one-sigma precision less than 3% is currently being achieved. Band-by-band analysis of the precision gives insight into the effects of atmospheric correction and surface reflectance uncertainties giving an understanding of the error sources in this approach. Variations in results are not seen between test sites and atmospheric effects are not the primary cause of day-to-day variability. These results are discussed with an emphasis on the current state of vicarious calibration accuracies/precision as well
as areas for improvement and future accuracy expectations.