A preferred method of ground-based vicarious calibration is the reflectance-based approach, which requires personnel to
be present at a test site during sensor overpass. The Remote Sensing Group at the University of Arizona developed an
instrumentation suite and methodology in 2004 to measure the surface and atmospheric characteristics in the absence of
personnel. Field campaigns typically occur at a rate of once per month during the academic year, and increase during the
summer months. The automated approach allows data to be collected during every overpass of large-footprint sensors
such as Terra and Aqua MODIS, and AVHRR, which are continuously collecting data.
The large-footprint-sensor site at Railroad Valley is 1 km2. In the absence of personnel, the surface bidirectional
reflectance factor is measured using five nadir-viewing radiometers that are currently located at the site. Their locations
are chosen based on the topography of the site in an effort to "completely" sample the 1-km2 area. This work quantifies
the uncertainty in predicting the surface reflectance of the 1-km2 area based on the point measurements of the automated
methodology. It also determines if the number of radiometers, and their positions, are suitable to characterize the site in a
spatial sense. These uncertainties are determined through the use of portable spectroradiometers, and high-spatial-resolution
QuickBird imagery.
|