Implication of spatial uniformity on vicarious calibration using automated test sites

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.