Research

The Remote Sensing Group in the College of Optical Sciences at the University of Arizona is best known for its work on the in-flight, radiometric calibration of remote sensing imagers. Radiometric calibration in this context refers to the ability to take the data from a sensor and convert it to a standard energy scale. Implicit in radiometric calibration is the ability to compare data between scales and sensors. In other words, absolute radiometric calibration allows data from individual sensors to be compared directly so long as traceability to known standards is maintained and estimates of absolute accuracy are themselves accurate.

The importance of radiometric calibration can be understood by realizing that as of 2005, there were more than 30 sensors, excluding dedicated atmospheric, meteorological, and defense satellites, in space for the purpose of terrestrial imaging of the earth-atmosphere system. These sensors are a variety of national and international efforts including sensors from every continent except Antarctica and more than 12 countries. The sensors range in spatial resolution from 0.6 to 1000 m with spectral resolution ranges from multispectral to hyperspectral spanning the wavelength regions from visible, near-infrared, through the thermal infrared and beyond to radar wavelengths. Swath widths and repeat visits also span a wide range of values all leading to an opportunity to study earth-processes on a global scale in an unprecedented manner. Knowledge of the radiometric calibration is vital to allowing the comparisons of these varied sensors.

Nearly all of the sensors follow a typical calibration chronology of preflight laboratory calibration followed by in-flight characterization. The success of the preflight calibration is contingent upon following a set of well-defined protocols and this should allow comparisons between the output of two sensors calibrated in different facilities. Unfortunately, there are many issues that can cause preflight calibration results to differ from those of the sensor in space (such as the launch of the platform). Thus, there is a strong justification for using on-board calibrators and vicarious approaches for radiometric calibration.

Vicarious approaches rely on data that are independent of the sensor itself. Such approaches include viewing the moon, desert sites, arctic regions, and cloud tops. The Remote Sensing Group developed and uses vicarious approaches that rely on in-situ measurements. Three methods, the reflectance , irradiance , and radiance based techniques, have been in used since the mid-1980s. The advantage that the in-situ approaches have is better absolute accuracy than the model-based approaches. Current results in the mid-visible portion of the spectrum show that the Remote Sensing Group can produce absolute radiometric calibrations to better than 2%, both in accuracy and precision.

The reflectance-based approach for which the group is best known relies on measurements of the surface reflectance of a test site at the time of sensor overpass. Atmospheric measurements are made concurrent with the surface measurements and the results from both are used as input to a radiative transfer code to predict at-sensor radiance that is compared to the sensor output to provide the calibration. The Remote Sensing Group travels to large, non-vegetated areas for our measurements. The group depends on an array of well-understood field equipment characterized in the RSG’s extensive radiometric laboratory. The laboratory, which is NIST-traceable, is also used for the absolute calibration of other sensors including laboratory transfer radiometers.