In-flight absolute calibration of radiometric sensors over dark targets using vicarious methods

The ability to conduct in-flight, absolute radiometric calibrations of ocean color sensors will determine their usefulness in the decade to come. On-board calibration systems are often integrated into the overall design of such sensors and have claimed uncertainty levels below 5%. Independent means of system calibration are needed to confirm that the sensor is accurately calibrated. Vicarious (i.e. ground-referencing) methods are an attractive way to conduct this verification. This research describes the development of in-flight, absolute radiometric calibration methods which reference dark (i.e. low-reflectance) sites. The high sensitivity of ocean color sensors results in saturation over bright surfaces. Low-reflectance targets, such as water bodies, are therefore required for their vicarious calibration. Sensitivity analyses of the reflectance-based and radiance-based techniques, when applied to a water target, are performed. Uncertainties in atmospheric parameters, surface reflectance measurements, and instrument characterization are evaluated for calibrations of a representative ocean color sensor. For a viewing geometry near the sun glint region, reflectance-based uncertainties range between 1.6% and 2.3% for visible and near-IR wavelengths; radiance-based uncertainties range between 6.8% and 20.5%. These studies indicate that better characterization of aerosol parameters is desired and that radiometer pointing accuracy must be improved to make the radiance-based method useful. The uncertainty estimates are evaluated using data from a field campaign at Lake Tahoe in June, 1995. This lake is located on the California-Nevada border and has optical characteristics similar to oceanic waters. Aircraft-based radiance data and surface measurements of water reflectance are used to calibrate visible and near infrared bands of the Airborne Visible/InfraRed Imaging Spectrometer (AVIRIS). The vicariously-derived calibration coefficients are compared to those obtained from a preflight calibration of AVIRIS. The results agree at the 0.3-7.7% level for the reflectance-based technique, which is within the believed method uncertainties. Finally, as a consequence of this research, the testing and refinement of radiative transfer codes applicable to oceanic environments is accomplished. These modifications lead to an improvement in the prediction of top-of-atmosphere radiances over water targets