A Consistent AVHRR Visible Calibration Record Based on Multiple Methods Applicable for the NOAA Degrading Orbits. Part II: Validation

Doelling, D., R. Bhatt, B. Scarino, A. Gopalan, C.O. Haney, P. Minnis, and K. Bedka (2016), A Consistent AVHRR Visible Calibration Record Based on Multiple Methods Applicable for the NOAA Degrading Orbits. Part II: Validation, J. Atmos. Oceanic Technol., 33, 2517-2534, doi:10.1175/JTECH-D-16-0042.1.
Abstract

Consistent cross-sensor Advanced Very High Resolution Radiometer (AVHRR) calibration coefficients are determined using desert, polar ice, and deep convective cloud (DCC) invariant Earth targets. The greatest AVHRR calibration challenge is the slow orbit degradation of the host satellite, which precesses toward a terminator orbit. This issue is solved by characterizing the invariant targets with NOAA-16 AVHRR observed radiances that have been referenced to the Aqua Moderate Resolution Imaging Spectrometer (MODIS) calibration using simultaneous nadir overpass (SNO) observations. Another benefit of the NOAA-16 invariant target–modeled reflectance method is that, because of the similarities among the AVHRR spectral response functions, a smaller spectral band adjustment factor is required than when establishing calibrations relative to a non-AVHRR reference instrument. The sensor- and band-specific calibration uncertainties, with respect to the calibration reference, are, on average, 2% and 3% for channels 1 and 2, respectively. The uncertainties are smaller for sensors that are in afternoon orbits, have longer records, and spend less time in terminator conditions.

The multiple invariant targets referenced to Aqua MODIS (MITRAM) AVHRR calibration coefficients are evaluated for individual target consistency, compared against Aqua MODIS/AVHRR SNOs, and selected published calibration gains. The MITRAM and SNO relative calibration biases mostly agree to within 1% for channels 1 and 2, respectively. The individual invariant target and MITRAM sensor relative calibration biases are mostly consistent to within 1% and 2% for channels 1 and 2, respectively. The differences between the MITRAM and other published calibrations are mostly attributed to the reference instrument calibration differences.

PDF of Publication
Download from publisher's website