Abstract
The Orbiting Carbon Observatory 2 and 3 (OCO-2 and OCO-3) instruments, launched in 2014 and 2019 respectively, employ multiple methodologies to refine the absolute radiometric scale. One technique, vicarious calibration, uses in-situ measurements taken at Railroad Valley, USA (RRV), along with near simultaneous measurements (within 0.5 hours) from the satellite instruments. The JPL field team has conducted measurement campaigns near every summer solstice since 2009 along with colleagues from the Greenhouse gases Observing SATellite (GOSAT) project, JAXA/NIES. Since 2012, additional ground datasets have been collected in the spring and fall to support additional targets acquired throughout the year. The RRV playa is also equipped with automated sensors developed and maintained independently by JPL and the Remote Sensing Group at University of Arizona. These automated data sets include surface pressure, surface reflectance, aerosol optical depth, and the ambient meteorological condition (temperature, wind speed, and relative humidity). The automated surface reflectance data sets provide a valuable time series, although the reported values are systematically lower than the spectroradiometer measurements collected by the field team. This is because the automated sensors do not directly measure the surface reflectance. They measure the reflected radiance at the bottom of atmosphere to estimate the surface reflectance but use a radiative transfer modeling (MODTRAN and the solar irradiance data) to estimate the downwelling solar flux incident on the RRV playa. We have combined the datasets collected during the field campaigns with these automated datasets to estimate scaling factors for the analysis of OCO-2 and OCO-3 measurements taken significantly before or after the campaigns. This combination could mitigate the systematic bias in the automated datasets and improve temporal coverage for OCO-2 and OCO-3. Other recent updates to the vicarious calibration data analysis procedure include adopting the TSIS-SIM solar irradiance spectra, and reprocessing data from the entire OCO-2 and OCO-3 missions using the ACOS Build 10 retrieval algorithm, and MODIS surface bidirectional reflectance factor version 6.1 data in the off-nadir correction to nadir surface reflectance to account for the impacts of OCO2/OCO3 viewing geometry. Results from the vicarious calibration analyses are presented here, reaching 3% accuracy of spectral radiance for the OCO collects closest to nadir, along with the estimated agreement with the satellite on-board calibrator results.
Vicarious Calibration of Orbiting Carbon Observatory 2 and 3
The Orbiting Carbon Observatory 2 and 3 (OCO-2 and OCO-3) instruments, launched in 2014 and 2019 respectively, employ multiple methodologies to refine the absolute radiometric scale. One technique, vicarious calibration, uses in-situ measurements taken at Railroad Valley, USA (RRV), along with near simultaneous measurements (within 0.5 hours) from the satellite instruments. The JPL field team has conducted measurement campaigns near every summer solstice since 2009 along with colleagues from the Greenhouse gases Observing SATellite (GOSAT) project, JAXA/NIES. Since 2012, additional ground datasets have been collected in the spring and fall to support additional targets acquired throughout the year. The RRV playa is also equipped with automated sensors developed and maintained independently by JPL and the Remote Sensing Group at University of Arizona. These automated data sets include surface pressure, surface reflectance, aerosol optical depth, and the ambient meteorological condition (temperature, wind speed, and relative humidity). The automated surface reflectance data sets provide a valuable time series, although the reported values are systematically lower than the spectroradiometer measurements collected by the field team. This is because the automated sensors do not directly measure the surface reflectance. They measure the reflected radiance at the bottom of atmosphere to estimate the surface reflectance but use a radiative transfer modeling (MODTRAN and the solar irradiance data) to estimate the downwelling solar flux incident on the RRV playa. We have combined the datasets collected during the field campaigns with these automated datasets to estimate scaling factors for the analysis of OCO-2 and OCO-3 measurements taken significantly before or after the campaigns. This combination could mitigate the systematic bias in the automated datasets and improve temporal coverage for OCO-2 and OCO-3. Other recent updates to the vicarious calibration data analysis procedure include adopting the TSIS-SIM solar irradiance spectra, and reprocessing data from the entire OCO-2 and OCO-3 missions using the ACOS Build 10 retrieval algorithm, and MODIS surface bidirectional reflectance factor version 6.1 data in the off-nadir correction to nadir surface reflectance to account for the impacts of OCO2/OCO3 viewing geometry. Results from the vicarious calibration analyses are presented here, reaching 3% accuracy of spectral radiance for the OCO collects closest to nadir, along with the estimated agreement with the satellite on-board calibrator results.