Atmospheric Measurement Techniques (Apr 2023)

Insights into 3D cloud radiative transfer effects for the Orbiting Carbon Observatory

  • S. T. Massie,
  • H. Cronk,
  • A. Merrelli,
  • S. Schmidt,
  • S. Mauceri

DOI
https://doi.org/10.5194/amt-16-2145-2023
Journal volume & issue
Vol. 16
pp. 2145 – 2166

Abstract

Read online

Clouds impose radiance perturbations upon Orbiting Carbon Observatory (OCO-2)-measured spectra. The Spherical Harmonic Discrete Ordinate radiative transfer Method (SHDOM) code is applied in both idealized bar cloud and scene-specific calculations of 1D and 3D radiances in order to understand 3D cloud effects for a wide range of gas vertical optical depths and solar- and sensor-viewing geometries for ocean and land scenes. SHDOM calculations for 36 scenes over the Amazon and the Pacific are co-analyzed with Moderate Resolution Imaging Spectroradiometer (MODIS) radiance-based cloud distance data and the OCO-2 Lite file rawXCO2 for both quality flag =0 (QF0; best quality) and quality flag =1 (QF1; poor quality) data. SHDOM calculations of the ocean and land scenes indicate that the 1D / 3D radiance intensity ratios and rawXCO2 decrease concurrently as the nearest-cloud distance decreases towards zero, especially for the ocean glint QF1 data, which provide the clearest evidence of 3D cloud effects in OCO-2 retrievals. Yearly analysis of OCO-2 O2 A-band continuum radiances indicate that 3D cloud-brightening events are predominant over cloud-shadowing events; therefore, 1D / 3D intensity ratios are predominantly less than unity. Bias corrected (bcXCO2) at cloud distances between 0 and 20 km are calculated for 20∘ latitude bands for 2015–2018. These zonal averages are used to calculate 3D-cloud-effect biases for bcXCO2 data (with a positive bias indicating that OCO-2 underestimates bcXCO2). Averages of 3D-cloud-effect biases, weighted by the number of Lite file data points in each of the nearest-cloud distance bins, in the Northern and Southern hemispheres, are 0.16 (1.31) and 0.26 (1.41) ppm (parts per million), respectively, over the ocean, and −0.13 (0.51) and −0.08 (0.47) ppm over land for QF0 (QF1) data.