Information content of OCO-2 oxygen A-band channels for retrieving marine liquid cloud properties

Abstract

Abstract. An information content analysis is used to select channels for a marine liquid cloud retrieval using the high-spectralresolution oxygen A-band instrument on NASA's Orbiting Carbon Observatory-2 (OCO-2). Desired retrieval properties are cloud optical depth, cloud pressure thickness and cloud-top pressure and the optimal channels depend on the atmospheric state, cloud properties and position within the OCO-2 swath. Based on information content criteria we select a micro-window of 75 10 of the 853 functioning OCO-2 channels spanning 763.5-764.6 nm and perform a series of synthetic retrievals with perturbed initial conditions. We estimate posterior errors from the sample standard deviations and obtain ±0.75 in optical depth, ±12.9 hPa in both cloud-top pressure and cloud pressure thickness, although removing the 10 % of samples with the highest χ 2 reduces posterior error in cloud-top pressure to ±2.9 hPa and cloud pressure thickness to ±2.5 hPa. The application of this retrieval to real OCO-2 measurements is briefly discussed, along with limitations and the greatest caution is urged regarding 15 the assumption of a single homogeneous cloud layer, which is often, but not always, a reasonable approximation for marine boundary layer clouds.

The operational cloud retrieval algorithms from TROPOMI on board Sentinel-5 Precursor

“…In particular, for deviations of 50 % in the cloud geometrical thickness, the retrieval errors in the cloud top height and cloud optical thickness are lower than 0.4 and 2.0 km respectively. Note that for the accurate retrieval of cloud geometrical thickness from the O 2 A-band either multi-angular (Merlin et al, 2016) or high-spectral-resolution (Richardson and Stephens, 2017) measurements are needed.…”

Reply to Referee #2

“…Several cloud retrieval algorithms based on measurements in and around the O 2 A-band at 760 nm have been developed for the GOME type of sensors: these include the ICFA (Initial Cloud Fitting Algorithm; Kuze and Chance, 1994), FRESCO (Fast REtrieval Scheme for Clouds from the Oxygen Aband; Koelemeijer et al, 2001;Wang et al, 2008), SACURA (Semi-Analytical CloUd Retrieval Algorithm; Rozanov and Kokhanovsky, 2004), UV/NIR (Diedenhoven et al, 2007), and ROCINN algorithms. These are all based on the independent pixel approximation (IPA), which is the assumption that the "radiative properties of a single satellite "pixel" are considered in isolation from neighbouring pixels" (definition of the American Meteorological Society).…”

Reply to Referee #1

“…The CRB approach works best with large pixels (Kokhanovsky et al, 2007), such as those from GOME (footprint: 320 × 40 km 2 ), where different types of clouds are combined, and errors in the cloud model are usually self-compensating. Furthermore CAL is more accurate than CRB for optically thin clouds (Rozanov and Kokhanovsky, 2004), and these kinds of clouds are the most frequent on a global scale. Previous studies using TOMS and GOME/SCIAMACHY measurements have demonstrated that a plane-parallel scattering cloud model is superior to a Lambertian reflectance cloud model for trace gas retrievals: Ahmad et al (2004) and Diedenhoven et al (2007) respectively.…”

“…Several cloud retrieval algorithms based on measurements in and around the O 2 A-band at 760 nm have been developed for the GOME type of sensors: these include the ICFA (Initial Cloud Fitting Algorithm; Kuze and Chance, 1994), FRESCO (Fast REtrieval Scheme for Clouds from the Oxygen Aband; Koelemeijer et al, 2001;Wang et al, 2008), SACURA (Semi-Analytical CloUd Retrieval Algorithm; Rozanov and Kokhanovsky, 2004), UV/NIR (Diedenhoven et al, 2007), and ROCINN algorithms. These are all based on the independent pixel approximation (IPA), which is the assumption that the "radiative properties of a single satellite "pixel" are considered in isolation from neighbouring pixels" (definition of the American Meteorological Society).…”

“…For TROPOMI/S5P, with significantly smaller ground pixels (footprint: 7 × 3.5 km 2 ), we have developed the more sophisticated ROCINN_CAL algorithm presented in this paper. In CAL, clouds are modelled as optically uniform layers of scattering water droplets -with this more physically realistic scenario, CAL is expected to be more accurate than CRB, especially for optically thin clouds (Rozanov and Kokhanovsky, 2004). For implementation of CAL model, a detailed description of the parameterisation of liquid water clouds in the forward model is provided in Sect.…”