Cloud retrievals from the TROPOMI UV/VIS/NIR measurements with aerosol signature

Kurzfassung

TROPOMI on board of Sentinel-5 Precursor (S5P) provides continuous daily distribution of several cloud properties, which are required as input for trace-gas retrievals. The operational TROPOMI cloud retrieval involves the ROCINN (Retrieval of Cloud Information using Neural Networks) algorithm, which retrieves the cloud height, cloud optical thickness and cloud albedo from NIR radiance measurements in and around the oxygen A-band (758-771nm). Within the ROCINN algorithm two different models are applied: the Clouds-as-Reflecting-Boundaries (CRB), where the cloud is a simple Lambertian reflector, and the Clouds-as-Layers (CAL), where the cloud is a homogeneous layer of scattering liquid-water spherical particles. ROCINN requires the calculation of the sun-normalized radiances from a radiative transfer model (RTM) for simulating clear-sky and cloudy atmospheres. The contribution of atmospheric molecules in the measured NIR radiances is incorporated into the forward model through the Rayleigh scattering model. However, the contribution of larger atmospheric particles (i.e., aerosols) has not been yet considered in the RTM simulations. Some TROPOMI cloud retrievals appear as contaminated by aerosols. OCRA (Optical Cloud Recognition Algorithm), the synergistic algorithm working prior to ROCINN, which computes a radiometric cloud fraction using a broad-band UV/VIS color space approach, often derives an elevated radiometric cloud fraction corresponding to the given aerosol conditions. When OCRA retrieves a radiometric cloud fraction above a certain threshold (e.g., the threshold is 5% for TROPOMI), ROCINN is triggered and tries to further retrieve the two additional cloud parameters height and optical thickness/albedo. In heavy aerosol loads, OCRA might return falsely elevated cloud fractions. Those false alarms can usually be identified because ROCINN retrieves a cloud height close to the surface level in such cases. Nevertheless, there are cases in which ROCINN cloud outputs do not refer to the surface properties of the scene, but to aerosol layers present in the same TROPOMI footprint. In this work, we evaluated the signature of aerosols in the TROPOMI cloud retrievals with TropOMAER (TROPOMI aerosol algorithm), which derives aerosol information in cloud-free and above-cloud aerosol scenes. Validation of the aerosol layers is done via synergistic ground-based measurements from a PollyXT multiwavelength-Raman-polarization lidar and an AERONET sun-photometer.