TEMPO/OCRA: Retrieval of radiometric cloud fraction from TEMPO UV-VIS measurements

Abstract

The Geo-Ring constellation comprises three geostationary instruments dedicated to air quality and atmospheric composition monitoring over the Northern Hemisphere with daytime hourly resolution. Following the successful launch of the European Sentinel-4/UVN (Ultraviolet-Visible-Near-infrared spectrometer) in July 2025, and in combination with the US-American TEMPO (Tropospheric Emissions: Monitoring of POllution, launched in April 2023, operational) and the South Korean GEMS (Geostationary Environment Monitoring Spectrometer, launched in February 2020, operational), the Geo-Ring constellation is now fully deployed.

The operational cloud products from Sentinel-4 will be based on the OCRA/ROCINN algorithm tandem. OCRA (Optical Cloud Recognition Algorithm) estimates the radiometric cloud fraction from UV-VIS measurements, while ROCINN (Retrieval Of Cloud Information using Neural Networks) builds on the OCRA output and O2 A-band measurements to retrieve two additional cloud parameters using optimal estimation theory. As the operational cloud products will not be publicly available before summer 2026, the OCRA algorithm may first be applied to the already operational Geo-Ring missions to assess the potential synergies with Sentinel-4 and other missions like EPIC (Earth Polychromatic Imaging Camera).

In this study, we adapt the OCRA algorithm to estimate the radiometric cloud fraction as observed by the TEMPO instrument from its UV-VIS measurements. The work addresses two main challenges. The first is the OCRA cloud-free map generation, for which we compare the differences between TEMPO-based cloud-free maps (i.e. mission-specific implementation) and adapted EPIC-based cloud-free maps (i.e. mission-synergetic implementation, which is especially important for the OCRA algorithm at mission start, when a mission-specific implementation is not yet feasible due to insufficient data availability). The second is the treatment of the surface reflectance effects in OCRA, for which we compare the differences between a data-driven approach using TEMPO measurements and a general approach based on RTM (Radiative Transfer Model) simulations.