Combined Retrieval of Cloud Parameters for TEMPO Measurements Over Canada Using Oxygen Dimer and B Band Absorption

Abstract Tropospheric Emissions: Monitoring of Pollution (TEMPO) provides valuable hourly measurements of airborne pollutants over North America. Data quality over snowy surfaces may degrade due to difficulty in accurately characterizing surface reflectivity and cloud properties, as snow and cloud are both highly reflective and therefore challenging to distinguish optically. Cloud properties for satellite‐based trace gas retrievals are commonly extracted from measurements of oxygen, through either direct absorption (A band, B band) or collisional absorption (oxygen dimer). Here we investigate a combination of the two methods, with potential for extracting additional cloud information over challenging scenes due to the different responses of the two methods to various atmospheric conditions. A combined retrieval is successfully applied to simplified simulated data, extracting cloud information under conditions which would cause either retrieval to fail on its own. A second technique is demonstrated on the simulated data which attempts to flag partially cloudy scenes in the absence of external surface reflectivity information. The individual retrievals are applied to TEMPO data and compared in order to investigate their compatibility for potential use in combined retrievals. Variability in effective cloud fraction is found to be on the order of 0.1, and a bias in optical centroid pressure (representing the altitude of the cloud) is observed between methods, up to 50 hPa for fully overcast scenes and higher for partially cloud scenes. Further investigation is required to determine if these discrepancies are due to retrieval error, instrument calibration, or physical differences between spectral bands.