1:45 PM - 2:00 PM
[AAS05-12] Validation and correction of TROPOMI tropospheric NO2 column density data using 4AZ-MAXDOAS at Chiba, Japan
Keywords:Atmosphere, Satellite observation, Nitrogen dioxide, Atmospheric science, Atmospheric chemistry
To quantify and solve the reported negative bias in tropospheric nitrogen dioxide (tropNO2) column data from the latest sensor TROPOspheric Monitoring Instrument (TROPOMI) aboard the Sentinel-5 Precursor satellite, we conducted ground-based four-different-azimuth-viewing multi-axis differential optical absorption spectroscopy (4AZ-MAXDOAS) observations at Chiba, Japan (35.63ºN, 140.10ºE, 21 m a.s.l.) in 2019. A comparison with simultaneous tropNO2 observations by 4AZ-MAXDOAS enabled an evaluation of the spatial inhomogeneity of tropNO2, which was considered responsible for a bias in results. Supported by inter-directional differences of up to 40% among 4AZ-MAXDOAS data, we found a significant horizontal spatial inhomogeneity in tropNO2 around the observation site. We then compared 4AZ-MAXDOAS data with coincident TROPOMI data. TROPOMI data had a 53% smaller tropNO2 column density as an annual average compared to 4AZ-MAXDOAS data, confirming the negative bias in TROPOMI data. However, the correlation between the magnitude of the bias and the coefficient of variance in four-azimuth directional data from 4AZ-MAXDOAS observations was not evident. This suggested that NO2 horizontal inhomogeneity couldn't fully explain the observed bias. We also applied a correction using the tropNO2 vertical profiles from 4AZ-MAXDOAS and the averaging kernels from TROPOMI. The correction reduced the bias to ~30% for annual average data, depending on the tropNO2 profiles used, indicating that there was a significant contribution of tropNO2 vertical profiles to the bias. These results not only quantified the reported negative bias but also emphasized the effectiveness of validation and correction using multi-directional ground-based data.