An ozonesonde evaluation of spaceborne observations in the Andean tropics

Satellite observations of ozone in the tropics have feedback from in situ measurements at sea level stations, but the tropical Andes is a region that is yet to be included in systematic validations. In this work, ozonesondes launched from the equatorial Andes were used to evaluate total column ozone (TCO) measured by spaceborne sensors TROPOMI/S5P (2018-2021), GOME-2/MetOp-B, OMI/Aura, and OMPS/Suomi NPP (2014-2021). Likewise, we evaluated tropospheric column ozone (TrCO) measured by the first two. Additionally, we evaluated TCO and TrCO from reanalysis products MERRA-2 and CAMS-EAC4. Results indicate that TCO observations by OMPS/Suomi NPP produce the closest comparison to ozonesondes (- 0.2% mean difference) followed by OMI/Aura (+ 1.2% mean difference). Thus, they outperform the sensor with the highest spatial resolution of current satellite measurements, namely TROPOMI/S5P (+ 3.7% mean difference). This overprediction is similar to the one encountered for GOME-2/MetOp-B (+ 3.2% mean difference). A positive bias with respect to soundings was also identified in TrCO measured by TROPOMI/S5P (+ 32.5% mean difference). It was found that the climatology used by TROPOMI overpredicts ozone in the troposphere when compared with the mean of Andes measurements, while both data sets are essentially the same in the stratosphere. Regarding reanalysis products, MERRA-2 compares better to ozonesondes than CAMS, both for TCO and TrCO (mean differences are 1.9% vs. 3.3%, and 11.5% vs. 22.9%, respectively). Identifying spaceborne ozone measurements that currently perform the best over the region is relevant given the present conditions of rapidly changing atmospheric composition. At the same time, ozonesonde data in this work offer an opportunity to improve satellite observations in the Andean tropics, a challenging region for space measurements.

Ozone correlations between mid-tropospheric partial columns and the near-surface at two mid-atlantic sites during the DISCOVER-AQ campaign in July 2011

The current network of ground-based monitors for ozone (O₃) is limited due to the spatial heterogeneity of O₃ at the surface. Satellite measurements can provide a solution to this limitation, but the lack of sensitivity of satellites to O₃ within the boundary layer causes large uncertainties in satellite retrievals at the near-surface. The vertical variability of O₃ was investigated using ozonesondes collected as part of NASA's Deriving Information on Surface Conditions from COlumn and VERtically Resolved Observations Relevant to Air Quality (DISCOVER-AQ) campaign during July 2011 in the Baltimore, MD/Washington D.C. metropolitan area. A subset of the ozonesonde measurements was corrected for a known bias from the electrochemical solution strength using new procedures based on laboratory and field tests. A significant correlation of O₃ over the two sites with ozonesonde measurements (Edgewood and Beltsville, MD) was observed between the mid-troposphere (7-10 km) and the near-surface (1-3 km). A linear regression model based on the partial column amounts of O₃ within these subregions was developed to calculate the near-surface O₃ using mid-tropospheric satellite measurements from the Tropospheric Emission Spectrometer (TES) onboard the Aura spacecraft. The uncertainties of the calculated near-surface O₃ using TES mid-tropospheric satellite retrievals and a linear regression model were less than 20 %, which is less than that of the observed variability of O₃ at the surface in this region. These results utilize a region of the troposphere to which existing satellites are more sensitive compared to the boundary layer and can provide information of O₃ at the near-surface using existing satellite infrastructure and algorithms.