Tropospheric Ozone Profile Retrievals from Combining the UV and Visible Spectra: GOME-2 and TEMPO

The height-resolved distribution of ozone in the stratosphere and troposphere has been generally retrieved from satellite backscattered ultraviolet (BUV) measurements in the 270-330 nm using an optimal estimation based nonlinear fitting, but there is always a concern about the accuracy of the retrieved ozone in the lower troposphere due to insufficient photon penetration down to the lower troposphere and the surface, especially at extreme geometries. Sensitivity studies demonstrated that adding visible measurements in the Chappuis ozone band to UV measurements in the Hartley/Huggins ozone bands can significantly enhance the retrieval sensitivity to the boundary layer ozone due to deeper photon penetration to the surface in the visible than in the UV. This additional sensitivity motivates the first NASA EVI (Earth Venture Instrument) TEMPO (Tropospheric Emissions: Monitoring of Pollution) instrument to measure earthshine radiance in the visible (540-740 nm) in addition to UV (~290-490 nm) to make better ozone pollution measurements. However, this retrieval enhancement has not been practically verified yet from existing measurements due to the weak ozone absorption in the visible and strong interferences from other trace gases (O₄, H₂O, O₂), surface reflectance and aerosols and the requirement of accurate radiometric calibration across different spectral channels.

The goal of this work is to present the reliable joint UV/visible ozone retrievals using GOME/2 measurements (240-790 nm), which will be demonstrated in terms of both fitting and retrieval quality through comparison against ozonesonde observations. To reduce the visible simulation errors, we will improve the convolution and interpolation processes of high-resolution absorption spectra onto instrument resolution and sampling interval. Temperature-parameterized O₃ and O₄ cross sections will be taken from Brion–Daumont–Malicet (BDM) and Thalman datasets, respectively. The HITRAN 2018 database will be used to make a look up table for H₂O and O₂ cross sections as a function of temperature and pressure instead of time-consuming line-by-line on-line calculation. We will also improve the characterization of the wavelength dependence of surface albedo using surface albedo/BRDF climatology derived from a combination of ASTER, USGS, MODIS, SCIAMACHY databases. Finally, we will perform soft calibration to account for inconsistent inter-calibration between UV and VIS measurements and time-dependent GOME/2 instrument degradation.