OMI total bromine monoxide (OMBRO) data product: algorithm, retrieval and measurement comparisons

• Published in: Atmospheric measurement techniques Vol. 12; no. 4; pp. 2067 - 2084
• Main Authors: Suleiman, Raid M, Chance, Kelly, Liu, Xiong, Gonzalez Abad, Gonzalo, Kurosu, Thomas P, Hendrick, Francois, Theys, Nicolas
• Format: Journal Article
• Language: English
• Published: Katlenburg-Lindau Copernicus GmbH 04.04.2019; Copernicus Publications
• Subjects: Additives; Air masses; Albedo; Albedo (solar); Algorithms; Analysis; Astrophysical observatories; Atmospheric composition; Bromine; Bromine compounds; Correlation analysis; Cross-sections; Distribution; Environmental aspects; Gases; GOME-2; Irradiance; Mathematical analysis; Measurement; Monitoring instruments; Ozone; Ozone monitoring; Pollution monitoring; Polynomials; Raman scattering; Raman spectra; Retrieval; Salt lakes; Satellite observation; Satellites; Seasonal distribution; Seasonal variation; Seasonal variations; Solar irradiance; Sulfur dioxide; Uncertainty; Volcanic eruptions; Volcanoes; Wavelength
• ISSN: 1867-8548; 1867-1381; 1867-8548
• DOI: 10.5194/amt-12-2067-2019

This paper presents the retrieval algorithm for the operational Ozone Monitoring Instrument (OMI) total bromine monoxide (BrO) data product (OMBRO) developed at the Smithsonian Astrophysical Observatory (SAO) and shows comparisons with correlative measurements and retrieval results. The algorithm is based on direct nonlinear least squares fitting of radiances from the spectral range 319.0–347.5 nm. Radiances are modeled from the solar irradiance, attenuated by contributions from BrO and interfering gases, and including rotational Raman scattering, additive and multiplicative closure polynomials, correction for Nyquist undersampling and the average fitting residual spectrum. The retrieval uses albedo- and wavelength-dependent air mass factors (AMFs), which have been pre-computed using a single mostly stratospheric BrO profile. The BrO cross sections are multiplied by the wavelength-dependent AMFs before fitting so that the vertical column densities (VCDs) are retrieved directly. The fitting uncertainties of BrO VCDs typically vary between 4 and 7×1012 molecules cm−2 (∼10 %–20 % of the measured BrO VCDs). Additional fitting uncertainties can be caused by the interferences from O2-O2 and H2CO and their correlation with BrO. AMF uncertainties are estimated to be around 10 % when the single stratospheric-only BrO profile is used. However, under conditions of high tropospheric concentrations, AMF errors due to this assumption of profile can be as high as 50 %. The retrievals agree well with GOME-2 observations at simultaneous nadir overpasses and with ground-based zenith-sky measurements at Harestua, Norway, with mean biases less than -0.22±1.13×1013 and 0.12±0.76×1013 molecules cm−2, respectively. Global distribution and seasonal variation of OMI BrO are generally consistent with previous satellite observations. Finally, we confirm the capacity of OMBRO retrievals to observe enhancements of BrO over the US Great Salt Lake despite the current retrieval setup considering a stratospheric profile in the AMF calculations. OMBRO retrievals also show significant BrO enhancements from the eruption of the Eyjafjallajökull volcano, although the BrO retrievals are affected under high SO2 loading conditions by the sub-optimum choice of SO2 cross sections.