The relevance of aerosol in the retrieval of tropospheric NO2 from satellite - a study of model data applicability

Abstract

Nitrogen dioxide (NO2) is a key pollutant in the troposphere, being one of the main precursors of tropospheric ozone, and source of nitric acid, as well as contributing to global climate change. Tropospheric NO2 vertical columns can be determined from satellite observations, although some uncertainties are still associated with the retrieval process. The conversion from measured slant columns to vertical columns is accomplished with airmass factors (AMF) that are determined by radiative transfer (RT) models. While the measurement (instrumental) conditions are well assessed, improvement is still needed regarding the a priori information of atmospheric characteristics required for the estimation of AMFs (e.g., vertical distribution of the gas, aerosol loading and clouds). This thesis presents a sensitivity study focused on the impact of aerosol on the tropospheric NO2 AMF. Optical properties, size distribution, and vertical distribution of the aerosol were varied within several scenarios. Overall, the results show a tendency for two main opposite effects. On the one hand, enhancement of the measurement sensitivity occurs by means of multiple scattering, when aerosol is mixed with the trace gas. On the other hand, a shielding effect by an aerosol layer located above the NO2 is also verified. The identified pivotal factors for the AMF calculations were the relative vertical distribution of aerosol and NO2, the aerosol optical depth and the single scattering albedo, as well as the surface reflectance. A case study was developed, focusing on the impact on the NO2 measurements of volcanic ash emitted from Eyjafjallajökull during the spring of 2010. Aerosol and NO2 data from the EURAD chemical transport model (CTM) were used to design scenarios for the RT calculations. A small variation of AMFs was found, revealing that, in the days and region analysed, the satellite observations of NO2 were not significantly affected by the mentioned eruption. Nonetheless, it was verified that the conclusions of the study are dependent on the accuracy of the CTM data, and on the approach employed to account for (and determine) aerosol optical properties. Such findings highlight the potential challenges that can be faced in the future if model data are used in satellite retrievals. In addition, a model evaluation performed within the GEMS project is described, where global stratospheric and tropospheric NO2 columns predicted by two chemical transport models MOZART and TM5 are compared with SCIAMACHY observations. The evaluation exercise allowed for the identification of flaws in the model systems, showing problems with the prediction of high levels of pollution in some regions (e.g., East-Asia), and with the simulation of NO2 concentrations during biomass burning events.