Impact of 3D radiative transfer on NO2 remote sensing over built-up areas

The retrieval of atmospheric trace gases from ground-based, airborne and satellite-based remote sensing instruments relies on the assumption of horizontal homogeneity when computing air mass factors (AMF) with radiative transfer models (RTM). However, this assumption is not justified in urban areas, where surface properties and atmospheric composition have high spatial variability. To study the effects of horizontal inhomogeneity on trace gas retrievals, we implemented 3D box-AMFs in the MYSTIC solver of the libRadtran RTM as well as an urban canopy module to account for the effect of buildings. We demonstrate the importance of 3D radiative transfer for several case studies: First, we simulate 3D-box AMFs for a ground-based MAX-DOAS instrument. Second, we calculate NO2 slant column densities (SCD) for an airborne imaging spectrometer measuring the emission plume of a power plant and the spatially inhomogeneous emissions of traffic in an urban area with mid-rise buildings. Finally, we apply 3D-box AMFs to measurements with the APEX airborne imaging spectrometer over the city of Zurich. Our case studies show that 3D effects are important for interpreting ground-based and airborne measurements when the NO2 field has high spatial variability. For example, 3D effects induce a spatial smoothing of NO2 SCDs that can partly explain why airborne NO2 measurements are spatially smoother than simulations with dispersion models. In addition, the presence of building shadows shields some NO2 resulting in a non-negligible underestimation of NO2 columns when not considered in the retrievals. 3D-box AMFs also improve the interpretation of NO2 SCDs measured by the APEX imaging spectrometer over Zurich. In conclusion, we show that 3D radiative transfer effects should be considered when retrieving trace gases at high resolution in urban areas to reduce systematic biases, to account for spatial smoothing effects, and to better relate slant column observations to the true (3D) trace gas distribution. 3D radiative transfer effects will also be relevant for satellite instruments that measure trace gases at the kilometer scale or at even higher spatial resolution.