Regularisation model study for the least-squares retrieval of aerosol extinction time series from UV/VIS MAX-DOAS observations for a ground layer profile parameterisation

Abstract

Abstract. The retrieval of tropospheric aerosol extinctions from MAX-DOAS observations of O4 using a small number of three or four extinction profile parameters suitable for boundary layer reconstruction is investigated with respect to the following questions. First, to what extent does this nominally over-constrained pure least-squares problem for the inversion of the radiative transfer equation require regularisation and how should parameters of the regularisation be chosen? Second, how can a lack of information in the under-constrained case be best compensated by using the information contained in a sequence of observations and by explicitly including intensities into the fit? The forward model parameterises the optical properties of the boundary layer aerosol by its extinction profile, single-scattering albedo and a Henyey–Greenstein phase function. Forward calculations are carried out online, i.e. without look-up tables. The retrieval uses a Tikhonov regularisation combined with an approximate L-curve criterion and empirical a priori information from the retrieval sequence based on previous valid solutions. The consistency of the approach is demonstrated in selected model case studies assuming a polluted urban scenario and westward viewing direction of the instrument. It is shown that a dynamic choice of the regularisation parameter is crucial for high aerosol load and large diurnal variations. The quality of the retrieval can be improved significantly if the retrieval sequence and thus the a priori is chosen according to the information content of the measurement series. Additional intensities improve the solution for all solar angles if suitably weighted. This flexible retrieval algorithm allows for reconstruction of aerosol profiles in the boundary layer for a wide range of viewing directions and extinctions. It can thus be applied to observational geometries scanning the sky in two angular dimensions and to retrieve further aerosol optical parameters in the boundary layer.

MAX-DOAS retrieval of aerosol extinction properties in Madrid, Spain

“…Therefore, the retrieved O 4 differential slant column densities (DSCDs) at different elevations can provide information about the impact of aerosol scattering on photon paths. By combining measurements of the O 4 absorption with radiative transfer model simulations, ground-based MAX-DOAS has been used in previous studies to determine atmospheric aerosol vertical extinction profiles and optical depths (e.g., Irie et al, 2008Irie et al, , 2009Li et al, 2010;Clémer et al, 2010;Hartl and Wenig, 2013;Hendrick et al, 2014;Vlemmix et al, 2015;Frieß et al, 2016).…”

MAX-DOAS observations of aerosols, formaldehyde and nitrogen dioxide in the Beijing area: comparison of two profile retrieval approaches

“…Stability is important for comparison with satellite observations if the number of available cases is very limited; accuracy over a wide range of profile heights is important if MAX-DOAS would be used to provide a climatology of profile heights for better a priori estimates in the satellite retrieval. The recent work by Hartl and Wenig (2013) provides indications that the Phillips-Tikhonov regularization method can be used for MAX-DOAS profile retrievals which are more stable and at the same time (potentially) less biased in a climatological sense. To our best knowledge, their method has not yet been applied to a long data set of real observations with a similar focus on the ability to retrieve accurate (first order) profile height estimates.…”

Ground-based MAX-DOAS observations of tropospheric aerosols, NO<sub>2</sub>, SO<sub>2</sub> and HCHO in Wuxi, China, from 2011 to 2014

“…Wang et al, 2014) or directly from the derived profile. The existing profile inversion schemes developed by different groups can be subdivided into two groups: the "full profile inversion" based on optimal estimation (OE) theory (Rodgers, 2000;Frieß et al, 2006Frieß et al, , 2011Wittrock, 2006;Irie et al, 2008Irie et al, , 2011Clémer et al, 2010;Yilmaz, 2012;Hartl and Wenig, 2013;Wang et al, 2013a, b) and the so-called parameterization approach using look-up tables (Li et al, 2010Vlemmix et al, , 2011Wagner et al, 2011). In comparison with the look-up table methods, the OE-based inversion algorithms are in principle easily applied to different species, measurement locations and instruments, but they require radiative transfer simulations during the inversion and can therefore be computationally expensive for large data sets.…”