A simple and versatile cloud-screening method for MAX-DOAS retrievals

We present a cloud-screening method based on differential optical absorption spectroscopy (DOAS) measurements, more specifically using intensity measurements and O-4 differential slant-column densities (DSCDs). Using the colour index (CI), i.e. the ratio of the radiance at two wavelengths, we define different sky conditions including clear, thin clouds/polluted, fully-cloudy, and heavily polluted. We also flag the presence of broken and scattered clouds. The O-4 absorption is a good tracer for cloud-induced light-path changes and is used to detect clouds and discriminate between instances of high aerosol optical depth (AOD) and high cloud optical depth (COD). We apply our cloud screening to MAX-DOAS (multi-axis DOAS) retrievals at three different sites with different typical meteorological conditions, more specifically suburban Beijing (39.75 degrees N, 116.96 degrees E), Brussels (50.78 degrees N, 4.35 degrees E) and Jungfraujoch (46.55 degrees N, 7.98 degrees E). We find that our cloud screening performs well characterizing the different sky conditions. The flags based on the colour index are able to detect changes in visibility due to aerosols and/or (scattered) clouds. The O-4-based multiple-scattering flag is able to detect optically thick clouds, and is needed to correctly identify clouds for sites with extreme aerosol pollution. Removing data taken under cloudy conditions results in a better agreement, in both correlation and slope, between the MAXDOAS AOD retrievals and measurements from other colocated instruments.