Extracted from Chimot, J., Global mapping of atmospheric composition from space – Retrieving NO2 from OMI, PhD book, Delft University of Technology (TU Delft), The Royal Netherlands Meteorological Institute (KNMI), July 2018. height and tropospheric
Remote sensing of particles is likely more complex than of trace gases as their description requires a multitude of parameters: size, concentration (sometimes as a function of particle size), particle shape, chemical composition, optical properties (scattering vs. absorption) and their horizontal and vertical distribution. All these parameters can generally not be simultaneously retrieved from one single satellite measurement.
Perhaps the most famous satellite sensors dedicated for aerosol observations are the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) on-board the Cloud Aerosol lidar and Infrared Pathfinder (CALIPSO) since 2006, and the Moderate Resolution Imaging Spectroradiometer (MODIS) on-board Terra and Aqua platforms. Both the CALIPSO and Aqua platforms fly together with Aura (including OMI) in the A-Train satellite constellation. By measuring the upwelling Solar radiation at the top of the atmosphere in several spectral channels (covering the visible and the thermal infrared), MODIS provides daily global information on the aerosol optical thickness (AOT or τ), defined as the extinction optical thickness integrated from the surface to the top of the atmosphere (Levy et al., 2013). MODIS aerosol retrieval is based on an ensemble of algorithms, each of them specific to diverse cases (bright surfaces such as desert, dark vegetation on continents, ocean) and the detection and filtering of clouds based on the thermal infrared channels (Remer et al., 2005, 2008; Levy et al., 2013). The MODIS τ products are usually recognized as reference by many studies. However, MODIS data do not contain information on aerosol vertical profile.
CALIOP is an active lidar instrument. It provides very accurate vertical atmospheric distribution of aerosol, together with their backscattering and extinction properties. However, it only looks at the nadir point of view and thus cannot provide a global coverage with a high frequency.
Observing the aerosol vertical distribution with a high revisit frequency (ideally daily) is likely one of the main remaining challenges for aerosol satellite remote sensing. To meet this objective, passive hyperspectral sensors are generally employed as they offer adequate spatial coverage with a good temporal resolution. The traditional technique is the slant column measurement of molecular dioxygen O2 which is well mixed and relatively stable in the atmosphere. This is a suitable proxy to determine the modified scattering height due to aerosol particles. Most of the algorithms focus on the O2-A band around 765 nm, on TROPOMI, OCO-2, GOME-2, and GOSAT space- borne sensors (Wang et al., 2012; Sanders et al., 2015). However, exploitation of this band remains challenging due to high sensitivity to surface albedo (brightness) and saturation (Nanda et al., 2017).
Although not optimized for aerosol monitoring, OMI allows to distinguish absorbing particles (from scattering) from its UV 330-388 nm band. The so-called UV Absorbing Index (UVAI) is derived by the OMI near-UV aerosol algorithms (OMAERUV and OMAERO) in the 330-388 nm spectral band (Torres et al., 2007). This benefits from a long heritage where similar techniques were developed for past sensors (TOMS, GOME and SCIAMACHY) (Torres et al., 1998). It relies on the measured change of spectral contrast, with respect to a pure Rayleigh atmosphere. Weakly absorbing or large non-absorbing particles are associated with near-zero or negative UVAI values. Furthermore, OMI-like sensors also include the O2-O2 spectral features in the UV and visible. The O2-O2 477 nm band is the strongest one, presenting a wider (over 10 nm) although weaker spectral absorption than O2-A. As analysed in my thesis research, this leads to high sensitivities for retrieving aerosol layer height in case of strong aerosol loading and with less challenges due to saturation.