Satellite remote sensing of aerosols using geostationary observations from MSG-SEVIRI

Abstract

Aerosols play a fundamental role in physical and chemical processes affecting regional and global climate, and have adverse effects on human health. Although much progress has been made over the past decade in understanding aerosol-climate interactions, their impact still remains one of the largest sources of uncertainty in climate change assessment. The wide variety of aerosol sources and the short lifetime of aerosol particles cause highly variable aerosol fields in both space and time. Groundbased measurements can provide continuous data with high accuracy, but often they are valid for a limited area and are not available for remote areas. Satellite remote sensing appears therefore to be the most appropriate tool for monitoring the high variability of aerosol properties over large scales. Passive remote sensing of aerosol properties is based on the ability of aerosols to scatter and absorb solar radiation. Algorithms for aerosol retrieval from satellites are used to derive the aerosol optical depth (AOD), which is the aerosol extinction integrated over the entire atmospheric column. The aim of the work described in this thesis was to develop and validate a new algorithm for the retrieval of aerosol optical properties from geostationary observations with the SEVIRI (Spinning Enhanced Visible and Infra-Red Imager) instrument onboard the MSG (Meteorological Second Generation) satellite. Every 15 minutes, MSG-SEVIRI captures a full scan of an Earth disk covering Europe and the whole African continent with a high spatial resolution. With such features MSG-SEVIRI offers the unique opportunity to explore transport of aerosols, and to study their impact on both air quality and climate. The SEVIRI Aerosol Retrieval Algorithm (SARA) presented in this thesis, estimates the AOD over sea and land surfaces using the three visible channels and one near-infrared channel of the instrument. Because only clear sky radiances can be used to derive aerosol information, a stand-alone cloud detection algorithm was developed to remove cloud contaminated pixels. The cloud mask was generated over Europe for different seasons, and it compared favorably with the results from other cloud detection algorithms - namely the cloud mask algorithm of Meteo-France for MSG-SEVIRI, and the MODIS (Moderate Resolution Imaging Spectroradiometer) algorithm. The aerosol information is extracted from cloud-free scenes using a method that minimizes the error between the measured and the simulated radiance. The signal observed at the satellite level results from the complex combination of the surface and the atmosphere contributions. The surface contribution is either parameterized (over sea), or based on a priori values (over land). The effects of atmospheric gases and aerosols on the radiance are simulated with the radiative transfer model DAK (Doubling-Adding-KNMI) for different atmospheric scenarios. The algorithm was applied for various case studies (i.e. forest fires, dust storm, anthropogenic pollution) over Europe, and the results were validated against groundbased measurements from the AERONET database, and evaluated by comparison with aerosol products derived from other space-borne instruments such as the Terra/- Aqua-MODIS sensors. In general, for retrievals over the ocean, AOD values as well as their diurnal variations are in good agreement with the observations made at AERONET coastal sites, and the spatial variations of the AOD obtained with the SARA algorithm are well correlated with the results derived from MODIS. Over land, the results presented should be considered as preliminary. They show reasonable agreement with AERONET and MODIS, however extra work is required to improve the accuracy of the retrievals based on the proposed method