Dust aerosol optical depth and altitude retrieved from hyperspectral infrared observations (AIRS, IASI) and comparison with other aerosol datasets (MODIS, CALIOP, PARASOL)
Résumé
Remote sensing of aerosol properties in the visible domain has been widely used for a better characterization of these particles and of their effect on solar radiation. On the opposite, remote sensing of aerosols in the thermal infrared domain still remains marginal. However, knowledge of the effect of aerosols on terrestrial radiation is needed for the evaluation of their total radiative forcing. A key point of infrared remote sensing is its ability to retrieve aerosol optical depth as well as mean dust layer altitude, a variable required for measuring their impact on climate. Moreover, observations are possible night and day, over ocean and over land. Our algorithm is specifically designed to retrieve simultaneously coarse mode dust aerosol 10 µm optical depth (AOD) and mean layer altitude from high spectral resolution infrared sounders observations. In this context, results obtained from 6 years (2003-2008) of AIRS observations have been compared to other aerosol sensors on the A-Train. First, our AIRS-retrieved 10 µm dust optical depth shows a very good agreement with the 0.55 µm Aqua/MODIS optical depth product, particularly for tropical Atlantic regions downwind of the Sahara during the dust season, even far from the sources. Comparisons with PARASOL non-spherical coarse mode product confirm the agreement found between AIRS and other AOD products from A-Train instruments. Second, time series of the mean aerosol layer altitude are compared to the CALIOP Level-2 products starting June 2006. For regions located downwind of the Sahara, the comparison again shows a good agreement with a mean standard deviation between the two products of about 400 m over the period processed, demonstrating that our algorithm effectively allows retrieving accurate mean dust layer altitude. A 6-year global climatology of the aerosol 10 µm dust optical depth and of the layer mean altitude has also been established, emphasizing the natural cycles of Saharan dust. This algorithm has been designed for processing high spectral resolution infrared sounders in general and is now applied to IASI observations. Thanks to IASI higher spectral resolution, the selection of finer channels for aerosol detection allows an even more accurate determination of aerosol properties. Results obtained from the first year of IASI observations will be presented and compared to other aerosol products and measurements.