ALBERT

Description: It has long been recognized that Saharan desert is the major source for long range transport of mineral dust over the Atlantic. The contribution from other natural sources to the dust load over the Atlantic has generally been ignored in previous model studies or been replaced by anthropogenically disturbed soil emissions. Recently, Prospero et.at. have identified the major dust sources over the Earth using TOMS aerosol index. They showed that these sources correspond to dry lakes with layers of sediment deposed in the late Holocene or Pleistocene. One of the most active of these sources seem to be the Bodele depression. Chiapello et al. have analyzed the mineralogical composition of dust on the West coast of Africa. They found that Sahelian dust events are the most intense but are less frequent than Saharan plumes. This suggests that the Bodele depression could contribute significantly to the dust load over the Atlantic. The relative contribution of the Sahel and Sahara dust sources is of importance for marine biogeochemistry or atmospheric radiation, because each source has a distinct mineralogical composition. We present here a model study of the relative contributions of Sahara and Sahel sources to the atmospheric dust aerosols over the North Atlantic. The Georgia Tech/Goddard Global Ozone Chemistry Aerosol Radiation and Transport (GOCART) model is used to simulate dust distribution in 1996-1997. Dust particles are labeled depending on their sources. In this presentation, we will present the comparison between the model results and observations from ground based measurements (dust concentration, optical thickness and size distribution) and satellite data (TOMS aerosol index). The relative contribution of each source will then be analyzed spatially and temporally.

Description: We compare MODIS (Moderate Resolution Imaging Spectroradiometer) satellite aerosol retrievals of spectral optical thickness and size parameters over ocean with the same quantities derived from AERONET (Aerosol Robotic Network) observations made at island and coastal sites. Over much of the globe, the satellite-derived quantities agree well with the AERONET quantities. However, in regimes dominated by desert dust aerosol, the agreement is less robust. In the dust regimes, the MODIS retrievals show greater spectral dependence and report smaller particle sizes than do the AERONET derivations. We suggest that the reason for this discrepancy is the nonspherical nature of desert dust particles, which the initial MODIS algorithm is not able to handle. Using the discrepancy between MODIS and AERONET derived spectral optical thickness as an asset, instead of a detriment, we reconstruct the aerosol phase functions that the MODIS algorithm would have needed in order to match the AERONET retrievals. No assumptions of particle shape are used in the derivation of these functions and the results are empirical total column, ambient phase functions. We compare the empirically derived phase functions with phase functions calculated from spheres and spheroids, both situations in which assumptions about particle shape must be made. The resulting empirical nonspherical phase functions will be included in future updates of the MODIS algorithm.

Description: Remote sensing of aerosol from the new satellite instruments (e.g. MODIS from Terra) and ground based radiometers (e.g. the AERONET) provides the opportunity to measure the absorption characteristics of the ambient undisturbed aerosol in the entire atmospheric column. For example Landsat and AERONET data are used to measure spectral absorption of sunlight by dust from West Africa. Both Application of the Landsat and AERONET data demonstrate that Saharan dust absorption of solar radiation is several times smaller than the current international standards. This is due to difficulties of measuring dust absorption in situ, and due to the often contamination of dust properties by the presence of air pollution or smoke. We use the remotely sensed aerosol absorption properties described by the spectral sin le scattering albedo, together with statistics of the monthly optical thickness for the fine and coarse aerosol derived from the MODIS data. The result is an estimate of the flux of solar radiation absorbed by the aerosol layer in different regions around the globe where aerosol is prevalent. If this aerosol forcing through absorption is not included in global circulation models, it may be interpreted as anomalous absorption in these regions. In a preliminary exercise we also use the absorption measurements by AERONET, to derive the non-aerosol absorption of the atmosphere in cloud free conditions. The results are obtained for the atmospheric windows: 0.44 microns, 0.66 microns, 0.86 microns and 1.05 microns. In all the locations over the land and ocean that were tested no anomalous absorption in these wavelengths, was found within absorption optical thickness of +/- 0.005.

Description: The ability of dust to absorb solar radiation and heat the atmosphere is one of the main uncertainties in climate modeling and the prediction of climate change. Dust absorption is not well known due to limitations of in situ measurements. New techniques to measure dust absorption are needed in order to assess the impact of dust on climate. Here we report two new independent remote sensing techniques that provide sensitive measurements of dust absorption. Both are based on remote sensing. One uses satellite spectral measurements, the second uses ground based sky measurements from the AERONET network. Both techniques demonstrate that Saharan dust absorption of solar radiation is several times smaller than the current international standards. Dust cooling of the earth system in the solar spectrum is therefore significantly stronger than recent calculations indicate. We shall also address the issue of the effects of dust non-sphericity on the aerosol optical properties.