[1] Large sets of filtered actinometer, filtered pyrheliometer and Sun photometer measurements have been carried out over the past 30 years by various groups at different Arctic and Antarctic sites and for different time periods. They were examined to estimate ensemble average, long-term trends of the summer background aerosol optical depth AOD(500 nm) in the polar regions (omitting the data influenced by Arctic haze and volcanic eruptions). The trend for the Arctic was estimated to be between À1.6% and À2.0% per year over 30 years, depending on location. No significant trend was observed for Antarctica. The time patterns of AOD(500 nm) and Å ngström's parameters a and b measured with Sun photometers during the last 20 years at various Arctic and Antarctic sites are also presented. They give a measure of the large variations of these parameters due to El Chichon, Pinatubo, and Cerro Hudson volcanic particles, Arctic haze episodes most frequent in winter and spring, and the transport of Asian dust and boreal smokes to the Arctic region. Evidence is also shown of marked differences between the aerosol optical parameters measured at coastal and high-altitude sites in Antarctica. In situ optical and chemical composition parameters of aerosol particles measured at Arctic and Antarctic sites are also examined to achieve more complete information on the multimodal size distribution shape parameters and their radiative properties. A characterization of aerosol radiative parameters is also defined by plotting the daily mean values of a as a function of AOD(500 nm), separately for the two polar regions, allowing the identification of different clusters related to fifteen aerosol classes, for which the spectral values of complex refractive index and single scattering albedo were evaluated. Citation: Tomasi, C., et al. (2007), Aerosols in polar regions: A historical overview based on optical depth and in situ observations,

The SAMUM field campaign in southern Morocco in May/June 2006 provides valuable data to study the emission, and the horizontal and vertical transports of mineral dust in the Northern Sahara. Radiosonde and lidar observations show differential advection of air masses with different characteristics during stable nighttime conditions and up to 5‐km deep vertical mixing in the strongly convective boundary layer during the day. Lagrangian and synoptic analyses of selected dust periods point to a topographic channel from western Tunisia to central Algeria as a dust source region. Significant emission events are related to cold surges from the Mediterranean in association with eastward passing upper‐level waves and lee cyclogeneses south of the Atlas Mountains. Other relevant events are local emissions under a distinct cut‐off low over northwestern Africa and gust fronts associated with dry thunderstorms over the Malian and Algerian Sahara. The latter are badly represented in analyses from the European Centre for Medium–Range Weather Forecasts and in a regional dust model, most likely due to problems with moist convective dynamics and a lack of observations in this region. This aspect needs further study. The meteorological source identification is consistent with estimates of optical and mineralogical properties of dust samples.

[1] Aerosol remote sensing requires techniques enabling the determination of aerosol optical thickness (AOT) over land surfaces, because the most important sources (continental aerosols, anthropogenic aerosols, biomass burning, desert dust, volcano eruptions and others) are on continents. Here a retrieval method for the AOT over land surfaces from top-of-atmosphere (TOA) radiance using nadir looking instruments of the ocean color type (like Ocean Color and Temperature Sensor (OCTS), Sea viewing Wide Field Sensor (SeaWiFS), Moderate resolution Imaging Sensor (MODIS) or Medium Resolution Imaging Sensor (MERIS)) is presented. It is scheduled as an off-line procedure for the ENVISAT radiometers SCIAMACHY and MERIS. The method is based on lookup tables (LUT) between the AOT and the aerosol reflectance for wavelength <0.67 mm. The aerosol reflectance is obtained from TOA reflectance accounting for Rayleigh path reflectance and the apparent spectral surface reflectance. Over land the surface reflectance is estimated by a mixing model of bare soil and green vegetation spectra, tuned by the normalized differential vegetation index (NDVI) of the satellite scene. The method has been tested and validated with SeaWiFS data and with aerosol properties of the closure experiment LACE-98 (Lindenberg Aerosol Charactrization Experiment). For short wave channels (0.412-510 mm) an agreement between the retrieved and ground-based data of 20% is achieved. Thus the method enables the investigation of AOT over land, yielding the regional turbidity situation as well as the identification of aerosol sources like large cities, large fire plumes, haze, small scale dynamical events and also thin cirrus clouds.

[1] Nearly pure Saharan dust was observed with an Aerosol Robotic Network (AERONET) Sun photometer, several Raman and high spectral resolution lidars, and airborne in situ instruments during the Saharan Mineral Dust Experiment (SAMUM) 2006 in Morocco. In the framework of a case study we present particle shape-dependent dust properties, i.e., backscatter coefficients, extinction-to-backscatter (lidar) ratios, and linear particle depolarization ratios. These parameters can be inferred from AERONET's latest version of the mineral dust retrieval algorithm. The parameters can be measured with multiwavelength Raman/depolarization lidar without critical assumptions on particle shape. Lidar ratios inferred from the AERONET Sun photometer data tend to be larger than lidar ratios measured directly with lidar. Linear dust depolarization ratios were derived for several measurement wavelengths from the data products of the AERONET Sun photometer. The depolarization ratios tend to be smaller than the depolarization ratios measured with lidar. The largest differences exist in the near-ultraviolet wavelength range. Particle axis ratios were determined with scanning electron microscopy. The axis ratio distribution differs significantly from the axis ratio distribution that is assumed in the AERONET retrievals. If the axis ratio distributions measured during SAMUM are used, the reproducibility of the lidar data products improves. The differences may in part be caused by an insufficient understanding of the light-scattering model that is used in the AERONET algorithm. The results of the present study will be used to develop a dust light-scattering model that will serve as the theoretical basis for the inversion of optical data into dust microphysical properties.

Abstract. Regular aerosol observations based on wellcalibrated instruments have led to a better understanding of the aerosol radiative budget on Earth. In recent years, these instruments have played an important role in the determination of the increase of anthropogenic aerosols by means of long-term studies. Only few investigations regarding longterm trends of aerosol optical characteristics (e.g. aerosol optical thickness (AOT) andÅngström exponent (ÅE)) have been derived from ground-based observations. This paper aims to derive and discuss linear trends of AOT (440, 675, 870, and 1020 nm) andÅE (440-870 nm) using AErosol RObotic NETwork (AERONET) level 2.0 spectral observations. Additionally, temporal trends of coarse-and fine-mode dominant AOTs (CdAOT and FdAOT) have been estimated by applying an aerosol classification based on accurateÅE andÅngström exponent difference (ÅED). In order to take into account the fact that cloud disturbance is having a significant influence on the trend analysis of aerosols, we introduce a weighted least squares regression depending on two weights: (1) monthly standard deviation (σ t ) and (2) number of observations per month (n t ).Temporal increase of FdAOTs (440 nm) prevails over newly industrializing countries in East Asia (weighted trends; +6.23 % yr −1 at Beijing) and active agricultural burning regions in South Africa (+1.89 % yr −1 at Mongu). On the other hand, insignificant or negative trends for FdAOTs are detected over Western Europe (+0.25 % yr −1 at Avignon and −2.29 % yr −1 at Ispra) and North America (−0.52 % yr −1 for GSFC and −0.01 % yr −1 at MD Science Center). Over desert regions, both increase and decrease of CdAOTs (+3.37 % yr −1 at Solar Village and −1.18 % yr −1 at Ouagadougou) are observed depending on meteorological conditions.