Status Ten Years

Ten years ago, aerosols were treated quite simplistically in global models. Tro-pospheric aerosols were characterized primarily by coarse size (r > 0.5 pm) dust, as a natural component, and by accumulation mode size (0.05 > r > 0.5 pm) (non-absorbing) sulfate, as the anthropogenic component (where r is the particle radius). Even by ignoring organic aerosol sources, it was not easy to convert industrial sulfur emissions via chemical and transport processes to somewhat realistic sulfate aerosol mass distributions to exact impact assessments of anthropogenic pollution (Langner and Rodhe 1991). Early evaluations of model simulations using satellite remote-sensing data were difficult. Estimates for mid-visible AOD by AVHRR sensor data (Stowe et al. 2002) have been available since 1981. However, AVHRR retrievals were limited to (deep) ocean regions, and they suffered from calibration problems, overpass time drifts, and sensor discontinuity on consecutive platforms. In contrast, since 1979, TOMS sensor data have provided estimates for AOD and ro0 even over (snow-free) land regions (Torres et al. 2002). However, a coarse pixel size increased the potential for cloud contamination, and data were provided for an energetically less interesting ultraviolet spectral region. In addition, accuracy of TOMS aerosol data depends strongly on an assumed aerosol altitude. Comparisons of wavelength-normalized AOD annual averages (Figure 3.1) between TOMS and AVHRR reveal, at best, similar overall patterns with evidence of signifi cant differences in magnitude. Detailed comparisons are not possible because of data inconsistencies (e.g., overpass time, pixel size, sample) and differences in retrieval assumptions for aerosols (e.g., particle size or absorption) and environment (e.g., solar surface albedo, definition of cloud-free conditions). Some aspects of these quantitative limitations can be overcome using reference values from complementary ground-based remote sensing. Solar transmission measurements (from the ground) have several advantages over solar refl ection measurements (from space): they have a well-defined background and are insensitive to aerosol absorption. Ten years ago, however, the number of monitoring sites that provided AOD and AnP by sun photometry

Global datasets AOD

Global datasets AOD

Figure 3.1 Annual mid-visible aerosol optical depth (AOD) fields at 550 nm wavelength from remote sensing. Ground-based remote sensing AOD data by AERONET (aer) have been combined into a gridded product and expanded for better viewing; they establish an AOD reference. In comparison, annual AOD fields of different multiannual satellite retrievals are given (MIS: MISR; Mc5: MODIS-collection5; Mc4: MODIS-collection4; AVn: AVHRR-noaa; AVg: AVHRR-gacp; TOo: TOMS-old; TOn: TOMS-new; POL: POLDER). Times with enhanced stratospheric aerosol after major volcanic eruptions have been excluded, and listed values are global averages of non-zero data.

Figure 3.1 Annual mid-visible aerosol optical depth (AOD) fields at 550 nm wavelength from remote sensing. Ground-based remote sensing AOD data by AERONET (aer) have been combined into a gridded product and expanded for better viewing; they establish an AOD reference. In comparison, annual AOD fields of different multiannual satellite retrievals are given (MIS: MISR; Mc5: MODIS-collection5; Mc4: MODIS-collection4; AVn: AVHRR-noaa; AVg: AVHRR-gacp; TOo: TOMS-old; TOn: TOMS-new; POL: POLDER). Times with enhanced stratospheric aerosol after major volcanic eruptions have been excluded, and listed values are global averages of non-zero data.

was small, and measurements at different ground sites lacked coordination and comparison of calibrations.

The Global Aerosol Data Set (GADS) was the first serious attempt to establish a global climatology for aerosol (microphysical and) optical properties. Using the available data, GADS combined the sampling information on aerosols from different (mainly surface in-situ) sites with modeling (Koepke et al. 1997). Aerosols in GADS are classified by size and composition into ten main components: water soluble, water insoluble, soot, sulfate, two types of sea salt, and four types of minerals. Each component is given a particular composition, size, and humidification factor (WMO 1983; d'Almeida et al. 1991). Based on assumptions of the ambient relative humidity, Mie simulations (assuming spherical shapes) are applied to determine the spectral aerosol properties for AOD, ro0, and g, via the OPAC software package (Hess et al. 1998). GADS aerosols are represented locally by a mixture of up to four components. GADS distinguishes between seven different altitude profiles, and horizontal resolution is 5°*5° latitude longitude grid. Data are provided for two months (January and July) to represent a winter and a summer season. The major shortcomings of GADS include the limited seasonality (e.g., it misses the biomass burning maxima of the southern hemisphere), the restriction to a few classes with predefined values for particle size and spectral absorption (e.g., there is now a consensus on a much weaker mid-visible absorption for dust), and the necessity to provide vertically resolved data for the ambient relative humidity to account for the water uptake of the particles.

Was this article helpful?

0 0

Post a comment