15 
[ODIS. The presence of 
rrnels due to chlorophyll 
ater absorption. Wet soil 
ipping capability, and a 
iquid water absorption 
ensing of aerosol and 
, p 39 , (derived from the 
the 2.1 pm channel, p 
vith vegetation cover or 
(Kaufman and Remer, 
bsorption using the total 
; 0.41, 0.47 and 0.66 pm 
•ns for the 0.41, 0.47 and 
onditions of MODIS and 
>ol model used in these 
iol optical thickness to a 
3 are based on the AFGL 
servations. Ground based 
function of the ambient 
ufman et al., 1994). The 
the measured radiances, 
applied. With analytical 
the Rayleigh scattering) 
described by Fraser et al. 
[ of 50x50 km (to exclude 
; 40th percentile is used) 
id boxes were no values 
1. If no such values are 
used and the correction 
iverse of eq. 1 and a look 
r each pixel is a result of 
50 km grid. An aerosol 
e next section. 
meida et al. (1991) as an 
expansion of the work of Shettle and Fenn (1979). The authors give a compilation of a large 
amount of data and tabulate the dominant type of tropospheric aerosols as a function of the 
latitude, longitude and the season. They also discuss stratospheric aerosol. From the 
physical aerosol properties: refractive index and size distributions, d'Almeida et al. compute 
the optical properties at the same spatial and temporal scales: extinction coefficient, single 
scattering albedo, asymmetry factor and phase function. Some studies were devoted to 
specific aerosols types: desert aerosols (d’Almeida, 1987, Shettle, 1984), maritime aerosols 
(Hoppel et al., 1990) or aerosols resulting from biomass burning in tropical regions (Crutzen 
and Andrea, 1990, Kaufman et al., 1992). The climatology is based on existing 
measurements, most of them taken at ground level. There is a danger that these 
measurements do not represent the whole atmospheric column or the properties of the 
ambient instruments. Hegg et al. (1994) showed recently, using data derived from the 
Sulfate Cloud And Radiation experiment - Atlantic (SCAR-A) that the aerosol size 
distribution varies with altitude in the North-East US. Optical measurements from ground 
based sun/sky radiometers (Kaufman et al., 1994 and next section) can supplement the 
climatology with aerosol measurements that are integrated on the whole column and 
describe the properties of the ambient aerosol. 
There may be a limited value in atmospheric correction based on aerosol climatology 
that can reduce the effects on global data sets from low frequency spatial variations in 
aerosol concentration and their seasonality. But aerosol climatology will be definitely 
important in supplying the appropriate aerosol model (size distribution, phase function and 
refractive index) that is needed for correcting satellite images using path radiance derived by 
the dark target approach. 
3.3 Ground based measurements with sun/sky radiometers 
The aerosol parameters needed for performing atmospheric corrections can be obtained 
from in-situ measurements. Aerosol optical thickness can be obtained from sunphotometer 
measurements (Volz, 1954; Flowers et al., 1969; Peterson et al., 1981). Flowers et al. (1969) 
carried on a network over the United States from 1961 to 1966, d'Almeida et al (1983) 
performed a similar experiment over North and West Africa from 1980 to 1982 with 11 
instruments, Holben et al. (1991) conducted similar effort in the Sahel from 1984 through 
1986 using 15 monitoring stations. Extension of the measurements to include the aerosol 
size distribution and scattering phase function was determined from inversion of solar 
almucantar measurements (Kaufman et al., 1994). Single scattering albedo can be estimated 
from the collection of particles on filters, preferably by aircraft sampling of the entire 
atmospheric boundary layer and measurements of their absorption. Alternatively it can be 
determined from accurate measurements of the downward flux or radiance (King et al., 
1979; Wang and Gordon, 1993). Ground-based measurements of the aerosol loading and 
optical properties is useful for establishing the aerosol climatology for a given region for 
which atmospheric corrections are applied. 
Sunphotometer network, the Background Air Pollution Monitoring Network 
(BAPMoN) has been operated at a global and daily scales. It is managed by the World 
Meteorological Organization and could be used for correcting satellite images. Nevertheless, 
a recent report (Forgan et al., 1994) pointed out the shortcomings of the BAPMoN 
measurements: poor calibration of the instruments and lack of good monitoring of the data 
quality. Measurements with an improved network of automatic sun/sky radiometers were 
conducted recently in the Amazon basin during intense biomass burning (Holben et al., 
1994); in Africa by several African and French laboratories (D. Tanre et al.) and in the 
Eastern US (Y. Kaufman et al.). The instruments measure in addition to the solar direct flux 
(e.g. sunphotometry) also the sky and aureole radiance distribution. The retrieved aerosol 
properties are the optical thickness, the size distribution from 0.05 pm to 10 pm, the aerosol 
total loading and the scattering phase function (Kaufman et al., 1994). In order to maintain 
high data quality the instruments transmit the data through satellite communication