828 
greenhouse gases (Kaufman and Gao, 1992). We can also retrieve the total water content in the vertical column, 
Wwater, from Eq. (11), which reads 
WWater = ' As- A 4 4) * WGas ' A 5 ] - A 4 * C0S6 lnR54 
(13) 
3- RESULTS 
3.1. Application to simulated data 
The validity of the modeled transmittances, Eqs. (12a) and (12b) and total water vapor, Eq. (13)), over the range 
of the adopted assumptions, particularly the lack of inclusion of the emissivity effect, has been checked from 
simulated AVHRR Channels 4 and 5 of NOAA-11 brightness temperature data. These data were calculated using 
the LOWTRAN-7 code (Kneizys et al., 1988) with the appropriate channel filter functions. Atmospheric profile 
input data were obtained from a set of 60 marine radiosoundings extracted from the TOVS initial guess retrieval 
(TIGR) (Scott and Chedin, 1981). These set of profiles cover a large range of temperature and moisture 
conditions. The calculation includes the effects of water vapor, molecular nitrogen, and the uniformly mixed 
gases (CC> 2 , N 2 O, O 3 , CO and CH 4 ). Three observation angles (0°, 23° and 46°), five surface temperatures, T-5, 
T, T+5, T+10, and T+20, (T is the first boundary layer temperature of the atmosphere) and 45 different 
combinations of Channels 4 and 5 emissivities (£4 and £5 ranging from 0.90 to 1.00 and five values of the 
emissivity differences, Ae=£ 4 -£ 5 , ranging from -0.02 to 0.02) were also considered for each radiosounding. From 
these simulations we have constructed Figure 1, which corroborates the correlation existing between the 
transmittances and the right-hand side of Eq. ( 8 ). We find A=0.98 and B=1.90, and therefore, 
FIGURE 1 .- Logarithm of the transmittance for AVHRR Channels 4 and 5 versus the logarithm of the 
transmittance ratio R 54 . The solid line is the linear regression of the points. 
(14a) 
(14b) 
0 
- 0.4 
- 0.8 
w -1.2 
c 
- 1.6 
- 2.0 
- 2.4 
- 0.8 - 0.7 - 0.6 - 0.5 - 0.4 - 0.3 - 0.2 - 0.1 0 
On the other hand we plot the total column content of water vapor computed from the 
radiosonde data against cos 01 nR 54 in Figure 2. From this figure we obtain, 
W=0.259-14.253 cos 6 lnR 54 - 11.649 (cosG lnR 54 ) 2 
(15)