829 
[he relationship is clearly non linear, for AVHRR/2 data. A linear fit was also tried, but it actually gave a 
curvature in the residuals, that disappear when a second-order polynomial is fitted. 
cos B In R _ 
54 
FIGURE 2,- Total water vapor at nadir versus the modeled transmittance ratio. The correlation coefficient (r) and 
the standard error of estimate (s) are given in the figure. The solid line is the quadratic regression of the points. 
Now, the performance of the numerical expressions given by Eqs. (14a), (14b) and (15) is to be 
analyzed. To accomplish this, a new simulation using five standard atmospheres included in LOWTRAN-7 code 
(tropical, midlatitude summer, midlatitude winter, subartic summer, and subartic winter) has been made. The 
results are given in Table 1. In spite of the fact that the table includes the worst situation (an observation angle 
of 46°) the results corroborate the validity of the modeled transmittance and the total water vapor. 
TABLE ].- Mean and standard deviation between the exact values oft 4 , T 5 and W calculated from LOWTRAN-7 
and the estimated values obtained from Eqns. (14a), (14b) and (15) using the simulated brightness temperatures 
in Channels 4 and 5, through R 45 , as a function of several combinations of the emissivity difference, Ae=£ 4 -es, 
over N adjacent fields of view. 
46° 
Ae 
At 4 
ax 4 
AT 5 
ax 5 
AW fern) 
aW(cm) 
- 0.02 to 0.02 
-0.013 
0.016 
- 0.010 
0.012 
-0.039 
0.079 
- 0.02 to 0.01 
- 0.022 
0.026 
-0.023 
0.028 
0.014 
0.066 
- 0.02 to 0.0 
-0.031 
0.037 
-0.035 
0.041 
0.064 
0.097 
- 0.01 to 0.01 
-0.016 
0.019 
-0.014 
0.017 
-0.024 
0.068 
- 0.01 to 0 
-0.025 
0.029 
-0.027 
0.032 
0.028 
0.070 
0 
-0.017 
0.021 
-0.016 
0.020 
-0.016 
0.063 
- 0.01 
-0.034 
0.040 
-0.039 
0.045 
0.078 
0.109 
- 0.02 
-0.047 
0.054 
-0.056 
0.065 
0.150 
0.183 
0.01 
0.004 
0.005 
0.011 
0.013 
-0.130 
0.160 
0.02 
0.028 
0.032 
0.043 
0.050 
-0.266 
0.310