SUNLIGHT AS A SOITRCE OF RADIATION 



to tliicknossos of 1 and 10 mm of distilled water a1 20°C calculated from 

 the accurate ah.s()ri)tioii coefficients of ('urcio and Petty (1951). The 

 water-vapor curve represents the transmission through 1.85 km of atmos- 

 phere along a horizontal path containing some haze and a total of 17 

 mm of precipitable water. That is, the water vapor in a column 1 cm 

 square and 1.85 km long would, if condensed to the liquid phase, form a 

 column of licjuid water 1 cm square and 17 mm long. The portion of the 

 water-vapor cur\'e for wave lengths longer than 0.9 /x was measured in 



1.00 



mm WATER 



17 mm WATER 

 VAPOR 



0.8 



1.0 



2.0 



22 



2.4 



2.6 



1.2 1.4 1.6 1.8 



WAVE LENGTH, /.^ 



Fig. 3-14. Solar inten.sity 'o and transmis.sion of water and water vapor. 



1949 by Gebbie et at. (1950). The short-wave-length portion of the 

 curve below 0.9 yu was from miscellaneous older measurements along 

 horizontal paths in the real atmosphere; the attenuation below 0.6 /x was 

 largely due to haze because water vapor is very transparent in this region. 

 The curves of Fig. 3-14 bring out the well-known differences in the 

 absorption of water in the liquid and vapor phases. For example, 

 beyond 1.4 /x the strength of absorption by liquid water is much stronger 

 than by water vapor. Thus, 10 mm of liquid water and 17 mm of water 

 vapor are opaque at 1.4 m, but, although the transmission of water vapor 

 rises at 1.65 /x to a high value, liquid water remains completely opaciue at 

 longer wave lengths. Also, it is seen that the absorption coefficients of 

 1 7 mm of water vapor and of 1 mm of liquid water are comparable on the 

 short-wave-length side of the 1.9-m water-vapor band, but, although 

 water vapor regains its transparency at 2 /x, 1 mm of liquid water does 

 not again become transparent. Furthermore, the wave lengths of maxi- 



