SUNLIGHT AS A SOURCE OF RADIATION 101 



physics, although it is of impcjrtaiR-e in the physical state of the atmos- 

 phere and in meteorology. 



The measurement of the spectral distribution of intensity in sunlight 

 and its correction for atmospheric attenuation has been a major function 

 of the Astrophysical Observatory of the Smithsonian Institution, begin- 

 ning in 1892 under the direction of S. P. Langley who invented the bolom- 

 eter and first measured the spectral distribution of energy in the solar 

 spectrum, and continuing under Abbott and others. Their measure- 

 ments have been made at Mt. Wilson, Mt. Whitney, and Washington, 

 D.C. A convenient summary of their work is to be found in the Smith- 

 sonian Physical Tables (Fowle, 1934b). Many details of method are 

 described in a later publication by Abbott et al. (1942). Independent 

 and, in some spectral regions, improved measurements of solar spectral 

 intensity have been made by other investigators. In nearly every case, 

 however, the measurements were scaled to fit the Smithsonian curves 

 which therefore remain the standards over most of the spectrum. 



In the ultraviolet and visible portions of the spectrum it is observed 

 that the atmospheric absorption follows an exponential law. Hence 



i = iV-^T^"'^, (3-2) 



where i and /n = the intensities of a beam of sunlight at the bottom and 

 top of the atmosphere, respectively, and a, the atmospheric attenuation 

 coefficient, refer to a wave-length interval from X to X + (fX; Z = the 

 zenith angle of the sun; and 7 = a factor which accounts for the curva- 

 ture of the earth, ^^alues for 7 are listed in standard tables; 7 is very 

 close to unity for Z < 80°. In the infrared for certain bands of water 

 vapor and other gases, Eq. (3-2) does not agree with the observed absorp- 

 tions. However, Eq. (3-2) is not wTong; the discrepancy is due to the 

 use of insufficient dispersion to resolve the narrow and complex structure 

 of many of the bands. 



The air mass M is defined by 



M = 7 sec Z. (3-3) 



From Eq. (3-3), .1/ - 1 for Z = and 7 = 1, and therefore .1/ is the 

 amount of atmosphere from the surface to space in a vertical direction. 

 Then 



(3-4) 



I = loC 



-aM 



To determine io on top of the atmosphere, i is measured for several 

 values of M and is plotted for each wave length on a logarithmic scale 

 against ]\I . The straight line thus obtained is extrapolated to zero air 

 mass, which gives lo when proper account is taken of the transmission of 

 the spectrograph and the spectral response of the accompanying bolome- 



