Table 5 — Solar irradianoe at sea level for air 

 mass between 387.5 and 712.5 m^, and the per- 

 centage of total visible energy within 25-m// 

 intervals over this spectral region 

 [From data of Mson, 1940] 



■'• Values of Moon (1940) given at lO-m/i 

 intervals. Values in table not followed by a 

 footnote were obtained by graphical interpo- 

 lation between lO-mu intervals. 



the percentage of downwelling visible radiant 

 energy has been computed at six depths by 

 integrating the data for the 11 wavelengths (see 

 table 10) for Types I, II, and III water and 

 intermediate types. The wavelength specificity 

 and half-band width of the irradiance meter 

 under conditions of Type I, II, and III may be 

 readily computed by multiplying, wavelength- 

 by-wavelength, the sensitivity of the detector 

 unit by the relative spectral energy distribu- 

 tion at any depth. These calculations reveal 

 that in Type I ocean water at 100-m. depth, 

 the maximum sensitivity is at 475 m^, and the 

 half -band width is of the order of 50 m^. At 

 50 -m. depth in Type III water, the peak re- 

 mains at 475 m/i, but the half-band width is 



somewhat broader--about 70 m^. The differ- 

 ence between the spectral sensitivity and 

 half-band width of the detector in air and im- 

 mersed in water is a natural consequence of 

 a system in which one of the component filters , 

 sea water, is in effect continually changing its 

 spectral transmission characteristics. In the 

 calculation of the percentage of downwelling 

 visible irradiance remaining at various depths 

 (optical depth), the half-band width of the de- 

 tector remains nearly constant and has a peak 

 sensitivity at 475 m/j. Thus all attenuation 

 coefficients derived from measurennents using 

 the above equipment and filters have been as- 

 signed to 475 m^. 



Observational Technique and Computations 



Measurements of submarine irradiance were 

 always made between 1000 and 1430 hours 

 local time and nnost frequently between 1230 

 and 1400 hours. The measurements were made 

 in the following manner: 



1. The ship was oriented to reduce the 

 likelihood of interference from the ship's 

 shadow, 



2. The irradiance meter was lowered with a 

 bathythermograph that had a 150-m. depth 

 range (or, occasionally, with an electronic 

 depth-sensing element) attached, until the out- 

 put of the submerged cell was about 10 fia.. 

 The instrument was then raised or lowered to 

 the nearest lO-nn. mark on the photometer 

 cable, 



3. After the meter reached equilibrium, at 

 least one and usually three sets of simul- 

 taneous readings of the output of the deck and 

 submerged cells were made, one observer 

 reading one meter only. On days of variable 

 cloud cover, it often took as long as 10 nninutes 

 to obtain three sets of readings at a given 

 depth. 



4. The wire angle was measured with a hand 

 inclinometer by visually superimposing the 

 edge of a conventional wire -angle indicator 

 with the conducting cable and reading the angle. 

 The meter was then raised to the next 10- or 

 20-m. mark. The outputs of the cells were 

 again read (see 3 above). This procedure was 

 repeated until the photometer was at a depth 

 of 5 or 10 m,, or until the output of the sub- 

 merged cell exceeded 1,000 ^a.. 



5. The output of the detector was then cor- 

 rected for departure from linearity of re- 

 sponse. These values have been corrected 

 again to an arbitrary but constant deck cell 

 reading to allow for changes in incident radi- 

 ation. The irradiance meter depths were com- 

 puted from the wire angle measurennents in 

 combination with the maximum depth recorded 

 by the bathythermograph, or by using the depth 

 indicated by the depth-sensing element. 



11 



