158 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 39 



of year indicates great differences in the composition of the atmosphere 

 and in the quantity of suspended particles of smoke and dust. Ob- 

 servations by Coblentz and Kahler* near Washington, D. C, show 

 that the total solar energy received at sea level may vary even on 

 clear days 25 percent (from 3.64 to 4.86 B. t. u. per square foot 

 minute) at noontime in September. The absorptive effect of heavy 

 smoke over large industrial areas can be judged by the simultaneous 

 observations showing that the intensity of sunshine in Chicago may be 

 only 55 percent of that at Madison, Wis.^ Except during great dust 

 storms, the depletion due to dust in the atmosphere is less obvious than 

 that due to smoke. The usual fine atmospheric dust, when considered 

 as including smoke, haze, and liquid particles, can be estimated as 

 causing about 10 percent depletion ^ at midday and, of course, more 

 as the sun approaches the horizon and the sun's rays pass more 

 obliquely through air for a greater distance. 



Solar energy received on a horizontal surface at the ground. — The 

 regular Weather Bureau pyrheliometers measure the insolation on a 

 horizontal surface inside an evacuated spherical glass. The readings, 

 being based on a calibration with an uncovered master pyrheliometer, 

 indicate the amount of solar energy received horizontally outside the 

 instrument. Figure 1 shows the average daily record for clear weather 

 at Fresno, Calif., in July and December, 1933. The most important 

 seasonal variation is the number of hours of sunshine per day — nearly 

 15 in midsummer, but only a little more than 9 in midwint-er. Another 

 seasonal variation that greatly reduces midwinter solar energy re- 

 ceived on a horizontal surface is the height of the sun in the sky. At 

 Fresno this is represented by cosines of the angles of incidence of 0.97 

 at noon in July and of 0.60 at noon in December. 



The graphic areas under the curves of figure 1 represent the total 

 energies received on clear days. Any cloudiness would reduce the 

 amount of energy received, and prolonged shade at noon would se- 

 riously interfere with the operation of a solar heater. The clear-day 

 records in figure 1 show the winter energy to be 37 percent of the 

 summer, but the total energy received during the whole month of 

 December was only one-sixth of that received in July. These figures 

 indicate that the average December cloudiness deprived Fresno of 

 more than half the maximum total energy. This is a seasonal effect in 

 the central valleys of California because cloudiness is .rare in the 

 summer. 



« Coblentz, W. W., and Kahler, H., A new spectropyrheliometer and measurements of 

 the component radiations from the sun. Dep. Commerce, Sci. Pap. Bur. Standards, vol. 

 16, Pap. 378, pp. 240-41, 1921. 



' Kimball, H. H., The distribution of energy in the visible spectrum of sunlight, sky- 

 light, and total daylight. Int. Uluminating Congr., Pap. No. 12* p. 10, 1928. 



