Feb. i6, 1920 



Determination of Normal Temperatures 



507 



This result will be accurate for normal temperatures, and the weather 

 forecast will help one to predict the actual temperatures, because it is 

 abnormally warm just before and cold just after the passage of a cyclone. 



A less scientific method but one requiring only arithmetic and a knowl- 

 edge of the mean monthly temperature for the place will now be given 

 for the solution of this problem. 



From figure 2 it is seen that the curve is approximately flat at the top 

 from July 1 5 to August 1 5 and at the bottom from December 1 5 to Jan- 

 uary 15. To get the mean daily temperature for any particular day of 

 one of these months gimply use the mean value given for the month. 

 For other months of the year the mean daily temperature changes by 

 one-third of i ° F. a day. 



Suppose it is desired to determine the temperature at Ogden on Sep- 

 tember 21 at I p. m. In Table I it is seen that the mean September tem- 

 perature for Ogden is 63° F. Since the twenty-first is six days removed 

 from the fifteenth, multiply ^3 ° by 6 and get 2°. Subtract this from 63°, 



Fig. 5. — Hourly variation of temperatures expressed in percentages of the mean. 



obtaining 61° as the probable mean temperature for the day. Turn 

 now to Table II, where it Is seen that the 3 p. m. temperature Is 128 per 

 cent of the mean. Hence multiply this percentage by 61° and get 78° 

 as the probable temperature. This value could be improved much by 

 getting the weather forecast from the Weather Bureau as explained 

 earlier. 



The first term of this equation is not 100 per cent, because the mean 

 temperature of the day is not quite the average of the maximum and 

 the minimum for the day. 



To determine the temperature at 2 a. m., expressed as the percentage 

 of the mean, divide 2 by 24 and multiply by 360° ; 30° is obtained. Intro- 

 duce 30 for Q In the above equation, and with the aid of a trigonometric 

 table and a little multiplication the desired percentage will be obtained. 



This equation will apply for any location where the mean daily tem- 

 perature variation decreases as the mean dally temperature decreases, 

 so as to give the ratio of the difference between the maximum and the 

 mean to the mean a constant value of 1.29, as in Utah. The writers 

 think it is of rather wide application. 



The fact that this ratio comes out constant for all seasons is a coinci- 

 dence, inasmuch as It holds only for the Fahrenheit scale of degrees and 

 this scale was arbitrarily chosen. 



