NO. I WEATHER DOMINATED BY SOLAR CHANGES ABBOT 



SOLAR PERIODICITIES 



It would be encouraging from a forecaster's standpoint if definite 

 periodicities should be found in solar variations. In table 3 are given 

 lo-dav mean values of solar radiation from 1918 to 1930.^ A ten- 

 dency towards the recurrence of a certain form of 8 months" period 

 was discovered in the lO-day means. To evaluate this periodicity, 

 the lO-day mean values were arranged in a table of 9 lines of 24 con- 

 secutive values each, beginning with May, 1924. Mean values of the 

 24 columns being computed, they resulted thus : 



8-month ('criod 



Direct Means" 40 41 42 41 44 41 41 42 41 43 4- 4° 4' 4^ 



Smoothed Means 40 41 4- 4- 43 4^ 4^ 4^ 4-^ 42 4i 4' 41 4i 



Smoothed Departures o +1 +2 +2 +3 +2 +2 +2 +2 +2 +1 +1 +1 +1 



Direct Means" 39 41 4° 38 41 4i 39 38 37 37 



Smoothed Means 41 40 40 40 39 39 39 38 37 37 



Smoothed Departures + 1 o o o — i — i — i — 2 —3 — 3 



" First two figures omitted. Thus for 1.940 calories, I substitute 40. Departures are given 

 from 1.940, omitting three figures. 



From these numbers a smoothed curve was drawn which gave the 

 departures from 1.940 calories. Subtracting these departures, the 

 original data were cleared of the 8-month periodicity from January, 

 1924, to December, 1930. It was then perceived that another peri- 

 odicity of 1 1 months seemed present. By a similar arrangement 

 in lines of ;^^ consecutive revised lO-day means of solar constant 

 numbers, the following values were computed, representing the 

 1 1 -month periodicity: 



ii-inonth period 



Direct Means 40 41 39 38 38 36 38 39 35 37 37 34 38 40 



Smoothed Means 41 40 39 38 38 37 37 37 36 36 36 37 38 39 



Smoothed Departures i o — i — 2 — 2 — 3 —3 — 3 — 4 — 4 — 4 — 3 — 2 — i 



Direct Means 40 41 43 44 41 40 38 42 42 40 42 45 43 46 



Smoothed Means 40 41 42 42 41 40 40 41 41 42 42 43 45 46 



Smoothed Departures o i 2 2 i o o i i 2 2 3 5 6 



Direct Means 44 45 43 43 41 



Smoothed Means 45 44 43 42 41 



Smoothed Departures 5 4 3 2 i 



As these two periodicities had been evaluated solely from results 

 of 1924 to 1930, I desired to see whether they were also in evidence 

 from 1918 to 1923. For this purpose, I made temj^lates fitting the 

 smoothed-curve departures for both periodicities. These templates 

 I traced again and again in their proper phases to fill the entire period 



^The best values are those obtained since January, 1924. Prior to August, 1920, 

 all observations were made in the outskirts of the city of Calama, amid dust and 

 smoke, and with less perfect equipment than subsequently. Prior to January, 191 9, 

 there was only one observation per day and by the " long " method. 



