SMITHSONIAN MISCELLANEOUS COLLECTIONS 



VOL. J'] 



VERIFICATION— RAINFALL 



In the following verification of precipitation predicted in the weekly 

 w^eather forecast, only rain fallen within the small zone in heavy 

 shading on the accompanying map, figure 5, has been used, and days 

 with fog and great humidity which did not materialize into rain 

 were not included. 



Atmospheric changes over the region of the River Plata usually 

 cover a great area, approximately indicated by dotted shading on 

 the map of an extension of about 600,000 km.^ besides the small 

 zone 50,000 km.' immediately around the River Plata. 



Table 13. — Villa Ortuzar Observatory, Chacarita, Buenos Aires, 



Lot. 34° 35' 15" S., Long. 58° 28' 15" W. H. 25 mt. 



Average rainfall in mm. 



(1901-1920) 



Jan. 

 79 



Feb. 



66 



Mar. 



TOO 



Apr. 



May 



72 



June 



51 



July 



55 



Aug. 



57 



Sept. 



73 



Oct, 



85 



Nov. 



100 



Dec. Year 



104 962 



Monthly rainfall observed, and deviations from the normal. 



Notwithstanding that the verification results over the greater area 

 would have improved the percentage, these w'ere not taken into 

 account, so as not to fall into the very common error made by long- 

 range forecasters who claim for verification purposes, atmospheric 

 phenomena occurring over great extensions of territories, and fre- 

 quently over the whole world. 



Table 13 shows the normal rainfall for Buenos Aires for a period 

 of 20 years from 1901 to 1920, and also the monthly values observed 

 during the years 1922, 1923, and 1924 with their corresponding 

 departures from the normal. The year 1924 was very dry compared 

 with the normal, the precipitation being 32 per cent under normal. 



With reference to tables 14 to 18, the column headed " Day " 

 indicates the days of the forecast week. As the prediction is issued 

 on Wednesday, Thursday is the first day, Friday the second, etc. In 



