H. A. Hazen — Prevailing Wind Direction. 461 



Aet. LI. — Determination of "prevailing wind direction;" by 

 H. Allen Hazen. 



Various methods have been adopted for the graphic repre- 

 sentation of the prevailing wind direction. Perhaps the sim- 

 plest of these consists in drawing radial lines for each of the 

 eight principal points of the compass and making the length of 

 each line correspond to the number of times the wind has 

 blown from that direction. A broken line joining the ends of 

 these would enclose an area, and the portions of this area be- 

 tween the radials would indicate approximately the frequency 

 of the wind. If the winds were not of the same velocity from 

 all points, this could be indicated by figures on each line de- 

 noting the mean velocity from each direction. The difference 

 would be inappreciable in the United States, and in fact wher- 

 ever there is a difference the velocity will be greater in the line 

 of the prevailing direction, so that, allowing for velocity, we 

 would simply intensify the mean direction. In no case would 

 there be any error in the mean direction ensuing from a neglect 

 of a slight variation in the velocity. 



Lambert's formula. 



Perhaps the best method of getting this quantity is by 

 means of Lambert's formula. The ordinary computation of 

 this formula is exceedingly tedious, especially where several 

 thousand computations are to be made. I have prepared very 

 simple tables, for my own use, which do away with the use of 

 logarithms and enable any one at a glance to take out the 

 quantity sought. The whole process is a simple mechanical 

 one and a child can perform the work without difficulty. It 

 has seemed wise to prepare these tables for wider use, because, 

 as the science of meteorology advances the demand is for in- 

 creased accuracy in its results. The examination of any 

 monthly mean chart with wind directions accurately projected 

 upon it will show distinctly the usual law of circulation of 

 winds about high and low pressures, and where the winds do 

 not seem to follow this law it can be inferred that there is some 

 local peculiarity which affects the wind. These tables are spe- 

 cially commended to voluntary observers of the weather for 

 the computation of their mean wind direction. 



We have Lambert's formula as follows : 



A = _ E-W + (NE-S W) cos 45° + (SE-NW) cos 45° 

 N-S+(NE-SW) cos 45° — (SE-NW) cos 45° 



In which N, NE, E, etc., represent the number of times the 

 wind has been observed from that direction during the period 



