Section 151, Atmospheric Turbulence 



Maps of pressure distribution give us an idea of the direction and Intensity of winds, while 

 maps of pressure anomalies show the deviations from average conditions. 



But when we average the data for any particular time interval, we thereby exclude all varia- 

 tions of the observed factor in the same value but with opposite sign. Thus, measuring the factors 

 of sea currents at one particular geographical point of the sea and then adding up the results, we 

 automatically exclude all periodic currents, and, in particular, tidal currents. These very tidal 

 phenomena, however, may be reflected in a most decided fashion on sea behavior and on ice con- 

 ditions.* The same reasoning may obviously be applied to the atmosphere. For the study of ice 

 behavior as conditioned by winds, it is necessary to know not only the direction and intensity of 

 winds but also their variability in time. 



The study of wind roses may be employed for solution of this problem. But it is difficult to 

 translate the wind roses into figures which may be manipulated in the future. Besides, wind roses 

 which are constructed for shore stations may have little resemblance to the wind roses for the 

 same time interval constructed for even the nearest points of the open sea. It is well known that 

 winds are distorted by local conditions. I have therefore preferred to use a different method, as 

 follows, recalling that: 



1. Atmospheric pressure is a meteorological factor which is most simply and accurately 

 determined by, and at the same time least dependent on local conditions. 



2. All other meteorological factors are either connected with pressure or depend on it 

 directly, which is why the isobars are the basic lines on a synoptic map. 



3. A sharp change in pressure always brings about a sharp change in the weather, and in 

 particular, a change in direction and velocity of wind. I have conjectured that we may assume that 

 a pressure change at any geographical point is an index of variability of weather, or in other words 

 an index of atmospheric turbulence. 



Let us suppose that on the barograph record at moment t^ we have a pressure p j^, and at 

 moment t2 a pressure P2- Let us call a the straightline distance between points on the barograph 

 tape defined by the coordinates t - and p, and the coordinates 1 2 3^*^ P2 • ^^^ ^^ suppose now that 

 the atmospheric pressure at the given point of the earth's surface has changed from moment t^ to 

 moment to not in a straight line, but in a more or less complex curve. Using a curve measuring 

 instrument, let us take from the barograph record the length of this curve in the same scale as 

 straight line a , and call it b . The ratio b/a = k will be the index of atmospheric turbulence** at the 

 given point for the given time interval. Obviously this index is not less than 1. 



*A theory of turbulence of sea in the same sense as we here speak of turbulence of the atmos- 

 phere has not as yet been propounded, but it should by all means be put forward. Here we may note 

 only that the wind-caused turbulence of the sea is in part characterized by atmospheric turbulence 

 while tidal agitation is characterized by the length of the orbit of aparticle for the tidal, phase, etc. 



**(Editor's Note--perhaps "index of large-scale variability inpressure" would be abetter phrase 

 here since turbulence in a meteorological sense refers to the irregular local transitory variations 

 in the general airflow and is manifested as gustiness, bumpiness, updrafts and downdrafts). 



430 



