ISENTROPIC ANALYSIS 
139 
the extreme south the velocities are 
hardly accelerated at all. Therefore, 
it is highly probable that lateral 
shearing stresses may be important in 
producing cross-isobar wind compo- 
nents, and thus also pressure varia- 
tions which are of a purely dynamic 
rather than thermal-advective nature. 
There has been accumulated an in- 
creasing mass of evidence, both on 
theoretical and observational grounds, 
that the magnitude of these lateral 
shearing stresses is sufficient to ac- 
count for such pressure variations. 
Thus by equating isentropic shearing 
stress within sloping parallel isen- 
tropic surfaces which intersect the 
ground at a normal angle and stress 
due to ground friction (and assigning 
reasonable values of lapse rate and 
wind velocity in temperate latitudes) 
Rossby [8] obtained a lateral stress 
of 167 dynes/cm’ corresponding to an 
isentropic shear of 2.5 X 10—’ sec— 
and an isentropic eddy viscosity of 
6.7 X 10° grams/em/sec. These val- 
ues agree fairly well in order of mag- 
nitude with those obtained from a 
study of the diffusion of water vapor 
actually observed in isentropic charts 
(Grimminger [9]) ; moreover, they lie 
between those found by Richardson 
and Proctor for diffusion over small 
distances and those obtained by 
Defant by considering cyclones and 
anticyclones as large-scale turbulent 
elements in the general circulation. 
The analysis of isentropic charts has 
shown that eddies of appreciably 
smaller size than extra-tropical cy- 
clones and anticyclones occur almost 
daily in the isentropic flow patterns. 
There is being gathered an increas- 
ing mass of observational evidence to 
substantiate the theory summarized 
above®. In the following pages we 
shall briefly discuss the practical re- 
sults of studies of some of this ma- 
terial. 
Before proceeding with this discus- 
sion it is well to repeat that the isen- 
tropic analysis suggests itself as a 
practical tool in synoptic meteorology 
for the following two reasons: 
1. It provides a method of identify- 
ing and following large-scale moist 
and dry currents and of anticipating 
their subsequent thermodynamic modi- 
fications. 
2. It provides a method for taking 
into consideration lateral shearing 
stresses and their hydrodynamical 
effects on the prevailing flow pattern. 
§ 2. PLOTTING ROUTINE 
The mechanical operation of pre- 
paring upper-air data in a form suit- 
able for isentropic analysis is quite 
simple. We require for the analysis 
two sets of upper-air charts: isen- 
tropic charts, for which an ordinary 
geographical base map suffices, and 
atmospheric cross sections which repre- 
sent vertical planes cutting through 
the atmosphere along desired lines. The 
choice of some convenient isentropic 
surface or surfaces for analysis must 
then be made. The deciding factor in 
this regard is the general elevation 
of the surface, which again depends 
upon the general temperature distrib- 
ution. Thus the isentropic surface 
§ = 290° which has been found useful 
in the United States during the win- 
ter months is far too low during the 
summer months. In general, the isen- 
tropic sheet finally chosen should be 
high enough to be above the layer in- 
fluenced by surface friction, and at 
the same time, low enough to show a 
large range of specific-humidity val- 
ues. It is obvious that over a large 
area of diverse topographical and 
climatic features these two criteria 
3A complete quantitative treatment of the 
underlying theory is at present in prepara- 
tion by C.-G. Rossby and his colleagues. 
