THE FORECAST PROBLEM 
and with the hydrometeors, in an attempt to find a 
numerical basis of extrapolation for forecasting of the 
short-range or daily types. The success of all such 
attempts has been notably indifferent, to the extent 
that none of them has found general quantitative appli- 
cation in practical daily forecasting. This lack of practi- 
cal application is caused partly by the large amount of 
work involved and partly by the lack of encouraging 
results thus far obtained. This failure to obtain good 
results doubtless stems from the same difficulties which 
cause the failure of the more elaborate computational 
techniques: notably insufficient quantity and accuracy 
of synoptic observations, difficulty in smoothing out 
local disturbances, and insufficient physical understand- 
ing of the atmospheric processes. For these same rea- 
sons, equally little can be hoped from these efforts at 
present towards the solution of the forecast problem. 
Rossby’s vorticity technique has been applied to both 
the daily and the extended forecasting of flow patterns 
in the middle troposphere with enough demonstrable 
success so that it has received some routine application, 
notably by the Extended Forecasting Section of the 
U. S. Weather Bureau [6]. It has proved useful as an © 
auxiliary forecasting tool under certain conditions, but 
the limitations which are imposed upon it, in part by 
the restriction of its use to clearly defined wave-patterns 
aloft, and probably more fundamentally by the neces- 
sary assumption of horizontal (nondivergent) air flow, 
are in themselves sufficient guarantee that this method 
cannot contribute significantly to the solution of the 
basic forecast problem. 
2. Physical Techniques of Forecasting the Weather— 
based on a consideration of the physical factors which 
may actively modify the existing state and trend of the 
weather as it progresses. It was mentioned above under 
the discussion of synoptic extrapolation techniques that 
the usual short-range and daily forecasting procedure 
entails the use of a few physical considerations at least 
as a supplement to the purely synoptic extrapolation. 
In the same manner, certain physical factors frequently 
form the basis of some of the statistical extrapolation 
techniques in extended or long-range forecasting. 
In short-range and daily forecasting the most effective 
use of physical considerations to supplement or modify 
the synoptic extrapolation techniques has developed 
from the Norwegian polar front theory, or as more 
commonly designated now, from air-mass and frontal 
analysis. The study of the physical properties of air 
masses, in particular the vertical distribution of moist- 
ure and the vertical stability, together with the physical 
factors which modify them, has proved most useful for 
the type of detailed local forecasting which is so impor- 
tant to aviation. The use of energy diagrams such as the 
adiabatic diagram, the tephigram, or the aerogram to 
indicate the physical probability of local convection, and 
the consideration of topographic influences, such as 
upslope or downslope motion, surface heating or cooling 
from land and water surfaces to change the air-mass 
properties, and the turbulence and mixing character- 
istics produced by wind and rough terrain—all of these 
739 
are physically modifying influences which are normally 
considered in routine short-range forecast practice. 
Likewise the physical concepts of frontogenesis and 
the accumulation of available potential energy supplied 
by air-mass convergence, of cyclogenesis by the develop- 
ment and occlusion of wave disturbances with all of the 
attendant weather cycle and sequences of hydrometeors, 
of the modification of frontal structure and condensation 
forms by orographic barriers and coast lines, all of these 
are physically modifying influences, the consideration 
of which has added much to short-range and daily 
forecasting. 
However, the prognostic potentialities of air-mass 
and frontal analysis and related physical techniques of 
forecasting appear to have been largely realized. Prob- 
ably some further refinement of forecasting detail may 
be effected along these lines, but no radical advance can 
be hoped for from this quarter. In fact, 1t can be asserted 
that the contribution of frontal and air-mass analysis 
to scientific weather forecasting has fallen far short of 
that which was hoped for in the early days of the 
development of this new school. The reason for the 
failure of frontal and air-mass concepts to solve more 
weather problems probably lies to a considerable extent 
in the following observational or hypothetical facts: 
a. The utter complexity of atmospheric conditions 
which are not even approximately represented by such 
concepts as homogeneous air masses and frontal dis- 
continuities. 
b. The fact that cyclogenesis as a process probably 
rarely if ever closely approximates the ideal Bjerknes 
wave-cyclone model. 
c. The fact that the primary impulse or drive of large- 
scale cyclogenesis and anticyclogenesis frequently origi- 
nates far outside of the developing center, hence cannot 
be identified or anticipated by local conditions. 
The physical factors which have been assumed to 
exert either modifying or controlling influence over the 
extended or long-period anomalous fluctuations of the 
large-scale weather patterns, and hence to require either 
secondary or primary consideration in extended or 
long-range forecasting, may be classified as follows: 
a. Continental (orographic barriers, coast lines, and 
extensive snow cover). 
b. Oceanographic (anomalous surface current flow, 
temperature, and polar ice conditions). 
c. Extraterrestrial (sun, moon, and planets). 
Of the continental factors mentioned, obviously the 
topographic features vary only during geological time, 
hence their influence is expressed in the seasonal nor- 
mal patterns. It is only as the circulation pattern is 
markedly anomalous by reason of some primary dis- 
turbing factor that topographic influences find expres- 
sion as secondary anomalous characteristics of the 
circulation pattern, notably in the windward and lee 
effects of a major orographic barrier on an abnormally 
strong cross-wind system. Likewise an extensively 
anomalous condition of continental snow cover doubt- 
less contributes to a minor degree to the persistence or 
intensity of an anomalous weather pattern, but much 
statistical analysis has failed to establish any signifi- 
