SHORT-RANGE WEATHER FORECASTING 
It is beyond the purview of this article to describe 
in any detail how forecasting should be organized, but 
all forecast organizations should provide for the follow- 
ing: 
1. An adequate method of selecting forecasters. 
Demonstrated interest in forecasting, ability shown in 
a practice forecast program, temperament and general 
intelligence, among other factors, should form the basis 
for selection. 
2. Verification of forecasts made by regular fore- 
casters. The weather service should have some method 
of ascertaining whether proper standards are being 
maintained, and whether the accuracy curve of indi- 
vidual forecasters is rising or falling or has leveled off, 
as well as some method of determining the comparative 
skill of individual forecasters. 
3. Satisfactory environment for forecasting. Ade- 
quate forecast performance requires a high level of 
concentration not possible in congested and noisy 
offices. Provision should be made for rapid preparation 
or receipt by facsimile of needed charts and maps, 
their efficient display, and the arrangement of other 
references and aids where they may be located and 
utilized quickly. 
4. Sufficient time for the proper analysis of data, 
application of techniques, reconciliation of conflicting 
inferences from the several charts and diagrams, and 
for the preparation and wording of forecasts. An hour 
and a half is estimated as the irreducible mmimum 
required on the average for the various forecast steps 
after completion of the six-hourly synoptic chart, and 
an additional sixty to ninety minutes is needed for the 
preparation of forecasts and warnings; communication 
schedules should be arranged accordingly. 
5. A forecast staff large enough to permit a forecast 
schedule with time for research, study, and relief from 
forecasting. Forecasting requires energy, enthusiasm, 
and imagination and the work should never be per- 
mitted to become routine or the forecaster to become 
stale. 
6. Freedom from interruption and from administra- 
tive and service work when on forecast duty. Although 
the forecaster is involved in problems of extreme com- 
plexity, he frequently must operate under conditions 
which permit constant interruption from phone calls 
and pilot briefing. Work performed under these con- 
ditions must inescapably be of inferior quality. Local 
forecasting can, of course, be combined with briefing 
and public service work. 
7. Proper scheduling of forecasters. There are still 
differences of opinion between and within forecast serv- 
ices with respect to the type of schedule which will 
result in forecasts of the best quality and continuity 
consistent with proper health safeguards for the fore- 
caster. Tests should be feasible. 
8. Testing of suggestions, new ideas, and techniques 
for the improvement of forecasting. Each weather 
service should have one forecast center adequately 
equipped and staffed for testing suggested new tech- 
niques in actual forecasting. 
763 
NEEDED RESEARCH IN SHORT-RANGE 
FORECASTING 
Some sixteen forecast units in the United States 
responded to an invitation from the author to indicate 
the areas of research of greatest urgency in short-range 
weather forecasting. The replies covered every phase of 
the whole problem of forecasting and reflected a general 
admission of the inadequacy of current techniques. 
There was general agreement that research should pro- 
ceed along two main lines: (1) toward the discovery of 
fundamental truths about the atmosphere, with em- 
phasis on the general circulation, and (2) toward the 
development of forecast techniques which would lead 
to improved forecasts in the immediate future. 
There is little doubt that the first is the more impor- 
tant, and that the second should be given a secondary 
role. Certaily the yield from the large amount of 
research on practical forecasting during the past decade 
or two has been exceedingly meager. 
Suggested topics for further research included: 
1. The general circulation of the atmosphere. 
Although many forecasters feel that the large propor- 
tion of research at the meteorological departments of 
universities which is focused on the broad problems of 
the general circulation of the atmosphere has little 
immediate and practical application to the problems of 
forecasting, it seems likely that any marked improve- 
ment in forecasting must come from this quarter. One 
may hope that discoveries will eventually be made 
that will revolutionize current forecast methods, but, 
even earlier, it seems probable that new knowledge of 
the general circulation would permit the preparation of 
better prognostic upper-air charts and other forecast 
aids. Continued intense research on the general circu- 
lation should have the highest priority. 
2. Aperiodie departures from the normal positions 
of centers of action and mean troughs and ridges. Of 
paramount importance to short-, medium-, and long- 
range forecasting are the causes of the departures of 
the centers of action from their normal location and 
intensity—anomalies which may persist for a matter 
of months. Of equal moment are the causes of the vari- 
ations of the zonal indices, which in turn permit, at 
times, excessive zonal or meridional flow; it is likewise 
desirable to learn how these variations can be forecast. 
3. The qualitative and quantitative forecasting of 
precipitation. All forecast units listed this forecast 
problem, and the majority described it either as the 
most difficult or as the most urgent. This problem in- 
cludes: 
a. Determining when and where precipitation will 
break out in a previously dry trough. 
b. Accurately timing the beginning of precipitation 
12-24 hr in advance. 
c. Forecasting the nocturnal type of thunderstorm. 
d. The degree of concentration of the convective 
afternoon thundershower. 
e. The development of “‘bursts” of rainfall, imsta- 
bility lines (squall lines), and other zones of conver- 
gence. 
