A PROCEDURE OF SHORT-RANGE WEATHER FORECASTING 
an “epignosis” of prognostic maps. When the diag- 
nostic-analytic task is finished, the analyst centers his 
attention upon the prognostic upper-air maps, maps II, 
prepared at the time ¢) — 12" and referring to t) + 125; 
these maps are readjusted, both by performing the first 
stage of prognosis upon maps I and by using the curves 
already drawn and only checking and correcting them 
in the light of the new observations now available. 
Finally, using maps I and II, the prognostic maps, 
maps III, for the time tf) + 245 are prepared. They form 
the basis for the forecast. At the next synoptic hour 
the analyst then starts with the maps II (prepared at 
time f) — 12" but readjusted at time to), on which the 
to + 12" observations have been plotted. 
The synoptic cycle described above not only im- 
proves the analysis and prognosis of the weather maps, 
but also shows the analyst to what extent the prognostic 
extrapolation has been successful in any particular 
weather situation. Thus, it reveals the influences that 
have been totally or partly neglected during the prog- 
nosis. In particular, we may hope that the application 
of the synoptic cycle may yield valuable information as 
to the nonadvective effects in the upper air. 
For practical purposes it is necessary to employ 
several meteorologists for the simultaneous prognosis 
of the system of surface and upper-air maps; a certain 
subdivision of work thus becomes necessary. However, 
a strict partitioning should not be maintained in prac- 
tice between the different parts of weather prognosis, 
for imstance by dividing (as is often done) a weather 
analysis center into a “‘surface” section and an “‘upper- 
air” section. Instead, the upper-air prognosis should be 
regarded as part of a single procedure involving the 
surface-map prognosis. The forecaster continually im- 
proves the surface-map prognosis from the results of 
the upper-air analysis and prognosis. On the other 
hand, he builds his system of prognostic upper-air maps 
on the prognostic surface map and must continually 
adjust his prognostic upper-air maps so as to profit 
from the recent improvements in the surface-map prog- 
nosis. 
The General Forecast. Over a very large prognostic 
region, the extent of which varies from one weather 
situation to another, a smoothed prognostic map is 
derived from the analyzed weather maps based on repre- 
sentative observations of a more inclusive analytic region 
(see Fig. 1). From this map system is issued the general 
forecast, which—for a forecast period more inclusive than 
the prognostic period of the maps—states in a semi- 
Lagrangian way the probable gross weather develop- 
ment, in the absence of marked local influences, for an 
entire administratively fixed forecast region situated 
within the central part of the prognostic region. In 
particular, the general forecast must tell specifically 
the what and where of the following weather phenomena, 
including their present state and position as well as 
their motion and development during the forecast 
period: (1) the main air currents, (2) the main cyclones 
and anticyclones, (3) the main air masses and fronts, 
(4) the main areas of cloud and precipitation that may 
occur within an air mass and at a front, (5) the regions 
767 
of especially strong wind, and (6) the general temper- 
ature distribution. Valid mainly for the synoptically 
representative stations, the general forecast is primarily 
of interest only to the weather service and for the 
strategic planning of region-wide activities such as 
communication, transportation, and aviation. 
The Area Forecast. The general forecast, which is 
quite insufficient for a specified smaller forecast area 
within the forecast region, must be amplified, detailed, 
and even supplemented by an area forecast which— 
although based on the maps and the general forecast 
for the region—takes into consideration important local 
effects that were purposefully neglected during the 
regional analysis and prognosis. At the area forecasting 
centers, the general analysis, prognosis, and forecast 
are received at 12-hr intervals from the regional analysis 
center by teletype and/or facsimile machines. In order 
to feel the pulse of the locally influenced present and 
LEGEND 
BOUNDARY FOR ANALYTIC-REGION ———BOUNDARIES BETWEEN ADJACENT 
SHORT-RANGE FORECAST DISTRICTS 
© GENERAL ANALYTIC CENTER 
@ AREA FORECAST CENTERS 
e LOCAL FOREGAST CENTERS 
— — BOUNDARY FOR PROGNOSTIC REGION 
BOUNDARY FOR FORECAST REGION 
BOUNDARIES BETWEEN ADJACENT 
FORECAST AREAS 
Fig. 1.—Analytic, prognostic, and forecast regions for the 
United States with its areas, districts, and localities. 
past weather of his area, the area forecaster may often 
be obliged to prepare, mostly on a scale greater than 
that of the regional map system, his own sequences of 
more frequent and detailed maps (special maps, or even 
surface maps limited to the extent of his analytic and 
prognostic area). Being based upon the synoptically 
unrepresentative observations of a dense area network 
of primitive auxiliary stations which do not ordinarily 
belong to the international network, the area analysis 
forms a supplement to, not a duplication of, the regional 
map analysis. Moreover, the area forecaster is able, by 
the use of a network denser in time as well as in space 
and from more recent observational data, to fix the 
position of fronts, isobars, etc., in his area with greater 
accuracy than is possible at the regional analysis center. 
To further both these purposes he also, if residing within 
his area, makes use of such supplementary detailed 
“unwritten”? observations as those he collects verbally 
from pilots and other contacts using the so-called local 
