A PROCEDURE OF SHORT-RANGE WEATHER FORECASTING 
if the curvature of its tendency profile is anticyclonic 
(cyclonic). An anticyclonic center experiences an ac- 
celeration (deceleration) if the curvature of the tend- 
ency profile is cyclonic (anticyclonic). Often a depres- 
sion with very strong winds on its forward side tends 
to become stationary and rapidly disappears. 
The foregoing extrapolation of the motion of the 
sea-level pressure systems is based largely upon a rec- 
ognition of certain events which had been observed to 
occur within the system in question (at the surface 
of the earth). However, the well-known steering “‘prin- 
ciple” or hypothesis had early been advanced that the 
ageostrophic upper flow, and the accompanying centers 
of change in the sea-level pressure, propagate with 
the real upper flow at a certain pressure surface, called 
the steering surface. Strictly speaking, this principle 
applies only to rises and falls, respectively, and not to 
the highs and lows. As a preliminary remark we cau- 
tion that pressure changes represent processes and 
therefore are liable to undergo modifications. They 
are not systems passively drifting with the flow. Even 
so, the movement of the sea-level systems is assumed 
to be closely associated with the pressure field at the 
steering surface, that is, with certain events observed 
outside the systems on the surface map. The following 
historical and critical account of steering—introduced 
in 1910—considers the Scherhag complex and culmi- 
nates in a discussion of Sutcliffe’s practicable rules of 
thermal steering. 
In 1931—three years before daily upper-air maps had 
been introduced in Germany—Miigge [53] had already 
formulated the loose rule that the rises and falls follow 
downstream the 500-mb isohypses, the speed of their prop- 
agation being roughly half the 500-mb gradient wind. The 
3-hr rises and falls seem to move in the direction of 
the 700-mb flow, while the 24-hr rises and falls move 
with the 500-mb flow. The movement of rises and 
falls, which are associated with the young and shallow 
pressure disturbances (highs, lows, troughs, and ridges), 
is very roughly determined by Miuigge’s empirical rela- 
tion so long as the development of the highs and lows 
in the hypsography of the steering surface is negligible. 
The falls have a tendency to follow the curved 500-mb 
isohypses better than the cuspal 500-mb isohypses, 
which the falls tend to cut across. On the other hand, 
the rises are inclined to follow the cyclonically curved 
500-mb isohypses and to depart from those curved 
anticyclonically. By their thermal influence in trans- 
forming the upper pressure field, the falls swerve to 
the left, the rises to the right of the established upper 
flow. The rises and falls of the older and more de- 
veloped disturbances are observed to move in the mean 
or predominant direction of the wave-shaped flow at 
about 400 mb to 300 mb. In this connection, constant- 
pressure maps higher than 500 mb should be consulted 
to select the steering surface with relatively straight 
absolute isohypses in the regions of the older and 
deeper disturbances. Pressure systems with closed iso- 
hypses up to the stratosphere, such as an occluded 
cyclone or a blocking high, are themselves considered 
steering centers, so that there is no well-defined rela- 
777 
tion between the 500-mb flow and the movement of 
their rises and falls. If a closed upper low is formed 
through a strong pressure fall at the surface of the 
earth, the fall tends to split, with the rapidly weaken- 
ing part swerving to the left, the stronger to the right, 
of the old path (occlusion stage of development). The 
cold lows in the upper relative hypsographies 
(Kaltlufttropfen) are not steered by the upper flow, 
but their movement can be predicted from other con- 
siderations (see Fig. 6). 
For preparing the prognostic map of 24-hr change in 
sea-level pressure, the following empirical relation was 
first suggested by Guilbert and Grossman and later 
formulated by Rodewald [66]. The 24-hr fall (rise) for 
to moves downstream wn the direction of the 500-mb iso- 
hypses for ty so that its position at tp) + 24" is on the 
axis of past 24-hr increase (decrease) in the 500/1000- 
mb hypsography’ for to. The prognostic sea-level pres- 
sure map for fo + 24" can be immediately obtained 
by simply adding graphically the prognostic map of 
24-hr change in sea-level pressure (obtained approxi- 
mately by the Guilbert-Grossman rule) to the sea- 
level pressure map for fo. 
In 1939 Ertel introduced his theory of singular ad- 
vection of the first order [26], according to which 
atmospheric pressure variations originate at discon- 
tinuities where changes of momentum occur. The best- 
defined and most persistent temperature discontinuity 
surface of the first order is the tropopause. By dis- 
regarding all other discontinuity surfaces (viz., fronts), 
Lucht [50] has attempted to adapt and apply Hrtel’s 
theory at the tropopause in order to find how the upper 
atmosphere is associated with variations in the sea- 
level pressure. Essentially, the adapted equation of 
Ertel gives the individual sea-level pressure changes as 
a quantity depending on (1) the solenoidal field of the 
mean temperature of the troposphere, (2) the hypsog- 
raphy of the tropopause, and (3) the stability dis- 
continuity of the tropopause. The tropopause discon- 
tinuity of the temperature gradient and inclination of 
the tropopause are apparently the most influential 
factors in deciding the speed of movement. The con- 
clusions from Lucht’s investigation mdicate, for the 
rises and falls, that the direction of motion can be 
predicted from the hypsography of the tropopause and 
that Ertel’s equation [26, eq. 50, p. 406] gives satis- 
factory values for the speed of motion. But Lucht’s 
efforts to apply Ertel’s equation are not persuasive, 
because of the sparsity of the data in 1939 for deter- 
mining the slope of the tropopause and the permanent 
difficulties in determining its temperature discontinu- 
ity. Lucht has thus attempted to substantiate and 
refine the observations made already in 1926 by Sttive 
and Palmén [56] and by von Ficker [29] that the rises 
and falls move in the direction of the isohypses of the 
tropopause. 
In an effort to extend the steering hypothesis to 
lows, Austin [5] has observed that the steered lows (1) 
7. Defined in Table I. 
8. Variously stated in [9, 66). 
