THUNDERSTORMS 
ment, for movement is in part comprised of propagation 
which in turn results from new growths or extensions. 
In the preceding section, measurements were described 
in which the propagation and dissipation components 
were removed from the thunderstorm motions. When 
this is not done erratic motions are indicated. At the 
end of the section on thunderstorm weather near the 
surface it was shown how certain areas immediately 
surrounding existing cells are favorable for the forma- 
tion of new ones. These frequently become attached to 
form a larger thunderstorm mass. In mountainous areas 
the propagation components in the motion usually out- 
weigh the wind components. Thus thunderstorms are 
seen to propagate along a mountain range even though 
the wind is blowing across the range at all levels, and 
sometimes there will be few or no cells drifting into 
the valley so that the wind component in the motion 
is relatively ineffective. 
Certain features of the earth’s surface provide an 
ideal location for new thunderstorm development. 
Mountains and, in fact, rugged relief of any kind as 
contrasted with smooth terrain will be favorable. 
Islands, peninsulas, and other pronounced heat sources 
are favorable for the formation of afternoon storms. 
In Florida it is found that the convergence of the sea 
air over the peninsula from both sides in the afternoon 
is favorable for thunderstorm formation [10]. Locally 
on the peninsula the dry land areas as contrasted with 
the numerous swamps and lakes are most favorable for 
new formation. 
In Ohio, a study of the initial appearance of 584 
radar clouds on 21 days showed a larger number of 
echoes from day to day over the rougher parts of the 
terrain. However, plenty of new echoes formed over 
the flattest, smoothest areas. Neither in Ohio nor in 
Florida was it possible to find a well-marked region 
from which the first thunderstorm of a series would 
develop. The thunderstorm hearth (Gewitterherde) de- 
seribed by von Ficker [13] was not apparent. 
In many parts of the world popular notions have 
developed concerning an apparent tendency for thun- 
derstorms to follow rivers. The idea is often expressed 
that the moisture provided by the river favors propaga- 
tion along the watercourse. Scientific studies indicate 
that, where thunderstorms propagate along rivers, con- 
ditions of the terrain rather than the presence of the 
water are the factors which are favorable for new 
growth. Many rivers have bluffs along their flanks and 
the immediately surrounding terrain is roughly dis- 
sected by tributary streams. In Ohio the maximum 
number of new formations occurred over the sharply 
dissected plains. The area studied is predominantly a 
peneplain at about 1000 ft through which streams cut 
down to the Ohio River (about 500 ft at Cincinnati). 
Squall Lines 
During the 127-day period from May 17 to September 
21, 1947, there were in Ohio 56 thunderstorm days on 
which extensive radar data were obtained. On 32 of 
these days lines of thunderstorms were observed. The 
lines on 6 days occurred along surface fronts, on 19 
689 
days were ahead of surface cold fronts, and on the re- 
maining 7 days apparently had no connection with the 
surface fronts. As is usual throughout the eastern United 
States, the pre-cold-front squall line was the predom- 
inating scene of thunderstorm activity. 
It was found that the pre-cold-front squall line is not 
quite parallel to the cold front but has an orientation 
averaging 13 degrees clockwise from that of the front. 
Its orientation is somewhat counterclockwise from the 
orientation of the overlying 700-mb contour lines. The 
movement of the squall line, as distinct from the move- 
ment of the individual storms composing the line, is 
not very closely related to the speed of the following 
cold front; in many cases it moves faster than the 
front. The individual elements of the line, or storms, 
move with the prevailing upper winds as described in 
a preceding section. 
The pre-cold-front squall line on the synoptic chart is 
frequently observed to last for periods as long as twenty- 
four hours, during which it may travel several hundred 
miles. When viewed on the radar, it was found that 
there is usually a zone of convective activity comprising 
several lines and, although the zone may persist for a 
considerable period of time, the lines within it are 
constantly undergoing change—new lines are forming 
and older ones are dissipating. 
The elements in a line range from a single isolated 
echo one to two miles in diameter to a large aggregate 
of storms appearing on the scope as a single, more or 
less homogeneous echo having a maximum dimension 
of over thirty miles. The number of separate elements 
in a line varied from 4 or 5 small echoes in the early 
stages to as many as 40 or 50 at the time of maximum 
echo intensity. For the most part, elements were found 
to form and dissipate on a given squall line, suggesting 
the existence of a preferred line formation. 
The squall line or, more properly, the squall-line 
zone was found to be at an average distance of 170 
miles ahead of the cold front, with extreme values of 
80 and 325 miles. Generally there was an area of 
relative inactivity between the cold front and the trail- 
ing edge of the squall zone; but on one or two occasions 
the convective activity finally extended back to the 
front itself. 
Generally, the thunderstorms associated with the 
squall lines were little different from any others except 
for a tendency to be more severe, especially in producing 
effects at the surface, such as strong, gusty winds. Since 
the storms come in a line, there is a more or less con- 
nected discontinuity line from the cold outflowing down- 
draft which is often mistaken for a true front. The 
return to tropical air-mass conditions afterward shows 
the local character of the discontinuity. The pressure 
“dome” of the combined storms often results in the 
creation of a long pressure ridge along the squall line. 
Sometimes a squall line of this character may after a 
time produce a true frontal discontinuity involving more 
cold air than can be accounted for by the downdraft. 
Of the seven squall lines that were not associated 
with fronts, three were related to some type of large- 
scale cyclonic disturbance, two were on the west side 
