THUNDERSTORMS 61 
(c) Associated with an _ oc- 
cluded front. 
3. Orographic thunderstorms. 
4. Thunderstorms in horizontally 
converging air currents. 
I. Arr Mass THUNDERSTORMS 
A. From local convection. The type 
of thunderstorm known for a long 
time as “local,” “heat,” or “local 
convective” may be singled out from 
other types in that it represents a 
form of penetrative convection due 
primarily to insolational heating in 
the lower layers of the atmosphere. 
No apparent frontal activity is in- 
volved, and the original cause appears 
to be purely thermal in nature. In 
summer these thunderstorms may be 
observed in almost every type of air 
mass. They are, however, by far 
most frequent in TG air, because the 
TG air has the most favorable moisi- 
ure and temperature distribution 
with elevation. Steep lapse-rate and 
high specific humidity are most favor- 
able for the development of Cu clouds 
into tall Cb. The tephigram under 
such conditions generally exhibits a 
large positive area and a compara- 
tively small negative area. As pointed 
out in the preceding article, it is im- 
portant to use critical judgment in 
selecting which particle of air will be 
assumed to penetrate the upper 
strata; furthermore, it is necessary to 
assume the temperature and moisture . 
that particle will probably have be- 
fore it ascends. Thus if one should 
choose the surface air particle having 
the temperature and moisture ob- 
served in the early morning hours 
he would, owing to the stability of 
the ground inversion, always obtain 
tremendous negative areas. In fore- 
casting a maximum temperature for 
the surface particle one should be 
familiar with the locality for which 
he is forecasting. It is common prac- 
tice to take for the maximum tem- 
perature of the day the surface tem- 
perature which would be potentially 
equivalent to that at the top of the 
ground inversion. The physical reas- 
oning behind this procedure is that as 
insolation warms the ground the 
stability of the overlying inversion 
precludes any appreciable upward 
transfer of the heat supplied the 
surface layer. Consequently, the 
lowest layer warms up rapidly until 
the ground inversion with its stability 
is completely wiped out. After this 
takes place most of the excess heat 
is carried upward by convection, and 
the temperature at the surface re- 
mains at a fairly constant maximum. 
In choosing the specific humidity of 
the particular element which is to 
rise through the surrounding air mass 
it is customary to assume that the 
specific humidity in the _ surface 
layers remains constant. This is 
logical because, exclusive of evapora- 
tion from the surface and mixing with 
the air above, the mass of water 
vapor per unit mass of air should 
remain sensibly constant in one and 
the same air mass; in a case where 
the specific humidity decreases 
rapidly upward from the surface, the 
mixing by convection and mechanical 
turbulence will lower the specific hu- 
midity of the surface air. In this 
case it would be erroneous to pre- 
sume a constancy cf moisture content 
during the day. In practice I have 
found it helpful in these cases to 
choose for the probable specific hu- 
midity of the rising particle the mean 
value of the observed specific humidi- 
ties through the lowest layers up to 
the condensation level, or to the top 
of any ground inversion or stable 
layer having relatively high humidity. 
A not uncommon case _ presents 
itself when above the moist surface 
air, presumably Tc, there flows a 
