Diurnal Weather Patterns—-L eopold 
THE TEMPERATURE INVERSION AND ITS 
DIURNAL CHANGE 
As an example of a summertime sounding, 
the September 14, 1946, radiosonde flight of 
1730 LST (0400 Greenwich time) is plotted in 
Figure 5. The base of the subsidence inversion, 
marked on the graph, was at an elevation of 
5,380 feet above mean sea level. The tempera¬ 
ture increased 1.6°C. through the inversion 
which is exactly the mean magnitude for 44 
soundings in the summer of 1947. Together 
with the accompanying rapid decrease in mois¬ 
ture at or near the same level, the inversion is 
apparently quite sufficient to limit the top of 
trade-wind clouds. Although direct correlation 
of cloud tops with the temperature inversion 
has not been made locally by means of airplane- 
meteorograph soundings, it is common observa¬ 
tion that the low-cloud tops in the area are at 
quite a uniform level over the ocean even when 
they are cumuliform in character. Moreover, the 
20 
15 
1500 FT. 
1 
1 u 
20 
15 
i 
a: 
* , 0 
1000 FT 
£ 20 
UJ 
a 
*0 
15 
Q 
Z 
> 1 o 
i— ' * 
500 FT. 
. . J|, 
20 
15 
10 
i 5 
> 
SURFACE 
/ 
< 
) 6 12 18 24 
HOUR 
Fig. 6. Wind speed at Honolulu—surface and 
aloft. Averages from pilot balloons for August, 1946. 
91 
writers own observations indicate that the 
height of cloud tops over the mountains is not 
much different than the general level of the 
cloud tops over the open ocean. The writer 
made a rough check with crude thermometers 
on the slope of Haleakala, the 10,000-foot 
mountain of East Maui. On the road to the 
summit, an inversion of temperature of 3°C. 
was easily distinguished, and it represented very 
closely indeed the cloud tops prevailing at the 
time. 
Until more careful measurements are made, 
it may be assumed that, as in the well-authen¬ 
ticated situation in Southern California (Nei- 
burger, Beer, and Leopold, 1945), the tempera¬ 
ture inversion is the top of the low cloud deck, 
and that the additional height of orographic 
clouds over the mountains is not of a large 
magnitude. 
The co-operation of the.U. S. Weather Bureau 
was solicited to obtain additional upper air data 
by special radiosonde flights. During a 2-day 
period, June 26 and 27, 1947, four extra flights 
were made, which, in addition to the regularly 
scheduled radiosondes, provided one ascent 
every 6 hours. The heights of isotherms and the 
inversion for the 2 days of special ascents are 
shown in Figure 7 plotted as a time-height 
cross section. On these particular days the sub¬ 
sidence inversion did not continue unbroken 
through the night, but as shown to be common 
in Southern California, disappeared at one level 
and reformed nearly simultaneously at another. 
The nearly isothermal layer at 7,500 feet, 1800 
LST June 26, already indicated the beginning 
of the inversion which was well established at 
7,700 feet at 0530 the following day. On the 
26th the inversion base reached its maximum 
height at 1030, and on the 27th at about 1500. 
The height of the inversion base for these 2 
days has been replotted on Figure 4C. The 
mean heights of the inversion base for the 
months of August, September, and November, 
1946, are plotted on the same Figure 4C at 
scheduled radiosonde times, 0530 and 1730 LST. 
