94 
PACIFIC SCIENCE, Voi. II, April, 1948 
what at night and are directly out of phase with 
inversion heights, as can be seen by comparison 
of Figure 4C and Figure 6. Though the aver¬ 
age increase is small, it appeared in September 
data as well as in the August data presented in 
the figure. That these changes of wind speed 
are related to the lower inversion at night is in¬ 
dicated by a simple calculation of hydraulics. 
Using the shape of the curve of Figure 4C and 
the plotted points showing mean heights of the 
inversion base for individual months, the maxi¬ 
mum and minimum heights of the inversion 
were estimated for August and September, 1946. 
The diurnal curves of wind speed at various 
levels aloft presented in Figure 6 for August, 
1946, were similarly computed for September, 
1946. 
Assuming no compression, the mean wind 
speed below the inversion should increase as the 
inversion decreases in height if energy is to be 
conserved. The ratio of wind speeds should be 
the same as the ratio of inversion heights if the 
inversion is a surface through which parcels of 
air do not pass. Table 3 presents these ratios. 
The diurnal curves of wind speed could not 
be drawn for elevations above 2,000 feet be¬ 
cause clouds limited the height of pilot-balloon 
observations. Nevertheless, the ratios are suffi¬ 
ciently close, considering the limitations of the 
data, to indicate that the diurnal variations of 
wind speeds aloft are the result of diurnal 
changes of inversion height as would be ex¬ 
pected from theoretical considerations. 
Loveridge (1924) and Jones (1939) attrib¬ 
ute the nocturnal rainfall at Honolulu to the 
radiative cooling at the top of the clouds. 
Soundings from Honolulu Airport are too far 
from the mountain crest to be representative of 
conditions in the orographic clouds. However, 
some nocturnal cooling at all levels seen in the 
Honolulu soundings is probably also true Over 
the mountains with additional cooling near 
cloud tops. Cooling at all levels implies a lower 
lifting condensation level or a lower cloud base 
at night than in the daytime. This is verified 
by observation. Higher nocturnal wind speeds 
aloft probably mean more turbulence and larger 
droplet size. All these factors would tend to 
provide a nocturnal maximum of rainfall. 
In so far as the city of Honolulu is concerned, 
many rain showers result from the blowing of 
rain droplets considerably leeward of the edge 
of the cloud producing them. Higher wind 
speeds would again tend to a nocturnal rainfall 
maximum in the city. 
SUMMARY 
The importance of convective shower activity 
in areas leeward of the main zone of orographic 
rainfall has not hitherto been brought out. 
Afternoon maxima of rainfall are observed in 
the center of the leeward plateau of Lanai and 
along the west edge of the Wahiawa saddle of 
Oahu. This implies that convective showers are 
an important source of moisture in many of the 
drier parts of the islands where only a moderate 
part of the moisture has been dropped from the 
air as a result of prior orographic lifting. It is 
likely that too much moisture has been extracted 
to give many local convective showers in the 
Lualualei area, though no gages were available 
there for analysis. Such areas must depend on 
TABLE 3. RATIOS OF MAXIMUM/MINIMUM INVERSION HEIGHTS AND RATIOS OF WIND 
SPEEDS, HONOLULU 
MAXIMUM AND 
RATIO MAXIMUM/ 
RATIO OF MAXIMUM/ 
PERIOD 
MINIMUM INVERSION 
MINIMUM INVERSION 
MINIMUM WIND SPEEDS 
HEIGHT IN FEET 
HEIGHT 
AVERAGE 500-2,000 FEET 
August, 1946 .... 
; 
o o 
o o 
1.39 
1.23 
September, 1946 . . . 
f 8,000 l 
l 6,000 j 
1.33 
1.44 
