1948—1949 Trials of the Schaefer-Langmuir Cloud-
Seeding ‘Technique in Hawaii —
By :
| LSB. LEOPOLD aid W.'A. MORDY | ~
Reprinted from Tellus Volume 3, Number 1
FEBRUARY.1951
1948—1949 Trials of the Schaefer-Langmuir Cloud-Seeding
Technique in, Hawaii*
LUNA B, LEOPOLD and WENDELL A. MORDY
Abstract
Fifteen tests during r948—49 involving thirty-seven individual cloud inoculations over the
jslands of Lanai and Molokai are described, With regard to the rdle of the dry ice in causing the
observed rainfall, the tests are too few in number to be considered conclusive, but the large rain
fall values associated with seeded clouds are of particular interest. The largest rains seem to be
associated with clouds where cloud top temperatures are slightly colder than o° C and such
clouds are also relatively thick. The relative importance of cloud thickness and temperature
cannot be determined from the available data.
There exists at the time of this writing
divergence of opinion concerning the efficacy
of seeding clouds with dry ice for the produc-
tion of rain.
The current status of the problem -was
succinctly stated by Brrczron (1949). “The
question..: is... no longer whether pre-
epinten can be released artificidlly, but rather
whether such a release can produce an appreci-
able amount of rainfall, and when and where
this could be done.”
The carly experiments in Hawaii have
already been reported (Lnorotp and Harsrzap
1948) and the present paper presents the results
of further seeding experiments conducted in
a more or less comparable manner. The
consistent location of seeded clouds from day
to day, the large number of rain gages, and
high elevation of the freezing level make the
Hawaiian experiments of some interest, though
both the earlier and the current sets of data
suffer the disadvantage of no observations
with radar.
In all but one of the trials discussed here, the
dry ice was ground to a size resembling granu-
« Published with the approval of the Director, as
Technical Paper No: 196 of the Pineapple Research
Institute of Hawaii.
lated sugar. The plane was a twin-engined
Beechcraft in which 4oo—s00 pounds of dry
ice were carried on a given flight. The dry ice
was dispersed from a hopper having a sliding
door, and from which 150 pounds of the
granulated material could be dropped out in
a time period of five to seven seconds. Flying
at a speed of r40 mph the seeding rate was
approximately 12.5 pounds per 100 feet of
horizontal flight. The abt towers seeded
were ordinarily of such a width horizontally,
about’ one-quarter mile, that the six seconds
during which the dry ice was dropped was
approximately the flight time across or throu h
the top of the tower. For a seeding run
plane was flown just through or just over the
cloud top.
The clouds seeded were those which occur
over the central part of the Island of Lanai or
over the Maunaloa area of the Western part
of the Island of Molokai. These clouds on test
days were cumulus congestus associated with
the cloud line formed :
tradewind and sea breeze. These clouds have
been described in detail in another paper
(Lzovoxp 1949). The cloud type is important
in considering the results because the con-
vergence in the lower levels where the op-
yy the interaction of
i
THE SCHABPER-LANGMUIR CLOUD-SBEDING TECHNIQUE IN HAWAIL 45
posing winds meet provides a source of ver~
tical motion in the upper cloud which, is not
found in the common, air mass cumuli. It is
possible, then, that these clouds possess the
prolonged “path of fall” of droplets referred
to by Burczron (r949, p. 3). even though
the total cloud thickness seldom exceeded
12,000 feet at the time of seeding.
Tt can be stated with some certainty that
when the cumulus clouds are capped by a tem-
perature inversion of 1°C or more as is the
normal case in Hawaii, seeding with dry ice
either produces a partial dissipation of the
cloud or no noticeable effect; the clouds do
not increase in height and the rain, if any, is
of insignificant amount, This is’ based on a
large number of observations during the
1947—1948 tests including not only tests made
by the authors but many others run independ-
ently by various companies in Hawaii, ‘The
conclusion was sufficiently well established
during the first year that all later tests reported
here were conducted on days when the Hono-
lulu 1500 Z (0500 LST) radiosonde indicated
the absence of the tradewind subsidence in-
version or that the inversion was less than 1° C.
Changes in the strength of the subsidence in-
Version occur rapidly in the Hawaiian area.
On several occasions the osoo LST raob
showed no inversion, on the basis of which a
seeding flight was made, However, by tooo LST,
after the flight had already begun a definite
inversion had formed, as could fe seen by the
concordance of cloud tops and by temperature
measurements taken. in fight.
Our experience, furthermore, indicates that
large horizontal wind shear through the levels
between base and top of cloud tends to blow off
the top of the cloud, and rain will not result.
An initial depression of the cloud top im-
mediately following seeding has been observed
by seatty all experimenters. Experience in
Hawaii indicates that the cloud will recover
from this initial depression and will grow
even higher than its pre-seeding cengition
only if the cloud was actively growing in
height at the time of seeding, Growing clouds
over Hawaii, usually have “harder”, more
well-defined tops than those not in the process
of growing which tend to be wispy-edged.
The wispy edges, the authors believe to be
cither a dissipating phenomenon ot a result of
ice-crystal formation in the clouds,
It is typical for moisture values to be moder-
ately large below the subsidence inversion,
decreasing sharply above, On days when the
inversion is absent, moisture occurs through a
much thicker layer, and experience points
toward large values through a deep layer as a
feature conducive to rain froma seeded cloud.
‘The Hawaiian experiments share with many,
others the difficulty of defining a practical and
definite control against which the results of
seeding can be compared. The tests reported
here were all made on days: of small or no
temperature inversion, On such days, all
clouds in the area tend to be thicker than usual.
The same conditions promote the growth of
the sea-breeze clouds. For these reasons, it is
obvious that the days chosen for test are the
days on which natural rainfall is most likely to
occur on. the test islands, Molokai and Lanai.
On the other hand, it is a fact widely known
in Hawaii that even on such days, when.
unusual build-up of cumulus clouds can. be
seen everywhere, dark threatenin; sea-breeze
clouds remain nearly all day over the Islands of
Molokai, Lanai and certain, other areas, but
seldom does more than a light shower occur.
One reason for this undoubtedly lies in the
fact that such a cloud, whose appearance
suggests imminent rain, often extends nearly
to the freezing level but seldom into it. The
freezing level over Hawaii usually lies between
14,000.and 18,000 feet, while the cloud tops
on days of no inversion generally lie between
10,000 and 13,000 feet. On such a day a few
cumulonimbus usually can be seen somewhere
in the Territory. It appears, then, that these
conditions are unusually favorable for the use
of dry ice as an. agent to start the rain process
which is potentially near but whose unaided
initiation is relatively infrequent.
Result of tests
In the 1948—1949 series, tests made on 6
days over the Island of Lanai and 9 days over
the Island of Molokai include adequate data.
Certain additional tests are omitted because
necessary data are not available or because
circumstances made the conditions of test
sufficiently different from the standard that
the tests cannot be considered comparable.
For example, some clouds were seeded which
lay entirely over the ocean and though rain
46 LUNA B. LEOPOLD and WENDELL A. MORDY
was observed, it could not be measured. In.
certain others, wind and temperature condi-
tions changed so rapidly during the test that
the Honolulu upper-air data were considered
completely unrepresentative of the: test area,
On a given day of test, one to five seeding
runs were made. Data on individual runs, in-
cluding summaries of cloud changes observed,
are ftiladed in Table 3 appended to this report.
The rainfall quantities measured are sum~
marized in Table 1. These data represent the
arithmetic meari rainfall in five gages except
where indicated.*
‘Table 1. Summary of all Hawaiian experiments,
1947-—1949-
Number of Test Days: 27.
Cases of Cloud Tops
Warmer than 0 °C
Cases ‘of Cloud Tops
Colder than 0 °C
No.) Rain at Ground | No, | Rain at Ground
of or of or
Cases} Other Notes _| Cases| Other Notes
1 | x25 1 2.188
1 [0.03 i 1.95?
6 | "Trace x 1.88 M
0.78 L
2 | Virga 1 0.60%
3 | Cloud dissipated 1 0.50
6 | No change observed] 1 02s
0 x Trace
1 | Cloud dissipated
8
1 Cloud went above freezing level after seeding and
during rain.
2 Mean of 5 gages in sceded area; rainfall figures
without superscript are for an individual gage under
seeded cloud.
M Molokai; L Lanai.
Tt will be noted that the rainfall quantities
measured in association with the experiments
reported here are considerably larger than
those reported’ by Orr et al. (1950), Coons
et al. (1948 a) and most other workers. In. the
Lanai, test of June 28, 1948, 1.75 inches fell in
45 minutes at Lanai City at the outer edge of
the cloud. A total of 3.40. inches fell in the
Palawai Basin directly “beneath the seeded
doud. This rain. began 19 minutes following
seeding. It is the largest rainfall on any day in
June in a 45-year record on that island.
© Gages used in this analysis: Lanai gage numbers
671, 672, 674, 681, 683; Molokai gage numbers 513,
$14, $18, $19, 522. Locations indicated by maps of Lo-
porn, Burn, Sripp (1948).
The interpretation of the data shown in the
table will be facilitated by a discussion of
certain problems which will immediately occur
to the reader.
Did any rain precede seeding, and was
rain occurring in the general area?
This question was stressed by Coons,
Gentry and Gunn (1948 c) who noted in a
particular test that radar echoes from snow
were existing or were forming in the imme-
diate vicinity at the time of seeding. Without
the facilities utilized ..., especially the radar,
it probably would have been concluded that
the one inch of snow was definitely a result of
the cloud-seeding operation (p. 18).”
Theirs is a cogent argument. From the
standpoint of agriculture, however, it is not
necessary to prove that dry ice can precipitate
rain when no rain would have fallen, The
question is rather: was the amount or timing
of rainfall affected by seeding?
The Hawaiian experiments were all run. on
days of unusually high probability of natural
rainfall, and rain usually did occur in the area
near or at the time’ of. rain from the seeded
cloud, However, rainfall amounts under the
test clouds were generally larger than. those
in the vicinity.
As mentioned earlier, it is difficult to
define a practical and definite control by which
we may compare the results of seeding against
circumstances which might have. occurred had
no seeding taken place. As a step in this
direction, results for the 8 experiments where
clouds colder than freezing were seeded are
presented in Table 2 and compared with rain-
Table 2. Mean rainfall amounts measured by
specific rain gages on days of test and similar
days of no test.
s{2l|3)el¢g
Pile yb yas
bya lala
Rainfall
(Rainfall values are average
of 5 gages, Lanai)
Non-seeded days:
No. of days..
% of total days.
seeded Days:
No. of days. .
% of total days
‘THE SCHAEFER-LANGMUIR CLOUD-SEEDING TECHNIQUE IN HAWAII 47
12500 189" 01.95"
is
a . ou? 28"
Q ozs" | or
a #60" i
Bioo0o oe
2
g
s
= eS
s
3 7500
2 00
5000 o_o
<4 <2 G 8 io
0 2
TEMPERATURE AT LEVEL OF CO, RELEASE (°C)
Fig. 1. Rainfall amounts associated with seeded clouds of given temperature and thickness. ‘Temperature given
‘are at the level of COs release, approximately at the top of the cloud.
fall amounts occurring on days where no
seeding was done and similar upper-air and
cloud ‘conditions existed.
Because there are only 8 such cases, a sta~
tistical test for significant differences in the
rainfall frequency distributions on these days
versus non-seeded days will not yield a sig-
nificant result, The results are nevertheless of
interest even if definite conclusions cannot be
reached.
In what ways did rain occurring after
seeding appear to be related to seeding?
As previously mentioned, rainfall in cach
case followed ‘the seeding by only a short
interval of time or rain intensity increased. In
most instances, rain endured for less than att
hour. In the case of Test 1 L and 5 L, however,
rainfall continued for approximately three
hours. The heaviest vainfalk intensity in both
of these instances occurred between 15 and 30
minutes following seeding. Clouds having
hard or well-defined edges at the top were
chosen, for seeding with the idea that rain in
such clouds had not yet developed. Lacking
radar this was the best evidence concernix
the state of the moisture droplets in the loud.
Other experimenters equipped with radar
equipment have indicated that this feature is
indicative of the lack of formation of ice
crystals within the cloud (Orr et al. 1950).
What Was the Relation between Rain Cloud
and 0° C Isotherm?
Data plotted in fig. 1 indicate that for the
15 days of experiments in 1948—1949 re-
ported here, a difference in the rainfall amounts
exists between those where the cloud-top
temperature was warmer than freezing and
those where the cloud-top temperature was
colder than freezing. Since the freezing level
in Hawaii tends to be at roughly a uniform
height, this represents, as indicated in fig. 1,
not only colder temperatures but also great-
er cloud thickness. A correlation of cloud
thickness with observed rainfall was indi-
cated by the previous Hawaiian experiments
(Lnovorp and Hatsrnap 1948), and its
importance has been stressed by Visrnour
(1950). Since even the coldest temperatures
observed at the cloud tops were only — 5° C,
itis difficult to evaluate the relative importance
of temperature and cloud thickness.
Conclusions
No definite conclusions can be drawn from
these data owing to the small number of
48 LUNA B, LEOPOLD and WENDELL A, MORDY
Table 3a. CO, Tests
Island
Data from rg00 Z i Z .
ata from 15 9 % Sounding Honolulu Height and ‘Thickness
Times of | Quantic
No, & Date Pressure y
Pressure | cicceapt | & Altitude | PPP TST’ | of COs | s, 1
at Base e at Moisture ase ae |
Break {
(Began 3 min,
scatter
1M 6/28/48 jor mb Og yoo mb 13: 53LST)} 100 Ibs 3,000 ft | 15,500 ft
9,800 ft 9,850 ft 14: 03 100 3,000 15,500
2M 6/29/48 730 mb 03 730mb | 12:00 150 3,000 8,000
8,775 ft 8,775 ft 14:20 150 3,000 8,000
3M 7/9/48 680 mb Ls yoomb | 13:17 100 2,900 11,000
10,600 ft 10,500 ft 13:30 100 2,900 12,100
13236 50 2,900 11,300
15:34 00 2,900 8,000
15:39 50 2,900 8,300
4M 8/3/48 632. mb 08 755 mb 14! 22 100 2,500 9,500
12,450 ft 7,900 ft 34! 32 100 2,500 10,000
rat 45 100 2,500 10,500 :
5 M 9/22/48 639 mb ed 63x mb 13: 02 100 4,000 11,500
11,425 ft 12,950 ft 13:19 00 4,000 12,500
13:29 100 4,000 13,000 |
6 M 9/24/48 525 mb o7 sasmb | 12:35 1s 2,500 13,800
17,100 ft 17,100 ft 12:47 7S 2,500 14,500
12: $5 150 2,500 14,800
7M 11/9/48 608 mb o.7 "6x0 mb 13119 150 4,000 8,160
13,425 ft 13,375 ft
8 Mrt/27/48 850 mb 043 None 142338 150 2,000 14,000 |
None 151048 150 2,000 13,100 |
9 M 6/13/49 585 mb Os $73 mb 13:35 100 2,500 12,900 4
14,400 ft 14,900 ft
1 Figures given are approximately
O,, Top —O,, Base
4
°K.
2 Fine granules.
2 Coarse granules.
‘THE SCHAEFER-LANGMUIR CLOUD-SEEDING TECHNIQUE IN HAWAII 49
in 1948—x949-
of Molokai,
Alt. & Temp. at
of Cloud] Tevet Toe Release
Mean,
Temp. Remarks on Cloud and Rainfall Roinfall
. from 5, Gagess
Thicknes| Alt, |, 707,
Sonding
12,500 ft| 14,300 ft] —2.0°C| 12: 00 LST weather report from Molokai indicated rain showers, Rain) 3.59 in.
12's00 | 14,000 | —1.2 | was not visible to pilot at altitude, Ice formed on plane while seeding,
Heavy rain was seen from plane at 14: 33 LST under seeded towers,
5000 | -8,000 | +9. | No zain, 0:00
§,000 8,000 + 9.0
sooo | rov7oo | + ss | Light showers began at 13; 20 IST from cloud seeded at 13: 17 LST.| 0.23
g200 | ratoo | 4.0 | A sudden increase in rain could be seen at 13: 18 LST by observers on]
Sgoo | 1x,500 | + $2 | Lanai (shown in photographs). At 13:47 LST a further increase of|
stoo | 8000 | +80 | rain was noted, At x3: 56 LST rain covered most of West Molokai
$1400 | 8300 | +82 | as seen from Lanai,
#eso | ors0o | cr $3. | Cloud dissipated in 2 minutes after final seeding. No rain associated] 0,00
oo | roo0o | + 4.3 | with seeded cloud. Clouds in vicinity showed wind shear with tops
8000 | 10,500 | + 3.0 | and bases separating,
7200 | 11,300 | +22 | No rain, Cloud dissipated. 0.00
8,200 12,100 + 2.0
8,700 12,400 + 2.0 i
irjoo | 134500 | —x1 | Rain was falling from cloud at time of seeding, At time coincident with} 0.25
rsce | 14000 | —20 | first seeding, observer on Lanai noted darkening of cloud and a second
t2300 | t4300 | —a3 | shaft of rain from base of cloud, At ra: 43 LST rain covered xo times
the area prior to seeding, Rain activity decreased at 1a: 55 LST and al
new shower began at 13:04 LST. Cloud tops flattened out beginning
. at 13204 LST.
6,160 | ‘8,000 | +8.0 | Cloud dissipated rapidly. 0.00
12,000 | 14500 | —3.8 |’ Rain at time of seeding. Cloud tops lowered after first seeding (r42 40| 0.00
trico | 12800 |—-ox | LST); at 15:03 LST rapid dissipation of cloud could be seen from|
Tanai,
10,400 | 1,500 | +02 | Showers over seeded area at 12: 15 LST, Of short duration and smalll 0.60
‘amount, Heavy rain observed from plane under seeded cloud. at
Sep.
14: 10 LST.
50 LUNA B, LEOPOLD and WENDELL A. MORDY
Table 3b, CO, Tests
Island
Data from 1500 Z Sounding Honolul ‘
Tiventon. 6 Honolulu Height and Thickness
No. & Date Paaweuce Times of Quantity
Pressure | guengena | & Altitude | Drop EST | of CO,
at Base Beh | at Moisture iaiaad Top,
Break,
iL 6/28/48 yor mb nd 700 mb 12:56LST| x00 lbs | 3,000 ft | 15,500 ft
9,800 ft 9,850 ft 13:13 Too 3,000 14,500
13:19 50 3,000 14,500
aL r1/9/48 608 mb o7 610 mb 13150 100 =| 1,800 12,000
13,425 Ft 13,375 ft
31 11/18/48 $50 mb 0.2 00 mb 14: 27 75% 2,000 ' | 10,500
15,900 ft 18,300 ft 14: 29 150% 2,000 10,600
14:33 is 2,000 10,500
4L 11/26/48 69x mb 03 692 mb 12:37 65 unknown | 7,000
10,200 ft 10,150 ft 12t 42 35 unknown | 6,800
SL 1/10/49 635 mb os 735 mb 11 40 100 3,000 14,000
12,350 ft 8,600 ft 10 44 Too. 3,000 14,000
133 43 200 3,000 12,000
x3: 49 100 3,000 12,000
6L 6/13/49 $853 mb OS 573 mb 100 3,000 14,400
14,400 ft 14,900 ft 100 3,000 14,400
4 Bigures given are approximately
O, Top —~ Op Bas
Oa Top ~ Op Base oe
7
* Ice was scattered more slowly than for other tests.
THE SCHAEFER-LANGMUIR CLOUD-SEEDING TECHNIQUE IN HAWAII
$1
in 1948—~1949.
of Lanai.
Alt, & Temp, at
of Cloud | Tevel Ice Release
Mean
Temp. Remarks on Cloud and Rainfall Rainfall
; from 5 Gages
MThickness] Alt. | ,,/033.
Sounding|
12,500t| 13,800f]— 1.0°C] Between 1 and 2 in, fell previous day, Cloud built rapidly in 2] 1.95 in.
11,300 [14,000 |—ra min, following ice but 14 min. after seeding, top of cloud was
11,300 | 14,000 1,000 ft, lower. Rain at Lanai City began at 13:15 LST and
in Palawai Basin (seeded area) about 13:15 LST.
10,200 | 12,000 |-++ 1,2 | Rain had occurred in hour previous to test, Tops of CU were] 0.06
flattened into As at 7,000 ft, At 33: 53 LST plane flew under!
seeded area and encountered light rain and moderate turbulence,
8,500 ] 10,500 }-++3.5 | Cloud top dropped soo ft 4 minutes after 14:27 LST, Large| 0.03
8,600 | 10,600 | + 3.2 central portion of cloud dissipated, At 15:00 LST rain was ob-|
8,500 | 10,000 | + 3.5 served from plane and ground ditectly below seeded cloud. Nol
other rain on Lanai or Molokai, Stopped raining at 15: 30 LST.
6,900 |-+ 9.8 | No rain, 0.00
6,700 | +10.2
11,000 | 14,000 [5,0 | Small showers fell in morning but were not associated with CO,) 2.18
11,000 | 14,000 |—5.0 drop in A, M, Cloud seeded was not highest cloud in area but|
9,000 | 12,000 }—3.0 was in formative process with well defined hard tops, Cloud|
9,000 } 12,000 |—3.0 was small in area, Rain began as a single shaft directly below
point where plane was seen to emerge from cloud (7oo ft below
top of cloud). Rain rapidly spread over wide area,
11,400 | 14,400 |—3.2 |.No indication of rising cloud tops after seeding, Top of cloud) 0.12
11,400 | 24,400 |— 3.2 dissipated after second seeding and became separated from lower!
portion, Tops became 8,000 ft by 13:25 LST. Moderate rain
was observed at 13: 30 LST when plane flew under seeded cloud,
52 LUNA B, LEOPOLD and WENDELL A, MORDY
tests and the lack of suitable control. Never-
theless, it is important that all pertinent data on
the subject of cloud inoculation be added. to
the literature. The present experiments are of
particular interest because of the large rainfalls
which were experienced under the seeded
clouds and very possibly related to the in-
oculation. :
Acknowledgments
The authors are indebted to Mr James
Medcalf of the Hawaiian Pineapple Company
whose cooperation and assistance in these
experiments contributed much toward the
completion of the data included in the tables.
Aircraft were flown by Mr Lloyd B. Osborne
of the Hawaiian, Air ‘Transport Service who
was of assistance not only in flight but helped
in many of the practical, details of carrying on
the experiments. Mr Kenneth Willey, Mr
George Moriguchi and Mrs Jane B. Briggs
were valuable assistants in assembling the
data for the tables and in preparing the
drawings.
REFERENCES
Buncerow, T., 1949: ‘The Problem of Artificial Control
of Rainfall on the Globe. Tellus, x, 1.
Coons, R. D., Ganrry, R. C., and Gunn, R., 1948 az
First Partial Report on the Artificial Production of
Precipitation Stratiform Clouds, Ohio 1948. ‘Bull.
Amer, Met, Soc., 29, 5.
— 1948 b: Second Partial Report on the Artificial Pro~
duction of Precipitation Cumuliform Clouds, Ohio
1948. Bull. Amer. Met. Soc., 29, 10.
— 1948 c: First Partial Report on the Artificial Produc~
tion of Precipitation, U. S. Dept. of Commerce Re~
search Paper 30.
Lnororp, L. B., and Hatstiap, M. H., 1948: First
Trials of the Schaefer-Langmuir Dry-Ice Cloud
Seeding Technique in Hawaii, Bull. Amer. Met. Soc.,
29, 10.
Inororp, L. B., Burn, $., and Stmp, C. K., 1948: A
Key to Rain Gages in Hawaii. The Hawalian Planters?
Record, Vol. 52, 3 and: 4.
Luororp, L. B., 1949: Interaction of Trade Wind and
Sea Breeze, Hawaii. Journal of Meteorology, 6, 5.
Onn, J. L., Fraser, D., and Pettit, K. G., 1950: Canadian
Experiments on Artificially Inducing Precipitation,
Bull. Amer. Met. Soc., 3%, 2.
Vurnttour, R. R., 1950: On the Formation of Rain in
Clouds not Reaching up to the Freezing Level. Journal
of Meteorology, 7, pp. 223226.
Esselte, Schlm 51
