WMOD00554

• Title: Seeding Summertime Convective Clouds to Increase Black Hills Rainfall
• Weather Modification - Doc Type: Report

for 3 years, logged only 27 shower days.
For an operational project, one should
plan on about 10 days per season
providing good seeding opportunities
over a small, fixed target area.

The data in table 2 on numbers of
shower and storm days and on average
rainfall per operational day in the
un seeded targe'ts indicate that shower
days contribute about 7 percent and
storm days about 93 percent of the
natural summer rainfall on the east side
of the Black Hills. (Contributions from
other types of days, which include
frontal overrunning situations, are
negligible after early June.) Combining
the data in table 2 for seeded targets
on shower days and un seeded targets on
storm days, we find that the shower days
now contribute some 16 percent of the
total rainfall, which is increased
thereby by about 12 percent.

The preceding calculations rest on
the assumption that results of Cloud
Catcher and the Rapid Project can be
extrapolated to larger areas. In fact,
interactions among the seeded clouds
might complicate matters considerably.
A seeding strategy for increasing
rainfall can not be selected
intelligently without having the target
area clearly in'mind. consider, for
example, the Rapid Project shower days
with southwesterly flow (table 2).
There is no doubt that, in both the
north and south targets, rainfall was
heaviest on days when the particular
target was seeded. [We do not know what
fraction of this favorable result was
due to suppression effects in the no-
seed target.] Therefore, if one wants
more rain in the north target area, he
should seed it; if one wants more rain
in the south target area, he should seed
it. However, if the objective is to
maximize total rainfall over the two
target areas, the recommended treatment
would be to seed the south target
always. Seeding the south target
yielded an average rainfall across the
two target areas of approximately 1.3
mm, compared to an average of only 0.65
mm when the north target was seeded.

Of course, the Rapid Project did not
test the more obvious treatment for
maximizing total rainfall, namely,
seeding clouds in both target areas on
the same day, whenever and wherever they
'appeared, nor, for that matter, did it
test the option of not seeding at all.
This exercise points out the need for
clearly defined objectives in selecting
a seeding strategy, as well as the
ambiguities that arise in the
interpretation of results from

randomized crossover experiments when
strong dynamic effects are present or
suspected.

4. SUGGESTIONS FOR FUTURE RESEARCH
Cloud Catcher I left some unanswered
questions about the effects of silver
iodide seeding on convective clouds,
particularly when the convective cells
are clustered together. It provided
evidence of significant rainfall
increases due to silver iodide seeding
of clouds of moderate size, but it did
not make clear how the individual cells
reacted to produce the apparent
increases. These unanswered questions
were to be addressed in Cloud
Catcher II, which was in the! field in
1971 and 1972. It used an S-band radar
to avoid attenuation by precipitation, ,a
floating target design, and quite
sophisticated programming in the on-line
computer which controlled the radar
scans, logged and processed the radar
data, and provided real-time graphic
displays to the project meteorologist
(Boardman and Smith, 1974).
Unfortunately, seeding suspensions
following the Black Hills flood of 1972
and program cuts in 1973 led to
termination of Cloud Catcher II before
its potential could be realized.
Lingering questions on the part of the
general public about whether or not
cloud seeding contributed to the 1972
flood make it doubtful whether seeding
of Black Hills summer clouds will be
attempted again during this century.
Nevertheless, this section considers
what a new round of experiments might
entail.

In assessing the status of seeding
of convective clouds, it is of some
value to compare the South Dakota
results with those obtained subsequently
in other places. statistical analyses
of the second Florida Area Cumulus
Experiment (FACE-2) failed to prove the
existence of rainfall increases for
either the total target area or for the
floating targets encompassing the cells
selected for treatment (Woodley et al.,
1983). Subsequently, Gagin at al.
(1986) used computerized cell-tracking
techniques in a detailed post:-hoc study
of convective cells within seeded and
unseeded cloud systems on FACE-2. They
found evidence suggestive of rainfall
increases for clouds that were treated
early in their lifetimes with more than
8 silver iodide flares. Their results
for such cases were in good agreement
with the estimates based on table 1,
namely, that rainfall from a moderate
shower cloud can be multiplied by a
factor of 2 or 3 by silver iodide
seeding. However, unlike Dennis et ale

8