WMOD00532

• Title: The Rain Stimulation Campaign in Panama Summer 1983
• Date: 6/1/1984
• Weather Modification - Doc Type: Report

t
I
I"
~
r

4. Assessment

Even when cloud seeding is done as an effil~rgency
measure, only for a short while, and an evaluation effort
is not planned or possible, an assessment question still
has to be asked: how well did it go? In this section, the
principal methods of assessment and their applicability
to last summer's operation will be briefly reviewed.

Departures from expected rainfall. In various
assessments, the expected rainfall has been taken as
the long-period normal, the rainfall occurring during
randomly selected unseeded occasions during the
operating period, or an amount based on comparison
with a nearby unseeded control area. None of the data
bases needed to establish an expected rainfall amount
was available in the present instance.

Cloud models. Computerized mathematical models
of clouds have been constructed that estimate the
extent of cloud and rain development when both the
state of the atmosphere and the strength of the
initiating convective impulse are known. The estimates
from calculations are different for impulse seeded and
unseeded clouds. Sometimes the differences suggest a
positive response to seeding, Evaluation is performed
by noting whether the observed cloud and precipitation
development better matches the seeded or unseeded
version of the model.

It is possible to generate a large volume of model
calculations, Drawing conclusions is difficult because of
two circumstances. First, sampling the state of the
atmosphere for a short time in a place like Panama is
somewhat like opinion polling: the result depends on
which section of the city is polled. The cloud may not
start in the section of the domain that was sampled.
Furthermore, the situation may be changed by the
latest events, often occurring after the samplin!~ took
place. Second, the initiating impulses in nature are
drastically nonuniform in strength, timing, and location.
As a consequence, developments over the mountains
or a strongly heated area are drastically different from
those over a weaker source of upward impulses, such
as the lake area.

A substantial volume of calculations based on :rawin-
sonde observations from last summer's operations has

been generated, but resources have not been available
for more than a cursory examination of them. Runs of
the model suggest that, in the climate of Panama,
relatively small atmospheric differences led to enormous
differences in the height that a cloud reached. Once a
cloud cleared certain hurdles, it was almost sure to
reach to the tropopause. Typical of these hurdles are
thE! presence of shallow stable layers near the ground
and the relatively high exposure of a small cloud to
evaporation at its border. Most hurdles were cleared if
thE~ cloud grew to exceed a diameter of about 2 km. It is
in the critical stage of growth just before this diameter is
reached when the model responds more sensitively to
seeding.

These theoretical expectations illuminate a practical
aspect of the situation in Panama. The greatest quantity
of :rain fell not from individual clouds, especially not
from the smaller ones, but from cloud clusters or very
larHe cloud aggregates. Each of these entities, however,
was fed by clouds that appeared at the borders of the
system and subsequently grew rapidly and merged with
it. The strength of the system appeared closely related
to the strength of development of the feeder clouds.
Indeed, the presence of strongly-developing small
clouds appeared necessary for initial formation of a
cluster or large aggregate. Thus, it seems that stimu-
lation of growth in a critical phase when the cloud is
relatively small may have had a great influence on the
contribution it made later to production or invigoration
of the large complexes from which most of thve rain fell.

These considerations guided the cloud seeding pilots in
thei:r selection of clouds for seeding. Whenever possible,
those seeded were feeder clouds on the up-shear edge
of clusters or large aggregations. Clouds that had
recEmtly arrived above the freezing level showed
vigorous growth but were still only a kilometer or two in
diameter, Most of the pilot reports of "good" responses
belonged in this category.

Modeling of the sort discussed above is widely regarded
as the wave of the future in weather modification
evaluation and a great deal of work is going into its
devE!lopment. However, it can never be better than the
observation on which the calculations are based. A

9