WMOD00470

• Project Name: Project Skywater
• Title: Environmental Impacts of Precipitation Management: Results and Inferences from Project Skywater
• Date: 6/6/1977
• Weather Modification - Doc Type: Report

Bulletin American Meteorological Society

cations of widespread and prolonged applications that
might crown their success, be assessed.
The pattern of results has given strongly significant
indications of effects on certain definite categories of
clouds combined with mostly weak and insignificant
effects averaged over the area and duration of the ex-
periments. Exceptions are results of the North Dakota
Pilot Project, which showed significant areawide in-
creases of the order of 12% (Dennis et al., 1974) and
possibly shorter programs of hygoscopic seeding in Texas
and Oklahoma (Bureau of Reclamation, 1976b). Thus,
the actual environmental impacts of Skywater activities
have to do mostly with certain individual storm events
for which a high probability of effect may be presumed.
Unfortunately, the means are not at hand for measuring
or even estimating the magnitude of these effects be-
cause of their extremely large variability under natural
conditions. For this reason, all environmental impacts
of Skywater activities are subliminal and defy identifi-
cation. The most that can be said is that they are too
weak to be manifest.
Impacts too weak to be manifest may nevertheless be
real and may even be environmentally and economically
important. Estimates of the magnitude of such subliminal
impacts have been made by inference from studies of
the environmental effects of natural precipitation varia-
tions. For example, it has been estimated from crop data
that each added centimeter of summer rainfall in west-
ern North Dakota increases the biomass production of
rangeland by '-'60 kg ha'l (Johnson, 1974). On this basis,
one may estimate that the North Dakota Pilot Project
caused some supposed increase in rangeland biomass
production, even though this impact is not directly
verifiable by observation. Such methods of estimation,
questionable as they are, provide the only presently
available means of assessing impacts of the Skywater
experiments or of taking such impacts into account in
decision making. None of the impacts so far assessed in
this manner are strong enough or certain enough to
warrant further consideration.

5. Current and future progress and .problems in
evaluating impacts

a. Research approach

Nearly all of the environmental research so far under-
taken in the name of weather modification has responded.
to particular potential hazards affecting identifiable in-
terest groups (such as the carrying capacity of winter
game ranges) or to specific environmental characteristics
seen as particularly important (such as the range of
principal forest species). This orientation was rein-
forced by the assessments by a working group of the
Ecological Society of America (Livingstone et al.) 1966,
p. 43) that prolonged application of precipitation man-
agement on even the mesoscale would produce "shifts
of range, local or complete extermination of species, and
at least an initial increase in weeds and pests." As indi-
cated by Steinhoff and Ives (1976), the research approach
was mostly aimed at understanding the physical proc-
esses relating these particular potential impacts to in-

499

creased precipitation, focused on the particular impacts
that it was thought might be most important.

b. Observational approach

As the work progressed, the methods used became in-
creasingly those of observing instances of gross altera-
tions in precipitation and other interacting environ-
mental factors. For example, in the alpine zone, snow
fences were erected that imposed snow loads and dura-
tions on the local vegetation equivalent to several hun-
dred per cent increase in precipitation. In the forest
zone, the range of study plots embraced a range of
climatic conditions larger by orders of magnitude than
the anticipated effect of precipitation management.
This method of assessment implicitly assumes that the
response of the biological system is linear with the
strength of the stimulus. K. Harper (personal communi-
cation, 1976) has proposed that, on the other hand, the
response is probably nonlinear (increasing as a higher
power of the stimulus) on the grounds that most or-
ganisms are provided with genetic means of coping with
environmental variations within a limited range and
become severely stressed only when this range is ex-
ceeded. If his view is correct, linear interpolation from
observed gross effects of snow accumulation changes will
overestimate the true effects. Since the reason for exag-
gerating the stimulus in the first place is to bring the
response up to the level of observability, it may prove
very difficult to verify the linearity, or nonlinearity, of
the response to the much weaker stimuli likely to result
from precipitation management.
There is a temptation to conclude that a response too
weak to be significantly observed must be unimportant
and to proceed with an action that involves a real but
subliminal risk of environmental damage, or subliminal
opportunity for environmental benefit. However, human
behavior is full of instances of risk-avoidance actions in
which the motivation is an image in the mind of the
possible adverse consequences rather than a physically
or statistically verifiable reality; in other words, a model.
A model of the stimulus-response relationship, if credi-
ble, becomes the basis for decision making (behavior)
even when it cannot be objectively verified at the re-
sponse level actually involved.

c. Process approach

This method is potentially promlSlng for investigation
of relatively strong, direct effects of weather modifica-
tion on particular parts of the environment. In its
present state of development, however, it is incapable
of comprehending or predicting the impacts of weather
modification on an ecosystem as a whole because the
interactions among components are so complex as to
defy analysis and simulation. It has come to be recog-
nized that, in the settings so far investigated, the poten-
tial disturbing effect of weather modification is over-
whelmed by other disturbances to the status quo, some
man made and some natural.
From this state of affairs, two broad conclusions are
now being drawn. First, since investigation of the points
of specific environmental impact thought to have the