WMOD00393

• Title: Weather Modification Glossary
• Prepared By: A. S. Dennis
• Date: 1/1/1987
• Weather Modification - Doc Type: Report

608 WEATHER MODIFICATION

Committee statisticians applied a gamma trans-
formation to normalize the data, and then estab-
lished a linear regression relationship between
the transformed target and control precipitation
amounts for the historical period. Then the con-
trol station precipitation during a seeded storm
was entered into the regression equation to esti-
mate what the target station precipitation would
have been in the absence of seeding, and the
result was compared to the observed target sta-
tion precipitation.
On the basis of a study of 299 seeded storms,
which occurred over several seasons in different
projects in the Western United States, the Advi-
sory Committee reported in 1957 apparent in-
creases in precipitation of 9 to 17% over the
amount that would have fallen without seeding.
The Advisory Committee also stated that there
was only a very small probability that the appar-
ent increases were due to chance. Although the
Advisory Committee's report was criticized on
various grounds, including the possibility of nat-
ural changes in target-control relationships, its
findings have been supported by later review
panels.
It is difficult to translate indications ofprecipi-
tation increases at individual gauges into esti-
mates of additional runoff from an entire water-
shed. The runoff from a watershed is not a linear
function of the precipitation falling on it. Precip-
itation gauges are sometimes sparse and tend to
be located in accessible areas at relatively low
elevations. In addition, accurate measurement
of snowfall, especially under windy conditions,
is extremely difficult. An approach that over-
comes some of these difficulties is to base the
evaluation on snow course data or, better yet,
on the streamflow data themselves. The histori-
cal regression approach is applicable to runoff
data. Care must be exercised to check on possi-
ble relocations of stream gauging stations and
changes in the behavior of a watershed due to,
for example, clearing of a forest.
The correlation between runoff amounts for
adjoining watersheds depends on watershed
characteristics. On the west side of the Sierra
Nevada in California, rocky watersheds allow
relatively little percolation of water into the soil
and evapotranspiration is slow. Most of the pre-
cipitation that falls becomes streamflow, either
immediately or as melted snowpack. Correlation
coefficients relating annual runoff from adjoin-
ing watersheds run as high as 0.9, and multiple
correlation coefficients as high as 0.97. The
close correlation permits the historical regres-
sion method to be applied with considerable

confidence. Analyses by a number of authors
have indicated increases in seasonal runoff in
rivers draining mountains in the western United
States, generally ranging between 5 and 15% of
the natural runoff.

C. RESEARCH ON SEEDING EFFECTS ON WINTER
OROGRAPHIC CLOUDS

A number of field experiments, some of them
involving randomized seeding trials, have been
carried out to refine estimates of the effects of
cloud seeding on orographic precipitation and to
determine those conditions most conducive to
precipitation increases by seeding. The largest
such project currently in existence is the Sierra
Cooperative Pilot Project (SCPP), which is con-
ducted by the U.S. Bureau of Reclamation and
cooperating agencies in the American River Ba-
sin in northern California.
Investigations of seeding effects in orographic
clouds have centered around the concept of
seedability. It is generally agreed that seedabil-
ity is associated with the presence of super-
cooled liquid water, although the presence of
supercooled liquid water does not guarantee the
existence of seedability. Searches for super-
cooled liquid water have been carried out with
new instruments including dual channel micro-
wave radiometers and aircraft equipped with op-
tical probes for counting, sizing, and classifying
cloud and precipitation particles. The research
on winter orographic cloud seeding has also in-
volved numerical modeling studies conducted
on large computers to simulate the trajectories
of snowflakes produced by both natural and arti-
ficial ice nuclei.
Present indications are that supercooled liquid
water is most likely to be associated either with
shallow stratiform clouds with tops in the - 5 to
-120C region or with moderate updrafts associ-
ated with embedded convective elements in the
general overcast. In maritime clouds, which are
often observed on SCPP, surprisingly high ice
crystal concentrations are observed on occa-
sion, for example, la-50 per liter at a tempera-
ture of -SOC. These concentrations, which
greatly exceed the measured concentrations of
natural ice nuclei, are attributed to a variety of
secondary ice crystal production mechanisms.
Some of the postulated secondary ice crystal
production mechanisms require the presence of
large cloud droplets, which are characteristic of
maritime clouds. In inland regions, such as Col-
orado, secondary ice is less commonly ob-