WMOD00528

• Title: The Feasibility of Enhancing Streamflow in the Silver Iodide in the Sevier River Basin of Utah bt Seeding Winter Mountain Clouds
• Date: 12/1/1991
• Weather Modification - Doc Type: Report

While several of the stonn periods with large SL W flux amounts lasted about 1 to 2 days, one lasted only
4.5 hours (March 13, 1987) and one lasted over 4 days (February 8, 1989). Variability is clearly a
common element in winter stonns as has been found in other locations. For example, Super and Boe
(1988a) showed the total flux for 11 early 1987 stonns over the Mogollon Rim of Arizona ranged from
1 to 419 Mg per meter of ridgeline. Three of the Arizona sto~s exceeded 230 Mg per meter crosswind
and the rest were under 95 Mg per meter. Similar amounts and distributions are seen in table 3.1. Thus,
a few stonns each winter have considerable SL W flux while all others have only limited flux.

It can be expected that seeding potential also will vary considerably, and only partially because of
variations in SL W. Considerable variation can be expected in the success of delivering ice nuclei to the
SL W cloud due to T &D variability, temperature dependence of AgI, time available for nucleation, ice
crystal growth and fallout, and other factors. The notion that seeding will result in a 10- to 20-percent
increase in precipitation for each stonn is erroneous. Most likely, properly conducted seeding produces
relatively large percentage increases in a small fraction of the stonns and limited changes in the others
as suggested by the statistical analysis of the Bridger Range, Montana, Experiment (Super, 1986) and that
of the Northern Wasatch projects (Hill, 1979).

Data from table 3.1 were used to investigate the relationship between stonn total precipitation far up the
windward slope and total SL W flux overhead. Figure 3-1 shows the plotted pairs of observations for the
27 stonns with available measurements, and the linear least square regression line. It is seen that SL W
flux tended to be larger in the larger precipitation producing stonns. This may seem a contradiction as
it might be anticipated that efficient stonns would convert most of the flux to precipitation and inefficient
stonns would have considerable flux and little precipitation. However, the observations show otherwise.
The situation in Utah is probably similar to the northern Arizona stonns discussed by Super and Holroyd
(1989) which had abundant SL W and limited snowfall during some (inefficient) phases and the converse
during other (efficient) phases. The inefficient phases tended to be at the beginning and ending of the
episodes when the cloud deck was shallow and cloud tops were wann. Middle portions of stonns,
associated with frontal passages, tended to have thick clouds with cold tops that apparently produced
abundant natural ice which converted SL W to snowfall.

The linear correlation coefficient for the data of figure 3-1 is 0.75, indicating about 54 percent of the
variance in snowfall was associated with SLW availability. While that is certainly a significant
relationship, almost half the variance is unexplained. (The correlation is considerably enhanced by the
two largest flux producing stonns as a value of 0.45 results without them).

Similar data from 11 northern Arizona stonns published by Super and Boe (1988a), plus stonn total
precipitation from the microwave radiometer site, were plotted on figure 3-2. A similar relationship was
found with a correlation coefficient of 0.88. Again, the larger flux producing stonns greatly enhance the
relationship. Finding similar relationships in Arizona and Utah, with larger flux amounts associated with
larger precipitation producing stonns, adds credibility to the results of figure 3-1.

The median stonn total precipitation for the cases plotted in figure 3-1 is 4.6 mm. All but 2 of the 13
stonns with below median snowfall had limited SL W flux, The greatest flux amounts were associated
with 7 of the 13 stonns with above median snowfall totals. This suggests that total precipitation over the
course of a stonn episode can be used as crude indicator of SL W availability with little flux expected if
precipitation is less than a few millimeters. However, all stonns with both precipitation and flux data had
some flux available if snowfall was observed. Hence, the occurrence of measurable precipitation suggests
that at least some flux probably existed. Thus, it is reasonable to examine precipitation records, as done

14