WMOD00412

• Title: Final Report on Utah Cloud Seeding Experimentation Using Propane During the 2003/04 Winter
• Date: 3/1/2005
• State: UT
• Weather Modification - Doc Type: Report

I

I

I

was provided both by considering the mean residuals from the pairs testing and from the mean double
ratio testing

II

Two other single partitions were attempted, and a number of dual partitions with one always HAS wind
directions of270 deg or less. That direction requirement alone reduced the population size to 69 EUs and
any second partition would reduce it further so dual partitioning attempts were recognized as challenging.
Detecting a real seeding effect at an acceptable level of statistical significance becomes problematic with
a small sample population. A number of the results were somewhat suggestive of seeding effectiveness.
Examples include when SL W was detected by icing rate sensors, when the warmer half of the EUs were
tested using HAS temperatures, and especially when the largest natural snowfall periods were elIminated
by limiting the control gauge SWE per EU to 0.05 inch. While the latter dual partition was quite
suggestive of seeding-caused snowfall increases, none of the additional partitions provided results as
convincing as the wind direction partition. Most of the resulting suggestions should be regarded more as
tantalizing than as conclusive. The underlying problem was very likely the limited population sizes being
tested when any two partitions were applied. If any actual seeding effects existed, the available sample
size was too limited to detect them to an acceptable level of statistical significance. Such results should
be considered inconclusive, neither demonstrating an actual seeding effect nor the lack of one. Only
testing a larger population than available would offer the possibility of providing more definitive: results.

I

I,

Ii

I

The most solid and important results of the statistical testing can be summarized as follows. It is highly
likely that a real seeding effect resulted from propane seeding when transport winds carried the seeding
plume over the target gauges. Overall, seeded units provided at least 20% more snowfall than nonseeded
units. If all hours containing the 69 EUs with proper targeting had been operationally seeded, the
estimated likely outcome would have been an increase of8% in the "seasonal" precipitation averaged for
gauges GTR and GSC during the 3.5 month period of experimentation.

I

I

I

A case study analysis was performed for three adjoining EUs on 21 December 2003, with two of the
EUs nonseeded and the middle EU seeded. By chance, this particular seeded EU coincided with a time
period having minimal natural snowfall, seldom observed during other EUs. The adjoining nonseeded
EUs had considerably more SWE than the seeded EU, both with highest accumulations at the control
gauge. But that gauge had only 0.009 inch during the seeded period while target gauges had 2 to 4 times
as much'SWE, increasing with distance downwind of the seeding site. The most convincing evidence that
seeding produced much of the increase was provided by examination of2D-C probe ice particle images
from TAR which were very similar to those published from pulsed seeding experiments conducted during
earlier winters. The seeded images were relatively small, generally uniform in size and shape and were in
much higher concentrations than observed during the adjoining nonseeded periods. Even with
conservative assumptions, it was estimated that seeded snowfall rates exceeded 0.010 inch h-l at the core
target gauges and likely approached twice that rate at the downwind gauge. Such seeding-produced
precipitation rates are in good agreement with earlier cases study experiments in Utah and elsewhere in
the Intermountain West.

I

I

I

While such rates are light snowfall, about half of all hours with detectable natural snowfall fall at the
rate of 0.010 inch h'! or less in the experimental area. Frequent hours with light snowfall are common at
higher elevations in the Intermountain West. The seasonal contribution of light snowfall is important
because of its high frequency of occurrence. Increasing the snowfall rate by 0.0 I 0 inch h-! for 200 hours
would result in a seasonal increase of 10% for a normal November through March winter, based on long-
term snow pillow observations in the experimental area. Examination of past icing rate sensor and other
SL W data from the Wasatch Plateau revealed that approximately 200 hours with available SL W could be
expected during a typical five month winter. Consequently, there is some statistical and physical basis to
expect that propane seeding might produce a seasonal S WE increase in the vicinity of 10% in thl~ limited
target area examined.

I

I

I

I
I

lV