Weather Mod Critical Issues Report

• Project Type: General OWC
• Project Name: Weather Modification
• Title: Critical Issues in Weather Modification Research
• Date: 1/1/2004
• Water Conservation - Doc Type: Final Report

.

I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

flank of the storm cells (Farley et aI., 2004a, Farley and Orville, 1999), in agreement with
recent observations (Thompson and List, 1999).

The NRC report quoted results from Smith et al. (1997) showing a 45% decrease
in crop damage due to hail suppression in a nonrandomized operational project, but cast
doubt on the results with some unpublished analysis using ratios. Further examination of
Fig,2,1 in the NRC report suggests that the analyst picked a starting point that would
bolster his point. In addition, the use of ratios in precipitation records requires extreme
care. Normally a scientist being corrected or challenged on a conclusion is afforded the
opportunity to respond, That such an opportunity was not offered here suggests a
possible negative bias on the part of the NRC committee, although we realize that NRC
panels do not normally invite comments and responses,

Other evidence exists indicating decreased hail damage during hail suppression
efforts. Mesinger and Mesinger (1992) examined 40 years of operational hail
suppression data in eastern Yugoslavia. After attempting to remove the effects of
climatic fluctuations during the period, they estimated that the hail suppression projects
reduced the frequency of hail between 15 and 20%. Rudolph et al. (1994) reported on
results from a randomized crossover hail suppression experiment conducted in northern
Greece during 1984-1988, Data were collected on 37 days from a total of 196 hailpads
spaced an average 4.5 km apart, Hailstone size distributions showed clear evidence of
beneficial treatment effects, Aircraft seeding using silver iodide (AgI) generators and
flares, primarily on flanking feeder cells, was employed. Target hailpad counts (impacts)
ranged from 38% to 100% less than control counts in all 12 size categories, with an
average reduction of 55%, On an annual basis, P-values ranged from 0.002 to 0.02.
Dessens (1998) in a long running operational program using AgI ground generators and
hailpads in southern France found a 42% decrease in hailstone number using target-
control analyses.

As in most research on operational programs in cloud seeding there are problems
oftargeting the seeding agent. Chemical tracers are a key to determining the extent to
which a target area is covered. Linkletter and Warburton (1977) found that during the
NHRE the AgI was broadly dispersed when weak, poorly organized storms were seeded,
but that the seeding agent was confined to only limited regions of the more vigorous
storms that had well-defined internal circulation patterns, In the 18 storms seeded in
1973 and 1974 only 50% of the 1973 storms and 70% of the 1974 storms had "seeding"
silver above background concentrations, Based upon theoretical predictions, less than
10% of the storms had enough silver to represent a significant seeding effect. Further
analysis of four storms in NHRE (Warburton et aI., 1982) revealed that the seeding
results appeared to fall into three categories; those where the AgI concentration was
relatively constant over a wide range of precipitation amounts; those where the
precipitation amounts were small and independent of silver contents; and those where
there is a positive correlation between silver concentration and precipitation amount. In
the cases with positive correlation, the seeding was associated with a precipitation
increase of about 1.7 mm depth of water per square meter.

13