WMOD00196

• Contract/Permit #: 2003-01
• Sponsor Name: Gunnison County, City of Gunnison, Crested Butte
• Project Name: WM Evaluation Report
• Title: The Development, Conduct and Evaluation of a Cloud Seeding Program for the Gunnison River Basin, CO During the 2003-04 Season
• Date: 8/1/2004
• County: Gunnison
• State: CO
• Weather Modification - Doc Type: Reports

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

years duration (20 years or more if possible). These records are partitioned into those located
within the designated "target" area ofthe project and those in a nearby "control" area. Ideally the
control sites should be selected in an area meteorologically similar area to the target, but one
which would be unaffected by the seeding (or seeding from other adjacent projects). The
historical data (e.g., precipitation) in both the target and control areas are taken from past years
that have not been subject to cloud seeding activities. These data are evaluated for the same
seasonal period of time as that when the seeding later was conducted. The target and control sets
of data for the unseeded seasons are used to develop an equation (typically a linear regression)
which predicts the amount of target area precipitation, based on precipitation observed in the
control area. This regression equation is then used during the seeded period, to estimate what the
target area precipitation should have been without seeding, based on that observed in the control
area. This allows a comparison to be made between the predicted target area precipitation and
that which actually occurred during the seeded period, to look for any differences potentially
caused by seeding activity.

This target and control technique works well where a good historical correlation can be
found between target and control area precipitation. Generally, the closer the target and control
areas are geographically, and in terms of elevation, the higher the correlation will be. Areas
selected too close together, however, can be subject to contamination of the control area by the
seeding activities. This can result in an underestimate of the seeding effect. For precipitation and
snowpack assessments, a correlation coefficient (r) of 0.90 or better would be considered
excellent. A correlation coefficient of 0.90 would indicate that over 80 percent of the variance
(r2) in the historical data set would be explained by the regression equation used to predict the
variable (expected precipitation or snowpack) in the seeded years. An equation indicating
perfect correlation would have an r value of 1.0.

In this particular case, one potential target site (McClure Pass, which measures data
appropriate for both the precipitation and snowpack evaluations) was considered questionable
due to its location near the northern boundary of the target area, where lesser seeding effects
might be expected to occur due to targeting considerations. Also, there was the question of
potential contamination due to its proximity to another historical seeding program over the

45