WMOD00559

• Project Name: Project Skywater
• Title: Project Skywater - Generalized Criteria for Seeding Winter Orographic Clouds - Summary Report
• Date: 3/1/1977
• Weather Modification - Doc Type: Report

f

The basic information used in this analysis was the rawinsonde data
at 6-hour intervals during exp,erimental events (these soundings give
the vertical distribution of t1emperature, moisture, and wind), 6-hour
precipitation totals summed from 2 hours before the sounding to
4 hours after the sounding, descriptions of the individual mountain
barriers and seeding networks, and tabulations indicating when the
seeding generators were on and off. If at least one generator was
on during the period'l hour before the start of a 6-hour block to
1 hour before the end of the block, the period was considered seeded.
Carryover contamination was not considered in this analysis.

Some of the projects had sounding data at intervals more frequent
than the 6-hour interval used in the analysis. Some projects had
other kinds of data, such as in-cloud measurements and radar obser-
vations. Although not used in the analysis itself, these data pro-
vided insight useful in interpreting results from the statistical
analysis.

An important part of the analysis was to identify where the effects
of seeding occurred as well as the magnitude of the "effect. Therefore,
the precipitation gages were glr:'ouped in similar regions in relation to
the mountain ba'rrier. The crest group included those gages in the
primary target area of the individual projects, generally those along
and on either side of the ridgfe crest. Other groups of gages were
considered as upwind or downwind.

II. Analysis Procedures

This st~dy considered 10 meteorological variables that depend on the
shape and size of the mountain barrier and on the characteristics of
the clouds. The variables werle grouped into four major physical cat-
egories: (1) the atmospheric mixing available to carry seeding mate-
rial to the appropriate portion of the cloud, (2) the water available
for conversion to precipitation, (3) the natural nuclei available to
start the precipitation process, and ,(4) the time available for nucle-
ation, growth, and fallout of the precipitation.

Using these variables, stratifications were performed by project and
groups of projects to determinle seed/no-seed ratios and statistical
probabilities, which suggest mlether seeding caused a change in pre-
cipitation. A nonparametric statistical test (Wilcoxon two-sample)
was used because it required no predeterm~ned assumptions about the
distribution of the data. Each of the variables was stratified to
optimize positive and negative seeding effects. Some single-variable
stratifications showed clear d,emarcation of seeding effects for one
project but conflicting results for another project. For other proj-
ects, the same stratification '~as completely uninformative.

5