WMOD00372

• Project Name: High Plains Cooperative Program
• Title: The Design of HIPLEX-1
• Date: 4/1/1979
• Weather Modification - Doc Type: Report

X. EVALUATION

The evaluation of HIPLEX-1 is envisioned as consisting of three
efforts. The first is a statistical evaluation of the physically
based experiment whose primary objectives are to establish proof of
changes due to seeding and to verify the physical sequence of events
leading to the changes. The second effort is concerned with the
detailed physical analysis of case studies to establish the cause-
effect relationship, or lack thereof, of any seeding-produce~ modifi-
cations to the cloud and its mechanisms. These two efforts will be
conducted in parallel during and after the randomized experiment.
The third effort, which will come at the conclusion of the other two,
will be devoted to final interpretation, integration, and evaluation
of the results of the first two efforts.

~'

~

During the experiment and its evaluation, there will be a continuing
interchange of data and results between the statistical and the
physical evaluation efforts. Important findings from the statistical
effort will help focus the case-study effort into more pertinent and
effective areas of exploration and, as the causative mechanisms for
modification are verified, the physical effort will generate. additional
a priori hypotheses that can be statistically tested in future years
of the program. In addition, the results of the physical studies
will provide direction of a posteriori statistical analyses of data
from previous years. An effective interchange between the two
evaluation efforts provides the background for major progress as the
experiment proceeds.

The core of the statistical evaluation will involve the 13 primary
response variables (representing 10 physical characteristics), which
are considered crucial to the statistical testing of the physical
hypotheses. The eight secondary response variables may yield informa-
tion about seeding effects, but they do not represent essential links
in the physical process. Both sets of response variables, p~us other
information, will be considered in the physical evaluation. '

A. Statistical Evaluation

The statistical evaluation of the HIPLEX-1 experiment involves five
issues: (1) the type of experimental design, (2) stratification,
(3) avoidance of bias, (4) statistical techniques, and (5) sample-size
estimates. These issues are briefly discussed below; a more thorough
discussion of statistical evaluation is presented in appendix E.

~

~

1. Experimental desi~n. - A randomized single-area design will be
used for HIPLEX-l. T e individual cumulus congest us clouds that
will be the test events for HIPLEX-1 can be thought of as, moving
target areas.

28