WMOD00464

• Title: The Management of Weather Resources - Volume II
• Prepared For: The Weather Modification Advisory Board
• Date: 6/30/1978
• Weather Modification - Doc Type: Report

A-3

reduce such instances in the future. The assignment of such responsibilities to a single person
of appropriate authority and willingness to act firmly may be the only way to assure that this
will happen.

Questions of calibration inevitably arise. Too many of us were trained in university
laboratories whem, often unwisely, the calibration of instruments was taken for granted. Con-
siderable thought and effort should be expended on which measuring devices are in special
need of calibration, and how this may best be done. In particular, doing this properly will avoid
the occurrence of such sentences as "Because of a radar calibration error, the seedable days of
the second year were biased toward more severe storms." It is, of course, not only radar cali-
brations that need attention.

Even with immediate oversight done well, there will be a need for auditing proc,esses at
various levels. Methods for checking -- on random samples -- the recording of data, both what
was written down and how it was identified, are likely to be essential. This is especially impor-
tant where the rl~corder is not blind to whether or not the day or other unit in question is
"seeded" -- and l~ven whether or not it was declared "suitable". Recording systems, automatic
or manual, cannot be trusted to function well without internal checks.

There may well be a place for an overall audit at a later stage. Those reslPonsible for the
design and management of clinical trials of new drugs are beginning to find the establiishment
of independent audit committees worthwhile.

* parallel analysis and deprjvacy *

A related approach, now in use in a major clinical (medical) trial, is to set up an alterna-
tive center for data analysis, in parallel with the main center, and to send copies of the data, as
generated, to both. Some find this unnecessary duplication of effort; others point to advantages
of dual analysis. In which weather modification studies such a procedure would be appropriate
is a problem that deserves thought.

The issue of deprivacy is a serious one. A common attitud~ of scientists, for which there
are many reasons, is that data is the personal property of the scientist gathering it -- that others
may see the details only as he or she prescribes. In major weather modification experiments,
the need for justice to be seen to be done overwhelms other considerations, at least in due
course. The details u and not just summaries -- neled to be available to serious workers. The
only questions arle when, and to whom.

The detailed results of exploratory phases, we suggest, should be freely availablle no later
than a year" after the conclusion of the exploratory phase. The detailed results of confirmatory
phases should also, we suggest, be freely available no later than a year aft€:r each planned
evaluation point (these ought probably not to be spaced closer than one every two years).

Perhaps th~ise results should be available almost immediately to competent groups with
real research interests and a written commitment not to make public their reports and analyses,
either orally or in writing, until the corresponding relPort of the main study is made public.

The results of "piggyback" experiments, also, should become freely available. How this
is best to be don.e requires careful consideration.

It is not a pure accident that "deprivacy" and "deprivation" share their first seven letters.
The policies just suggested do deprive those conducting the experiments -- and, in a different
way and to a different degree, those responsible for financial support and policy managlement n
of things of vaIUl~. We do not recommend deprivacy on any ground of abstract principle; rather