WMOD00439

• Title: HIPLEX-1: Experimental Design and Response Variables
• Date: 4/4/1984
• Weather Modification - Doc Type: Report

510

JOURNAL OF CLIMATE AND APPLIED METEOROLOGY

VOLUME 23

permitted by increased understanding, leading ulti-
mately to an area seeding experiment.
2) A detailed multistep physical hypothesis was for-
mulated with emphasis on the expected differences
between seeded and nonseeded clouds. For each step
of the hypothesis, response variables compatible with
, the measurement capabilities available for the exper-
iment were defined, so that the expected differences
could be monitored. For the response variables asso-
ciated with each step, indications were provided of
how (at least implicitly), when, and where the seeded/
nonseeded differences were to be measured. This aspect
of the experimental design is important because it pro-
vides for determining not only whether the seeding
produces increased rainfall, but also whether the phys-
ical mechanism is correctly understood.
3) Physical measurements of the characteristics of
candidate clouds were made and analyzed in real-time
to determine their suitability as experimental units.

Because of the approach that was used, HIPLEX-
I has been able to make a meaningful contribution to
the undet;"Standing of precipitation processes (natural
and modified) in cumulus clouds in spite of the un-
timely termination of the experiment and the small
sample size that resulted. HIPLEX-I was a test of
"static" seeding with dry ice, carried out to produce
microphysical effects in the clouds. The use of dry ice
permitted timely introduction of the ice nucleating
agent into the desired regions of the clouds, and ver-
ification of the production of ice crystal plumes by the
seeding was included in the experiment. The physical
and statistical evaluations of HIPLEX -I will indicate
the degree to which the physical hypothesis describes
the actual precipitation processes in the test case clouds.

Acknowledgments. We wish to acknowledge the
contribution of Richard Eddy in compiling the HI-
PLEX-I design document. Glenn Brier also contrib-
uted to the experimental design. Dr. Sumedha Sen-
gupta prepared the randomization sequences, and Jack
McPartland supervised their execution in the field. The
Learjet seeding aircraft was furnished and operated by
Colorado International Corporation, while the Uni-
versity of Wyoming provided the King Air cloud phys-
ics aircraft. The successful conduct of HIPLEX-I is
due in considerable measure to the accomplishments
of these two organizations in developing the equipment
and operating procedures for the project aircraft. A
special acknowledgment goes to the many unnamed
members of the HIPLEX-I field crews, whose dedi-
cation and competence helped make the conduct of
the experiment a success. Barbara Brown was the an-
alyst who "boxed" the radar echoes as described in
Section 7; she, John Middleton, and David Priegnitz
carried out the computations of the radar response
variables. We are grateful to Paul Lawson for supplying
Figs. 2, 3, and 4.

Preparation of this manuscript was supported by
the Division of Atmospheric Resources Research, Bu- '
reau of Reclamation, U.S. Department of the Interior,
under Contract 8-07-83-Y0009.

APPENDIX

Synopsis of Procedures for Calculating Primary
Response Variables

In specifying the locations from which data were
taken to determine the aircraft response variables, three
navigational "pointers" were used. They were defined
as follows:

Pointer I: The location where the maximum cloud
liquid water concentration was observed on the pre:-
treatment pass.
Pointer 2: The location within a cloud where the
maximum ice-crystal concen~ration was observed on
the first pass after treatment (2-min pass).
Pointer 3: The location of the center of the precip-
itation shaft as first encountered at the + 100C level.

Each pointer provided a floating frame of reference,
in which integrating the displacements determined
from the aircraft heading and true airspeed to zero
brought the aircraft back to the same point in the air
(that is, drifted with the environmental wind). On the
passes from two to about 17 min after treatment, the
cloud physics aircraft attempted to return through the
location of Pointer I. Thereafter it attempted on each
pass beneath the cloud to go through (in order of prioJr-
ity): the center of a radar echo, if one appeared; the
center of the visible rain shaft, if any; Pointer 3 aft(~r
it had been set during a pass through the rain shaft
beneath the cloud, which subsequently disappeared;
or the vertical projection of Pointer I. Pointer 2 was
used to determine when to expose a decelerator slide
on the cloud penetration about 5 min after treatment.
The flight path segments specified for computing
each variable were based on estimates of the diffusion
of the seeding plume that are discussed in Appendix
D of Bureau of Reclamation (1979), with allowanc(:s
for possible displacement of the seeding track from
the path of the cloud physics aircraft on the pre-treat-
ment pass and for wind shear between flight levels. To
make the definitions of the response variables fully
objective, default values were specified to cover situ-
ations where the measurements did not produce a value
of a particular response variable in accordance with
the established criteria. The default values were chosen
so that they would tend either to have no effect 011,
or to reduce, the separation between seeded and non-
seeded cases in the MRPP analysis. The individual
variables were calculated as follows:

CIC2: The average number concentration of ice
crystals observed by the Particle Measuring Systems
(PMS) 2D-C probe (depolarization channel) during a

~