WMOD00439

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

504

JOURNAL OF CLIMATE AND APPLIED METEOROLOGY

VOLUME 23

estimates wherever. the null hypothesis was rejected,
is evident from the statement of the physical hypothesis
in Table 2.

6. C~)Oduct of the experiment

A partial simulation of the planned experiment was
carried out using data obtained on HIPLEX-I type
clouds during the preliminary field investigations, along
with estimates of probable seeding effects (Mielke et
al.. 1984). The results suggested that, for the three
response variables considered (CIC5, PIC8, TFPI),
about 50-150 test cases would be needed to reject the
null hypothesis. Section IV of the design document
(Bureau of Reclamation, 1979) estimated 30-45 test
case clouds per summer, so that an experiment of 2-
3 years duration was anticipated. Regrettably, nature
failed to match the climatological estimate and so a
longer experiment was needed. It IS equally regrettable
that budgetary mandates terminated the experiment
before it could be extended beyond two years.
The HIPLEX-l experiment began in the summer
of 1979 and continued through the summer of 1980.
An initial start was made early in 1979, but it quickly
became evident that the experimental procedures were
not completely satisfactory. The use of the HIPLEX
Z-R relationship for subcloud regions that contained
substantial amounts of precipitating ice particles was
regarded as questionable. Therefore a provision was
added to constrain the RERC calculation to cases where
the center of the radar beam at the lowest tilt angle
(0.70) was at or below the + WOC level; the presence
of any ice particles would then be less likely to affect
the calculation. ~ome details of the flight procedures
were clarified; some computational procedures related
to where average values were to be taken were changed;
and some provisions further specifying default values
were added. Perhaps the most important lesson from
this experience was that, before launching such an ex-
periment, at least one trial case should be completed
and run all the way through the data reduction process
in order to uncover any unsuspected deficiencies in
the procedures. .
After the procedures were modified, the experiment
was restarted and 12 test cases were recorded during
the latter half of the summer. The summer of 1980
was extremely dry in eastern Montana and only eight
additional test cases were logged that year. Especially
in 1980, the measured liquid water concentrations of
many of the candidate clouds were not sufficient to
meet the selection criteria. Numerous additional clouds
not meeting these criteria were also studied each sum- .
mer, but not as part of the randomized seeding ex-
periment. HIPLEX-I was pre-empted by the Coop-
erative Convective Precipitation Experiment (CCOPE)
in 1981, and then terminated. As a result, only 20
total cases are available for analysis, with seven clouds
being assigned to Class A-I and 13 to Class B.

a. Experimental procedures

The HIPLEX-I experiment was carried out within
the framework of a general operations plan covering
all aspects from forecasting to debriefing of the missions
(see Appendix C of Bureau of Reclamation, 1979). In
the conduct of HIP LEX-I, the key elements were thl~
seeding aircraft, the cloud physics aircraft, and thl~
SWR-75 radar. The seeding aircraft was a Lea1jet, whik
the cloud physics aircraft was a King Air. Appendix
G of the design document (Bureau of Reclamation,
1979) describes the operating capabilities and instru..
mentation of the aircraft. All of the instrumentation
was calibrated according to the normal procedures; in
addition, the aircraft instrumentation was calibrated
during tower fly-bys and checked in comparison flights
involving other project aircraft.
Detailed operational procedures were spelled out
for the aircraft (see Section VIII of Bureau of Recla-.
mation, 1979). In brief, the seeding aircraft determined
the cloud top temperatures and rate of growth, and
conducted the treatment runs at or near the cloud tops
(Fig. 2). The cloud physics aircraft made an initial
pretreatment pass through each candidate cloud to
determine whether it met the selection criteria (see
Section 3). If it did, the cloud was treated and the
cloud physics aircraft then made a series of passes
through it to look for the development of ice crystals
and the initial precipitation particles.
The first two post-treatment passes were targeted
for 0.3 km below treatment altitude, or usually around
the -80C level, at about 2 ~in and 5 min after treat-
ment, respectively. (The treatment time was taken as
30 s after activation of the selected dry-ice dispenser,
to allow time for the pellets to reach the intended
region of the cloud.) They were made in a direction
perpendicular to the treatment pass (Fig. 3) in an at-
tempt to intersect the seeded curtain. Subsequent passes
at about 8, II, 14, and 17 min after treatment (or
until an echo appeared on the aircraft 3-cm weather
radar) were targeted for the -50C level (Fig. 4). Mean-
while, the seeding aircraft continued to monitor the
cloud top for 10 min after treatment, to observe any
changes in its height or temperature. It also collected
data on ice crystals in the top of the cloud, for the
physical evaluation, when the cloud top remained
above the aircraft's lowest allowable flight level.
The cloud physics aircraft then descended to the
+ WOC level to monitor the precipitation falling from
the cloud. It made a series of orthogonal passes, at
about 3-min intervals, through the rain shaft. Mean-.
while the SWR-75 radar conducted a continuous series
of 3-min volume scans beginning at an elevation angle
of 0.70. The next elevation step was 1.00, and sub-
sequent increments were 10 for nearby test cases and
0.50 for cases beyond about 80 km range. Observations
continued until the termination of the test case, 40
min after treatment. Close time lines were specified