WMOD00430

• Title: A Diagnostic Technique for Targeting During Airborne Seeding Experiments in Wintertime Storms over the Sierra Nevada
• Date: 7/7/1988
• Weather Modification - Doc Type: Report

JULY 1988

RAUBER ET AL.

815

crest. These accelerations had no effect on any of the
trajectory calculations. The trajectories that did pass
into region C did so only in the lower regions well
upwind of the crestline (see Fig. 3). To estimate the
wind components in region C, the region was divided
into five additional flow channels of equal pressure
depth. The U-, V-, and w-components of the flow were
then calculated in a manner similar to regions A
and B.
The method to determine the wind fields for target-
ing applies to flow in a stable or neutral atmosphere.
In many storms, convective motions were also present.
In conditions typical of fixed target experiments, weak
convection often occurred, generally embedded in the
larger-scale orographic cloud. The depth, frequency,
duration and strength of these convective motions were
functions of the time and location in the storm and
were, in general, unpredictable.
An estimate of the possible errors that convective
motions may introduce to targeting calculations has
been calculated by King (1984). King studied the effects
of several cloud parameters on targeting in stratiform
clouds containing weak embedded convection that oc-
cur over western Victoria, Australia. He found that
varying the magnitude of a constant updraft in his tar-
geting calculations from 0.0 to 0.5 m S-I, all other vari-
ables staying the same, caused an overall 16% change
in the transit time of a particle from its nucleation
point to the surface. In SCPP clouds, the mean vertical
velocity due to orographic lift encountered by a particle,
based on the terrain slope and typical horizontal ve-
locities, was in the range 0.2-0.4 m S-I. Convective
motions, positive or negative, could impose on these
orographic motions brief periods of vertical velocities
exceeding 1 m S-I. Averaged over the total particle tra-
jectory, the mean vertical velocity including convection
would most likely fall within the range of :to.5 m S-I
of the value for pure orographic flow. Using King's
estimates, it is probable that convection could cause
errors in the total transit time up to 16%. The corre-
sponding displacement of the impact point would de-
pend on the intensity and location of the convection
when encountered by the particle and the horizontal
wind speeds. In SCPP experiments, it was likely that
the variations in seeded particle trajectories imposed
by convection were averaged out by seeding over a
sufficiently long time period with several seedlines.
Typical experiments continued for 1-2 h with five-ten
seedlines placed within the clouds.

b. Microphysics

1) PARTICLE GROWTH RATES AND HABITS

Ice crystal habits and particle axial growth rates are
highly dependent on temperature and supersaturation
(Hallett and Mason 1958; Kobayashi 1961; Magono
and Lee 1966; Ryan et at 1976). Because seeding oc-

curred at a variety of cloud temperatures, a critical
aspect of targeting was to determine fall velocity as a
function of particle habit and size. Accurate targeting
required a reasonable estimate of ice particle growth
rates and fall velocities.
The parameterization developed for targeting during
SCPP operations used the linear growth rates for the
basal (c) and planar (a) crystal axes specified by mea-
surements of Ryan et al. (1976). They found crystal
axial growth rates to be maximum along the a-axis at
-I70C and maximum along the c-axis at -60C. A
distinct minimum in growth rate along both axes oc-
curred at -lOoe. Axial growth rates measured by them
were incorporated in the targeting calculations. The
measured growth rates extend through the first few
minutes of crystal growth. However, measurements of
Fukuta and Wang (1984) extend up to 20 min. Fukuta
and Wang's data suggest that the linear growth rates
reported by Ryan et at apply at times as long as 20
min. Crystal habits were determined on the basis of
temperature and crystal axial ratios. The algorithm is
discussed in following sections.

2) CLOUD LIQUID WATER AND RIMING

Th{: importance of accurately specifying the liquid
water content of a cloud when calculating particle tra-
jectories during seeding operations has been demon-
strated by King (1984). He showed, using model cal-
culations within stratiform clouds, that the trajectories
of cloud and precipitation particles growing by riming
depend primarily on liquid water content, and not on
the precise details of cloud temperature, crystalline
shape, graupel shape or size. Unfortunately, in natural
clouds, the liquid water content and distribution of
liquid are unknown and generally variable in time.
During SCPP, two measurements ofliquid water con-
tent were available on a real time basis, the first from
aircraft and the second from a ground-based radiome-
ter. The relationship between liquid water depth mea-
sured by the radiometer and actual cloud liquid water
content depends on the depth and distribution of liq-
uid-bearing cloud layers. A climatological analysis of
SCPP radiometric measurements by Heggli (1986)
suggested that liquid was confined primarily to the
lowest 1-2 km layer of cloud. An approximation of
the average cloud liquid water content was obtained
operationally by assuming that the liquid water was
distributed uniformly over a kilometer-deep layer
above the radiometer. The average liquid water content
of the cloud (g m-3) would then be numerically equal
to the liquid water depth measured by the radiometer
(millimeters).
Aircraft measurements ofliquid water content were
made at the minimum obstruction clearance altitude
(MOCA) on a descent westward from the crestline to
approximately the melting level. Measurements were