WMOD00474

• Title: The Evolution of Raindrop Spectra in Warm-Based Convective Storms as Observed and Numerically Modeled
• Date: 12/12/1978
• Weather Modification - Doc Type: Report

2306

JOURNAL OF THE ATMOSPHERIC SCIEN'::ES

VOLUME 35

Fig. 2b shows growth rate and growth time distribu-
tions for the 32% of all first echoes which exhibited
vertical development. Growth time is shown to be quite
short with 58% of all echoes that grew, having ceased
growth 5-10 min after detection. The growth rates
shown in Fig. 2b are quite low-typically 2--4 ms-I. This
observation may be the result of the coalescence rain
process in that (to a first approximation) the rates are
an algebraic subtraction of particle fallspeed from up-
draft speed. Thus, when ice particles constitute the
first echo, updraft speed should be approximately 2-3
m S-I greater than echo growth rate, whereas raindrops
(contributing to Z) are more likely to be in the 5 m S-I
range. It follows that updraft speed is typically 7-9
m S-I if the echo is primarily composed of water drops
as indicated for most of these data.
Echo mergers were tabulated for the period of ob-
servation, and it was found that approximately 37%
of all echoes were involved. The majority (58%) of
echoes that merged did so within 10 min of first echo,
with virtually no mergers beyond 30 min, The physical
significance of this statistic is that it places a lower
bound estimate on the frequency of multi cell storms.
It was found that multicell storms also evolve from a
succession of cells within a previously existent echo.
It follows that multi cell storms are a very common
phenomenon in west Texas during July and August
and may be the single most important storm type with
respect to total rainfall production.

4. Preliminary assessment of data

a. Context of the observations

On 13, 14 and 15 August 1975, raindrop spectra from
several storms were sampled 300 m below cloud base.
These storms were simultaneously viewed by radar
during the echo life history. The characterististics of
echoes during this period differed from the seasonal
norm in that early echo heights were higher and growth
rates were larger than the values indicated for the entire
period of radar observations. First echoes typically
appeared at -lOoC with several colder than - 200e.
Most echoes indicated at least some vertical growth
subsequent to detection, with correspondingly longer
lifetimes on the order of 1 h. These factors suggest ice-
phase involvement early in the precipitation history.
Cloud-base temperatures ranged between 12 and 150C
during aircraft sampling. During the period of aircraft
sampling echoes exhibited the typical single-cell
structure; however, one echo subsequently evolved
into a large multicell storm. Horizontal gradients of
radar reflectivity factor in all cases were strongest on
the western or southwestern boundary of the echoes.
When significant updrafts were encountered by the
aircraft, they were along the western or southern
flank of the rainshaft with magnitudes of 4-12 m S-I
::1::2 m S-I. The southwestern flank is also the upwind,

upshear side of the storms, although the maximum wind
shear within cloud depth was only 2X 10-3 S-I. Weaker
gradients of reflectivity factor were located along the
northern and eastern echo boundaries where vertical
air motion was either too weak to measure or down-
drafts of 4-6 m S-I were present.
The rather general characterization of cloud and echo
structure discussed to this point serves the purpose of
placing detailed drop spectrum measurements into'
context. The important points with respect to that
context are the warm cloud-base conditions, nonsteady
single-cell structures with updrafts located on the up-
wind flank of the storms. While the ice phase was
heavily involved in the precipitation process, a large
depth of cloud (? 1.5 km) was warmer than aoe. To
this extent the authors consider the rain falling below
cloud base to be somewhat isolated from" cold rain"
processes, with condensation, coalescence and breakup
mechanisms dominating drop spectra results.

.P

b. ExamPles of spectra

While it is not possible to present a few drop-size
distributions as being "representative" of the entire
data set, it serves an illustrative purpose to show ex-
amples. Fig. 3 shows eight selected spectra which
indicate a variety of distribution forms encountered.
In each panel one high and one low rainfall rate spec-
trum is indicated. Each spectrum has been averaged

...
'E
o

-2

-3

-2

-4

~

o
z -6
'---'
e

(!)
3 -2 \\ Obs.(M-P) ()
\ \ R:21.5(31.5) C
\ \ dBZ: 47 (44)
~\
\

\ \

\
\ \
R:4.2(B,B) \
dBZ'38 (33) \
-6
o 1.5 3.0 4.5

-4

-4

-5

-5

-6
o
D(mm)

FIG. 3. Examples of drop spectra for high and low rainfall rate
regions of storms. Note the large deficit of small drops and excess
of large drops compared to Marshall-Palmer distributions (dashed)
plotted for equivalent rainfall rates.