WMOD00419

• Title: Convective Rain Rates and their Evolution During Storms in a Semiarid Climate
• Date: 8/8/1984
• Weather Modification - Doc Type: Report

1606

MONTHLY WEATHER REVIEW

VOLUME 112

'C"
.J:.

25 dBz ATI THRESHOLD (149 cases)

NDCMP SUMMER 1980 DATA

30 dBz ATI THRESHOLD (117 cases)

..
E
.ll:
"in 1 au
....
<t 104.0
>
ex: 103.5
w
I- 103.11
Z
.... 102.&
<t
ex: 102.0
CJ
W 101.5
I-
Z 101.0
W 100.5
:E
i= 10'
<t
W 10.0.5
ex:
<t A

0 0
~~ 6~
o 0
;04
4
5 7 9 11 13 15 17 >17

- 2

:~: , z;. " :/.,

~!!~~~!J(
2 ~ ~ G' ~ ~ ~ . ~~
o 0 O~2 --: ~ -3 5
. . . '\"4.1/~ . .
2 2 2 2 !_2-2~ ~8
22222222\1l
3 5 7 9 11 13 15 >17

AVERAGE RAIN RATE (mm hr-')

B

FIG. 4. Two-dimensional frequency plots (number of cases) of R versus A TI.
Two A TI reflectivity thresholds were considered, 25 dBz (A) and 30 dBz (B).

(up to 10 km2 h) were eliminated. In Fig. 4B, the
. distribution is so scattered that Ii. cannot be at all
considered independent of AT!. We may recall here
that this analysis is performed in order to model echo
clusters for areal rain volume estimation purposes.
The scheme mayor may not work if the" rainfall of
seeded versus nonseeded clusters has to be compared.
That could be the subject of another study.
To confirm the fact that large scatters of the rain
rates are generated by small clusters, several regression
equations were determined (for the summer 1980
data) using a 25 dBz A TI threshold, after points with
small A TI values had been discarded. Table 1 sum-
marizes the results. They indicate a decrease in the
mean and a considerable decrease in the standard
deviation. and the range of Ii. values as the clusters
with small ATI values (<10 km2 h) are deleted. In
such a situation, Ii. = 4.0 mm; thus, most of the
scatter in the rain rate distribution is generated by
very small and short-lasting clusters. By considering
only clusters with A TI ~ 10 km2 h, the standard

deviation and. the range of the rain rates are consid-
erably reduced, from 7.68 to 1.55 mm h-I, and 57.59
to 11.39 mm h-I, respectively. The relative frequency
of the mode (3-5 mm h-I) increases from 48 to 65%.
Thus the 25 dBz A TI threshold (besides the 20 dBz
threshold previously noted) seems to become an
acceptable threshold for rain volume computations
through a given rain rate if small clusters with A TI
< 10 km2 h are discarded.
Earlier in this section, it was mentioned that a 25
dBz threshold is less satisfactory than a 20 dBz.
threshold in rain volume computations through the
A TI technique, if all clusters are considered. Later, it
was noticed that if the troublesome small clusters are
deleted, the 25 dBz threshold also becomes satisfac-
tory.
For a low radar reflectivity threshold (i.e., 20 dBz),
much echo area that is not likely to be associated
with any appredable amount of precipitation on the
ground will be included in the calculation (Ionescu-
Niscov, 1982; Doneaud et al., 1984). A good com-

TABLE 1. Average rain rate (mm h-I) regression parameters for NDCMP 1980 and 1981 summer data determined
using a 25 dBz A TI threshold after points with small A TI values had been discarded.

1980 1981
A TI cutoff (km2 h) None 10 25 None 10 25
Number of clusters in
the sample 163 149 131 107 587 583 505 420
R (mm h-I) 5.48 ' A.48 4.09 4.10 4.97 4.73 4.88 5.07
Standard deviation
(mm h-I) 7.68 4.44 . 1.55 : l.35 4.84 2.19 2.22 2.23
Range (mm h-I) 57.59 53.01 11.39 11.12 98.20 16.44 16.44 16.22