WMOD00445

• Title: Implications of Time Dependent Convective Z-R Relations for Radar Precipitation Analysis
• Weather Modification - Doc Type: Report

,(.
,

II. .1; :

"'
~

j"
o
-

-
~'jl
10
* -
*
~""
u
-....

.....................

.

6

..
I
o
-
100

tOl
R (MM/HR)

102

Figure IlIA

relationships of buildin~ and decaying rain shafts
in this data set, which was gathered in a very
warm, wet sounding with a cloudbase temperature
of 16 .C.

From the data in figure IV we can deduce the mean
Z-R relations:

Z . 763Rl.365
Z . 400Rl.560

(early)
(late)

(3)
(.4)

The corresponding model';'predicted Z-R relationships
show the same relation to each other but to a mu.;h
larger extent, as given below:

Z . 89l2RO.875
Z . l77Rl.375

(early)
(late)

(li)
(tl)

Thus, the model \~ould associate an "early" 3D-dB::
echo with a rain rate of 0.1 mm/hr and a "late"
30-dBz echo with a rain rate of 3.50 mm/hr. The
mean observational Z-R relations would predict the
rain shafts at 30 dBz to be 1.2 mm/hr (early) and
1. 8 mm/hr (late). If we use extreme rather than
mean values of A and Bfrom fi~ure IV, then the
rain rates predicted at a constnnt 30 dBz become
0.7 mm/hr (early) and 2.6 mm/hr (late).

3. CONCLUS IONS

The one-dimensional model which has inadequate
horizontal mixing and wind-shear overpredicts the
temporal evolution of the Z-R relation and, of
course, cannot address the spatial variations at
all. Nevertheless, sampled rain shafts confirm
that the temporal evolution predicted by the model
does indeed exist in the real world. The implica-
tions for rainfall estimation by radar (where only
. the sixth moment of the drop-size distribution is'
knolon) are profound in that the observed variabil,.
ity of the drop-size distribution is somewhat sys..
tematic in both time and space. Thus, different
Z-R relations apply at different times and loca-
tions within the same rainshaft. While aircraft
sampling-vol~~e considerations do not permit com-
plete separation of these spatial and temporal

N
o
-

o
o
-
*
C'J toO

tol
R (MM/HR)

102

Figure I1IB

'variations, it is clear that they do exist and
should be considered before making use of radar
:reflectivity to estimate convective rainfall rates.
rn practice this would be quite cumberso~e but not
:impossible to execute, since information as to the
llge of the rain shaft and position relative to
4~nvironmental winds is known to some extent.

!f the effect of a seeding treatment were to
jlncrease the duration of the mature stage of a rain
lihaft with respect to the duration of the building
litage, then the rainfall calculated with a constant
~:-R relation from radar data would underoredict the
!ieeding effect. On the other hand, a "paired"
c:loud experiment where the unseeded cloud tended
t:o be sampled later in its life cycle could easily
s;how 100 percent seeding effects where in fact none

o
(IJ

co

\0
CO . .
D D O'
"1' D
D
D
D
N
D
D
0
I
0,0 500. 0 1000 ' 0 1500 . 0
A
Figure IV