WMOD00478

• Title: Field Experimentation in Weather Modification
• Date: 3/1/1979
• Weather Modification - Doc Type: Report

---

~-
~

- ---,.r' H..~~ '" .--,_ ,

Dawkins qnd Scott: Field Experimentation in Weather Modification

73

E-daYs, respectively. Thus, the category W-days is ob-
jectively defined by the direction' of the pibal ,winds at
various levels from 10 A.M. to midnight of the experi-
mental day chosen to conform with Braham's wind
criterion as interpreted by us with the aid of meteorol-
ogists. There was never any effort made to search out
the category of days where seeding produces the most
negative effects, as stated by Braham, although such an
effort would not be unreasonable.
We are indebted to H.C. Chin and M. Neiburger for the
stratification of the experimental days into air-mass days
and frontal days. Actually, all four classifications tried
turned out to be very effective: the stratum of experi-
mental days not conforming with Braham's winds aloft
hypothesis and the stratum not conforming with air-
mass days always showed highly significant or significant
decreases in precipitation ascribable to cloud seeding.
The stratum conforming with Braham's hypothesis
showed little or no effect of seeding. The middle panel of
Figure B illustrates the striking difference between the
effect of seeding on Br-days and on not-Br days. The
lower panels contrast lV-days and E-days. On E-days
the loss in 24-hour precipitation ranges from 63 percent
to 43 percent per experimental day, all losses significant
or highly significant. On W-days, no ring shows a signifi-
cant change. For the category air-mass days, each ring
also shows a negative effect ascribable to seeding, ranging
from a 61 percent decrease to a 28 percent decrease (not
significant in Ring E), with the entire area showing a
significant 43 percent decrease. Braham's hypotheses as
to which type of days would be better for seeding appear
to be correct, in the sense that Braham-days are much
less worse candidates for seeding than not-Braham days.
.The next question was: Where are these apparent
negative effects coming from? This is a study that re-
quires close cooperation between meteorologists and'
statisticians. A glance at Figure A suggests that the some-
what less pronounced loss due to seeding which is ap-
parent in Ring E might be due to the extensive Missouri-
Mississippi River systems flowing along more than half
of the ring, perhaps connected with the heavy industry
there and its inadvertent seeding effects. A preliminary
study partially confirms this suggestion, but there are
insufficient data for a complete study, and there are
other possible explanations, such as downwind effects.
A study of upwind-downwind, as well as sidewind,
effects was undertaken using hourly precipitation data.
Hourly data are available only from recording gages, but
they permit a study of effects over time as well as in
extent. As shown in Figure C, the entire circle was split
into four sectors and two rings, each 90 miles wide. The
four subareas, Downwind 90-180 miles, Downwind 0-90
miles, Upwind 0-90 miles, and Upwind 90-180 miles for
the two pairs.of weather types E-days and W-days, are
shown in the upper panels of Figure C, and for Frontal
and Air-mass days in the lower panel. The solid lines
show three-hour moving average hourly precipitation
plotted each hour of the experimental day until noon of

"'~-_"""'~_ .,_"...L':.~;~."";";":;"'~_, :,_d....;.;..c,,-':......k:;.,:....:..,.,

the next day; the dashed lines give the same information
for not-seeded experimental days. On E-days, the de-
crease ascribable to seeding is easy to see. It is less striking
but still apparent on air-mass days. We see immediately
that the late afternoon and early evening buildup in
precipitation, which is the chief contributor to precipita-
tion on not-seeded days, just does not materialize on
seeded days of the E category, no matter which subarea
is under study. Less pronounced effects are seen for air-
mass days, but the decapitation of storm buildups per-
sists. Why does 'Seeding do this? This is more a question
for the meteorologist to study, but it does require co-
operation with statisticians.

C. Comparison of Average Hourly Precipitation
from Midnight at Start of Experimental Day to Noon
of Following Day (Three-hour Moving Averages In
900 Sectors Downwind and Upwind, for E- and W-
Days, and for Frontal and Air-Mass Days)

.t2 l' "" :~-:::.,,----. [~. "". .^
'0 -....' \
aOI " OOWNWIND \
~ ~ j-L-\,L 90-1_ -',
a. "o<J - ~ - ...._" ..
o . ., .

'Oll~'/'''\ ,/'" ~::..I/t)
,,-...,' \,'
, "J

~l . 1\ . {\, -~

, \' O.90mi.
~""" ! -- -\,- -~}
-'0~
o .

.Oll 1\ /\" ,~-, lOu~I.~'
I \ I ,_-"", '.. UVfRI.

~
00 12 24 56
hours

1\
" \ ,..-,
~V'A .
"r-J;i 'V r
,/ '.......",,"
---'

o

24

36
hours

12

FRONTAL days
1',
l \
\
\
\
\
\

DOWNWIND~AIR-MASS days

~-l8Omi. ,...._
~,,' ~-',...)
-_....-~ '--/
. .

.02l
c
.2
i.~
~ -
Cl.

o

l ,"
I \
DI, ~/\ .
"-f,F- - ~
--,'
o' .

l ... .
.llI ....../"C'\..__ ,/-'/ A'
-:,~ '-~ \

o . .

,o'l~/\ ,,.'
I '-
-..J '"

o . . .
o 12 24 36
. hours

~~^\ ~ t,_
0-10"",. I """'......... I

~
"l' .

UI'WIND
O.tOri

UI'WIND
10 - IIlOmi.

o

12

24 36
hours