WMOD00399

• Title: WMO Training Workshop on Weather Modification for Meteorologists - Lecture Notes
• Date: 12/1/1979
• Weather Modification - Doc Type: Report

- 54 -

membrane filter which retains the particles in the air. The number of ice nucl~i on the
filter is then determined by placing it in a box held at a given supersaturation and
temperature and counting the number of ice crystals that grow on the filter (Gagin and
Aroyo, 1969). Problems encountered with the filter method are 1) the volume efiect which
means that a relative decrease in IN-concentration is measured with increasing :volume
I
sampled. The reason is that during development of filters the large number of ~articles
other than IN act as vapour sinks. It must be remembe2ed that of the aerosol pa~~icles
with a concentration of ~ 105 cm-3 of surface air, 10 cm-3 are CCN and only 10 - 10-3
cm-3 act as IN at T = -200 C. 2) Only freezing and deposition nuclei can be activated at
supersaturations with respect to ice Si - 1 above or below water saturation respectively.
Contact nuclei may also be lost because they can pass through the pores of a ~ilter with
nominal diameter of 0.05 V.
Some results of world wide measurements of ice nuclei are shown in Fig. 8~ On the
average the number of ice nuclei per liter of air active at a particular supercooling ~T is
Ni = a exp(b~T). With a = 1o-5(liter-1 of air) and b ~ 0.6 this equation gives ~ 1 active
ice nucleus per liter at -200 C and an increase of about a factor of 10 for every 40 C fall
in temperature. For deposition nuclei the supersaturation with respect to ice is probably
the more important parameter than supercooling and their number can be approxi~ated by a
relation similar to the one in eq. (7): NO = A(Si - 1)a with A and a being con~tants. Values
of a range from 3 to 8 depending on the IN-source. (a = 3 for air in rural NE qolorado and
8 for polluted air in St. Louis).

"'s

W
..J
u
~IOOO
C)
z
i
a::
o 100
't
...
~
....
o 10
a::
...
CD
:::I!
=>
Z

IEURO~

::::RICA ,~~{

'SO'...RICA rr
6AUSTRAUA /1f
136/

" .. /~1

I.
I
1

Fig. 8~ Range of median number corlcentration
of IN as function of temperature for various
geographic locations: 44 stationsJ The dashed
line represents NIN = 10-5 exp(0.6H). (From
Bigg and Stevenson, 1970).

/

-10 -15 -20
TEMPERATURE (.C)

4. Raindrop size distributions

Precipitation over much of the world reaches the ground as rain. Its most Icommonly
measured characteristic is rainfall rate, measured with a recording raingauge ~uch as the
one described below, or simply the daily rainfall amount, measured with a raingauge. For
purposes of a weather modification experiment it is often preferable to measure precipita-
tion by means of a radar, in order to minimize the sampling error. For the cor~ect inter-
pretation of the radar reflectivity factor Z and for the calibration of the radar, the
cloud physicist and the radar meteorologist are therefore interested in the size frequency
distributions of the raindrops or raindrop spectra. The existence of a unique ~elationship
between rainfall rate R (mmh-1) and the drop size distribution N(O) would impl~ that the
radar echo intensity Z (= !N(O)06dO) and R were also uniquely related. In fact; Marshall
and Palmer (1948) found a remarkable relationship: When plotting the number of 'drops per
unit volume and diameter interval they obtained distributions which could be approximated
by an exponential law (see Fig. 9)

N(O) = No exp(-AD)[m-3mm-1]-

where No and A are the intercept of N(D)
depends on rainfall rate, A = 4.1R-0.21.
parameter No is a constant given by No =

(8)

.

at 0 = 0 and A(mm-1) is the slope par~meter which
Remarkably, they also found that the intercept
8000 m-3mm-1 . Using these values a classical Z-R