WMOD00298

• Title: Quantifying Ice Nucleation by Silver Iodide Aerosols
• Date: 5/1/1990
• State: CO
• Weather Modification - Doc Type: Report

I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

33

for the initial time to detect a signal. This maximum error is one data

record (L:.t - l5s in this study), and is equivalent to an error in

nucleation temperature of 0.250C in a 2.5 m s-l equivalent updraft

experiment. Studies are underway to implement a temperature dependent

transfer function based on Morrison's data and further DCC pulse

nucleation experiments (Rogers and DeMott, 1990).

The results of using the deconvolution procedure on some

experimental data are shown in Figure 3.6. In this experiment, ice

nucleus aerosols were injected directly into cloud at -80C. After

normalizing the ice crystal numbers measured falling from the chamber

in time to the total number of aerosols inj ected, these values were

smoothed with a binomial filter; typically a nine point filter was

used. This was done to keep the technique numerically stable. The raw

ice signal and the smoothed signal are shown in Figure 3. 6a. The

deconvoluted signal produced after the application of the transfer

function is shown in Figure 3.6b. The deconvo1uted signal was sometimes

smoothed with a 5 point binomial filter prior to further analysis for

nucleation rates. This smoothed signal is also shown in Figure 3. 6b.

Whenever smoothing was performed, the initial signal was always

truncated so as not to create artificial data before ice formation was

initially observed.
The primary measurement of humidity in the experiments performed

was made using two optical condensation type dew point hygrometers. A

prototype infrared transmittance hygrometer manufactured by the Ophir

Corporation (Nelson, 1982; Kahan, 1989) was also installed in the

chamber. This particular sensor and its performance characteristics

have been described by DeMott and Rogers (1989). Its potential use as