WMOD00412

• Title: Final Report on Utah Cloud Seeding Experimentation Using Propane During the 2003/04 Winter
• Date: 3/1/2005
• State: UT
• Weather Modification - Doc Type: Report

I

I

I

However, such cases were rejected whenever partitioning excluded EUs with HAS median wind
directions greater than 270 deg true. That partition was based on investigations of previous plume tracing
experiments discussed in Sec. 4, which indicated that HAS winds with a northerly component w(~re
unlikely to result in the seeding plume reaching the target gauges.

I

The noisy icing sensor data from the first stonn was examined by one of the authors soon aftenvard. He
had worked with data from the same model sensor collected from a different mountain location, and
realized that something was wrong. But sensor mismounting never occurred to him as mounting
instructions would be expected to accompany the instrument. Moreover, nothing was mentioned by the
field personnel who mounted the instrument. The author requested that the declaration threshold be
changed from 34 to 70 Hz to reduce false declarations of EUs with the noisy data. That was dom~ on 16
December.

I

I

The second stonn passed on 21 December. Data were still noisy but the higher threshold limih~d the
number of declared EU pairs to only three, all with typical southwest flow at the HAS seeding location.
The icing sensor heater switched on only during one of the six EU periods. But some Rosemount icing
sensor trips were recorded at the SIR station, only 46 m higher in elevation than the HAS. A photo taken
near the HAS documented that the area was in cloud about midway through the six EUs. While the
Goodrich icing data were still suspect, SL W cloud was observed in the vicinity by other means during the
second stonn and seeding potential likely existed.

I

I

I

A series of digital photos of the HAS instruments were received by the authors via email on 22
December. The problem with the icing sensor rod being tilted 45 deg off vertical was immediately
recognized. The sensor was remounted, with its mounting ann pointed southwest-ward into the
prevailing wind, and the sensing rod vertical, on 24 December before the third storm arrived. The cooling
vanes below the sensor assembly were pointed northeast rather than southwest because of concerns that
rime ice and snow buildup on them might adversely affect observations. Cooling vanes were unnecessary
during winter use but were part of the sensor assembly. The EU declaration threshold was left at 70 Hz
until more confidence was gained in the sensor's data.

I

I

I

Goodrich icing data during many EUs beginning with 25 December 2003 through 07 January 2004
appeared very reasonable. Frequencies decreased from the approximately 40,000 Hz no ice value to
39,400 where the heater would turn on. The frequency would then return to the no ice value over about 3
min after which it would began to decrease again in a generally linear fashion. A few exceptions
occurred when icing appeared to cease before the end ofthe 40 min EU, followed by a gradual increase in
frequency interpreted as sublimation from the sensing rod. No further stonns with EUs passed after 07
January until early February. Finally, on 05 February, the EU declaration frequency was set back from 70
Hz to the original 34 Hz over 10 min duration value as the noisy icing data problem had not reappeared
since the remounting of24 December. But Murphy's Law remained in play.

I

I

A large storm passed on 07-08 February, producing over one inch of SWE at some gauges and over two
feet of new snowfall. Goodrich frequencies were again noisy during most but not all the 12 EUs which
resulted, raising the likelihood offalse EU declarations in some cases. Yet, for example, the 7th EU plot
(not shown) was a classic case of expected sensor frequency behavior as shown in Fig. 12. HAS winds
remained from the southwest to west-southwest throughout the 6 pairs of EUs. While snowfall was heavy
and winds were strong, the stonn passage of25-26 December produced over twice as much snowfall with
stronger winds without apparent problems for the icing instrument. Close examination of photos of the
instrument's mounting plate and horizontal arms showed they were higher than necessary. It was
suspected that rime ice buildup on the mounting structure and icing sensor housing may have at least
partially blocked the sensing rod. Visibility was certainly limited by cloud according to webcam views of
the HAS area, and other icing sensors in the experimental area reported SL W presence. But the Goodrich
sensor was again producing noisy data during much of the 07-08 February stonn.

I

I

I
I

26

I