WMOD00484 - Laserfiche WebLink

Home Browse Search

OcTOBER 1988 SUPER, BOE, HOLROYD AND HEIMBACH 1149 the delay times from the counter used over the Grand Mesa ranged from only 27 to 34 s for total counts of 500 and 50, respectively. In any event, each estimated plume position should be approximately centered on its actual position so that the analyst knows where to search for enhanced IPC due to seeding. More serious underestimation of the AgI plume can come about from depletion of the AgI itself through scavenging by cloud droplets and by ice particles. Ice particle scavenging is believed to be limited (e.g., Radke et al. 1970), although field observations are scarce. As Flossmann et al. (1985) point out, the efficiencies with which aerosol particles are scavenged, either by nucle- ation or impaction with drops, are not well known. I: Whatever the mechanisms involved, the apparent rapid depletion of AgI shown in Parts II and III in SLW cloud and/or in high IPC zones (which implies SL W at an earlier time) can seriously degrade the acoustical counter's ability to delineate the entry edge of a seeded zone. Nevertheless, finding the core of the AgI plume located within a zone of high IPC strongly suggests that AgI seeding caused the ice particle en- hancement. Conversely, failure to find any AgI within such a high IPC zone might suggest that the IPC max- imum was a natural phenomenon. e. Seeding generators All ground-based seeding was done with the Mon- tana State University generator developed and used for the Bridger Range Experiment (Super and Heimbach 1983). A 3% by weight AgI solution, complexed with NH4I in acetone, was consumed at 30 g AgI h -I. It was sprayed through a hypodermic needle into a jet of propane, where the solution was atomized prior to consumption in the bum chamber. Calibration of this system by the CSU Cloud Simulation Lab (Garvey 1975) showed an output of 6 X 10 13 ice nuclei g -I at -lOoC for natural tunnel draft. This is the appropriate curve to use because all seeding was done at sites with very light winds. Output increased to 3 X 1014 ice nuclei g-I at -l2OC and 7 X 10 14 by -160C. All aircraft seeding was done with the instrumented Turbo Commander upwind of the Grand Mesa. An Aero Systems, Inc. generator was used with a 2% AgI- NH4I-acetone solution, burned at about 80 g AgI h -I. The solution was sprayed into a bum chamber where it was mixed with air and ignited. Generator tests by the same CSU Lab indicated a yield of 5 X 10 14 ice nuclei g-I effective at -lOoC, increasing to near 1015 ice nuclei g-l for temperatures between -12 and -20oe. .f! f Suiface-based wind and icing rate detectors Rosemount icing rate detectors designed for tower use were operated during the experiments, in close proximity to heated anemometers and wind vanes. The icing rate detectors were calibrated in a small wind tunnel in the same manner as the airborne unit dis- cussed in section 3c. Several series of tests were run with (~ach sensor, and individual values were within :t 15% of the series means, which were typically near 0.07 g. The mean mass required to initiate a deicing sequence, its cross-sectional area, and the mean wind speed were used to estimate SL W contents, as discussed by Boe and Super (1986). This assumes unity collection efficiency for cloud droplets encountering the 6.35 mm diameter rod of the sensor. Underestimation of actual SL W contents is therefore likely when small droplets are present and winds are light (Hindman 1986). A more serious underestimate is believed to have occurred due to the "lifting" of cloud base above the sensors, at least above the flat-topped Grand Mesa. It was visually observed that cloud base was often above the 70 m tower on the middle of the mesa even when considerably lower on the windward side. A vertically pointing microwave radiometer (not available during the experiments discussed in Part III) detected SL W with more than three times the frequency of the nearby tower..mounted icing rate detector on the Grand Mesa (Super et al. 1986). Similar comparisons were not done on th(~ Bridger Range. Consequently, while detection ofSL W by a tower-mounted probe is a clear indication of the presence of supercooled cloud, the converse is not necessarily true. Tower-mounted Hydro-Tech wind sensors were used on the Grand Mesa. They were robust and internally heated, and performed well under icing conditions. The wind speed sensor on the Bridger Range was operated above heat lamps, whicll were usually sufficient to keep it ice free, although some instances of riming were ob- served. The Bridger Range wind direction sensor was unheated, but no instances were observed in which it did not respond because of riming. Horizontal winds above the surface of the Grand Mesa were measured with an acoustic sounder at 30 m resolution up to 570 m agl. These winds, and es- pecially those measured by a small network of unheated surfacl~ stations around the mesa, generally exhibited significant differences from the ambient winds mea- sured by National Weather Service rawinsondes re- leased from Grand Junction, about 45 km to the WNW. Such terrain-caused distortions are important to the problem of targeting the seeding materials, as discussed in Holroyd et al. (1988). Flow distortions over the Bridger Range were only qualitatively ob- served. g. Sw:face precipitation measurements Surface precipitation measurements were made only on the Grand Mesa. A network of 7 Alter-shielded high-msolution gauges were operated through the ex- perimental period. In addition, snow particle charac- teristics were recorded at the "Snow Lab" on the Grand