WMOD00484 - Laserfiche WebLink

Home Browse Search

OCTOBER 1988 SUPER, BOE, HOLROYD AND HEIMBACH 1147 the 2D-C probe is about 5 L per 100 m offlight. Anal- yses for concentrations, size spectra, habits and cal- culated precipitation rates were performed according to the computerized techniques described by Holroyd (1987). Particles classified as graupel were assigned masses appropriate for graupel-like snow. Linear par- ticles were considered to be planar crystals viewed on edge because the temperature regimes were not indic- ative of the development of needle or sheath habits. Questionable images, those with centers apparently out of the field of view, and those smaller than 0.1 mm (about 4 pixels long) were rejected. The habits of the remaining "valid" 2D-C images were estimated, and masses and terminal velocities were assigned for pre- cipitation rate calculations, It is noteworthy that Hol- royd achieved reasonable agreement between snow- board samples and his method of precipitation esti- mation, which uses a ground-based aspirated 2D-C probe. This was in spite of artificial increases in ice particle concentrations caused by the aspirator, in one case by a factor of2.4 (Holroyd 1986). The use of an aspirator is not recommended by Norment (1987), a conclusion based on numerical studies simulating flows with water droplets and no turbulence. However, that study did not investigate the response to ice particles in air of normal turbulence. The size spectra were not changed significantly by the aspirator in the Holroyd case study, but more recent work by Deshler (1988) has suggested that the aspirated probe can oversample smaller particles and undersample larger particles. Also included in the King Air data acquisition system was a PMS 2D-P optical array probe. Analogous to the 2D-C probe, the 2D-P had a 200 Jlm resolution and 200-6400 Jlm range. The primary difference between the probes lies in their magnification optics. The sam- pling volume of the 2D-P is about 170 L per 100 m of flight. The ice particle size spectra calculated from the 2D- C probe were compared with those from the King Air's 2D-P probe on several passes through seeded cloud. Good agreement was found in the overlapping 0.6-1.0 mm range, but the 2D-C concentrations were only 25% to 75% of the 2D-P values for the 1.0-1.6 mm range. The 2D-C concentrations were even smaller fractions of the 2D-P values at larger particle sizes, yet a sub- stantial portion of the calculated precipitation rates from the 2D-C data will be shown to have been con- tributed by particles greater than 1.0 mm diameter. Because of these and other uncertainties, 2D-C precip- itation rates estimated according to Holroyd's scheme should be interpreted only in a relative sense, with the expectation that they are usually underestimates, es- pecially when particles larger than about 1.0 mm are involved. Quantitative comparisons between seeded and nonseeded snowfall estimated from the 2D-C probe should therefore be treated with caution. Unfortu- nately, surface precipitation observations were not practical in the Bridger Range, so 2D-C precipitation estimates at the lowest aircraft sampling level offered the only documentation of a key variable. Usually 10 to 30 valid hydrometeor images were stored in each of the two buffers of the 2D-C (or 2D- P) probe data acquisition system. A buffer would fill, and while it waited for the opportunity to write the information to magnetic tape, the second buffer would be filling. The rate of buffer filling depended on particle concentration and size, but the data systems were lim- ited to recording no more than one 2D-C buffer per second on the Turbo Commander and either a 2D-C or a 2D-P buffer per second on the King Air. When encountering high IPC, a buffer might fill in a fraction of a second, so that sampling might occur only during a portion of the period between recording times. A 62-channel, one-dimensional spectrometer probe (PMS 260X) covering the size range 10-620 Jlm was also carried by the King Air. Particles occulting the end elements of the 64-element optical array were ex- cluded. The first three channels are difficult to interpret because sample volume is a sensitive function of par- ticle size; the 260X concentrations referred to in Part II were for the size range 40-590 Jlm. Since the 260X particle: concentration was routinely calculated for each second of flight (unlike the 2D-C), it is referred to for general information on cloud conditions. c. Aircraft liquid water observations All airborne SL W measurements over the Bridger Range were derived from the PMS Forward Scattering Spectrometer Probe (FSSP). The size bins selected for the FSSP started at 1.0 Jlm. In all plots of SL W content, the value was set to zero unless the FSSP detected at least 1 () droplets cm - 3. This eliminated frequent cases in which one to a few relatively large particles cm-3 (presumably ice) would yield a calculated SL W content of 0.0 1-0.03 g cm -3 , while the other liquid water sen- sors detected no SLW. A Johnson-Williams (J-W) liquid water content meter, factory reconditioned prior to the field season, was used over the Grand Mesa. Once zeroed outside of cloud, it exhibited little drift and appeared to have a resolution near 0.02 g m -3, As shown by Strapp and Schemenauer (1982), a properly maintained J-W sys- tem can yield measurements of reasonable accuracy. The J-W device was compared with a Rosemount 871 FA icing rate meter (designed for aircraft use) using observations from 18 March 1986, one of three exper- imental days discussed in Part III. The Rosemount meter, which cycles by heating the probe to shed ice once a given mass adheres to it, was calibrated in a small, locally designed wind tunnel with 3 m S-I flow using droplets near 10 Jlm diameter. The mass of ice required to cycle the unit was carefully weighed after each of 11 tests and found to range from 0.032 to 0.045 g. The mean value of 0.039 g was used, along with the aircraft true air speed, probe cross-sectional area, and