Atmospheric Sciences. (Senior Division 2002).

Ebert, R.L., and P.S. Market. Department of Atmospheric Science, University of Missouri-Columbia. ASPECTS OF THUNDERSNOW EVENTS IN MISSOURI. Thirteen (13) individual events of thundersnow were identified over Missouri for the period 1961-1990. Surface observations from St. Louis, Kansas City, Springfield, and Columbia, Missouri, were used to further identify events that featured thunder with snow only (11). The sparse spatial data coverage and limitations with the dataset preclude the compilation of a regional climatology. However, of the ten (10) events with sufficient upper air data identified over Missouri, more than half (7) featured moderate or heavy snowfall, and four (4) of those featured heavy snowfall at some point during the event. The synoptic setting of these events falls into two categories: 1) the presence of a transient cyclone (8 of 10), and 2) isentropic upglide over an arctic frontal boundary (usually as return flow west of an anticyclone; 2 of 10). In addition, 7 of the 10 cases occurred in southerly or southwesterly flow aloft (500 mb), 2 featured west or northwest flow, and one was indeterminate.

Lam, J.E., P.S. Market, and A.R. Lupo. Department of Atmospheric Science, University of Missouri-Columbia. THE AFFECT OF AN URBAN HEAT ISLAND ON DOWNWIND LOCATIONS. This work constitutes a further investigation of the Columbia Heat Island Experiment (COHIX) data first presented last year. With this study, we will show that the affects of an urban heat island can be transported downwind. Data were examined from October 2000 to May 2001. Of the 243 days in the period, 95 did not involve a wind shift or precipitation at some point during the day. On these 95 days, a dominant wind direction was determined (northeast, southeast, southwest, and northwest) as was the quadrant of the city that experienced the highest temperature that day. The quadrant from which the wind was blowing and the quadrant of the highest maximum temperature were correlated (-0.267) at a 99% confidence interval. This indicates that the heat from the urban area was being transported downwind by the prevailing flow. The data were stratified by clear/cloudy sky conditions as well as by wind/not-windy days. Of greatest interest is the complete lack of evidence of thermal advection on days with winds in excess of 10 knots, indicating the affects of mixing. Days with winds less than 10 knots have a correlation between the quadrant from which the wind was blowing and the quadrant of the highest maximum temperature of -0.360 valid on the 99% confidence interval.

Lupo, A.R., E. McCoy, S. Allen, A. Akyuz, E.P. Kelsey, C. Halcomb, E. Aldrich, D. Beiger, E. Wise, M. Edwards, and D. Schmidt. Department of Atmospheric Sciences, University of Missouri-Columbia. THE PRESENTATION OF TEMPERATURE INFORMATION IN TELEVISION BROADCASTS: WHAT IS NORMAL? In a typical weather broadcast, observed daily temperature information such as maximum and minimum temperatures are shown and compared to normal. Such information, however, does not accurately describe whether or not that particular day is fairly typical for that time of year or truly an unusual occurrence. Thus it is suggested that the presentation of temperature information can be augmented with elementary statistical information in order to give a more meaningful presentation of weather information without the need to explain the basis of such statistical information. A study of the climatological maximum and minimum temperatures over a 30-year period for Columbia, Missouri is performed in order to provide the rationale for displ aying a 'typical" temperature range along with daily observations. This information was incorporated into television weather broadcasts at KOMU TV-8, the campus television station and local NBC affiliate.

Lupo, A.R., D. Albert, R. Hearst, C. L. Allmeyer, and P.S. Market. Department of Atmospheric Sciences, University of Missouri-Columbia. INTERANNUAL VARIABILITY OF SNOWFALL-TO-LIQUID WATER AMOUNTS IN SOUTHWEST MISSOURI. Forecasting the snowfall amounts for an approaching synoptic system is one of the most formidable challenges for weather forecasters. By using data acquired from the Missouri Climate Center and the National Weather Service Forecast Office in Springfield, Missouri, a climatology of snowfall-to-liquid (LS) ratios was derived for Southwest Missouri. A climatology of Southwest Missouri snowfalls was derived and then compared to a climatology using observations from the Springfield Weather Service Office only. These climatologies were very similar overall. Thus, LS ratio data, which is only available to Springfield, Missouri observations alone, should adequately represent LS ratios for Southwest Missouri. It was found that winter snowfalls had a higher LS ratio that those in the fall or spring seaso n. Overall, most snowfall events had an LS ratio of less than 12 inches of snow to 1 inch of liquid (12:1). Higher LS ratios were associated with northwest flow regime snow events, while lower ratios were found with southwest flow regime events. When examining the interannual variability, the percentage of low LS ratio snowfall events in El Nino years was significantly higher. This mirrored the greater frequency of southwest flow regime snowfall events in these years.

Market, P.S., and M. Bodner. Department of Atmospheric Science, University of Missouri-Columbia. NOAA/Hydrometeorological Prediction Center, Camp Springs, Maryland. THE CHICAGO-DUPAGE, ILLINOIS, THUNDERSNOW EVENT OF 31 JANUARY 2002. A case study is presented of a thundersnow event not associated with a strong cyclone, and well north of a surface quasi-stationary frontal boundary. Standard analyses reveal the presence of an 850 mb frontal zone, broad southwesterly flow at 500 mb, and the right entrance region of an anticyclonically curved jet streak at 250 mb all over the region at the time of the event. Initial fields from the Rapid Update Cycle (RUC) model at the time indicate the presence of ample moisture and strong frontogenesis in the presence of weak conditional symmetric instability over the region. A small area of conditional symmetric instability was diagnosed at -850 mb just south of the thundersnow event location. Aircraft soundings from the time of the event suggest little in the way of instabilit y over northern Illinois, and the lightning detection network failed to detect cloud-to-ground lightning strokes there. Still, lightning and thunder were observed with the event at DuPage, Illinois, suggesting the presence of vertical motions strong enough to generate charge separation.

Podrazik, K., and P.S. Market. Department of Atmospheric Science, University of Missouri-Columbia. STATISTICAL MODEL PERFORMANCE IN FORECASTING AVIATION-SENSITIVE WEATHER VARIABLES. The Aviation Model's model output statistics (AVN MOS) guidance were collected for the period of September 2000-August 2001. These output were compared to actual surface observations from the Columbia Regional Airport (COU). In particular, variables germane to aviation operations (cloud cover, ceiling, and visibility) were examined. Days without cloud cover or low ceilings, and those with high visibilities were excluded from the study. Skill was assessed for multiple model runs valid at the same time. Not surprisingly, the forecast skill of AVN MOS for these variables improves as the lead-time before a forecast hour dwindles.

Podzimek, J. and M. Podzimek. Cloud and Aerosol Sciences Laboratory, University of Missouri-Rolla. NEW DIVISION AND PARAMETERIZATION OF SNOWFLAKES. A new division of snowflakes is suggested based on the analysis of snow crystal and snowflake samples collected over eight years in the Midwestern United States. The division is rooted in the prevailing presence of specific snow crystals in aggregates and contains the following six types of aggregated snowfall elements: Needles and sheaths; small columnar and plate crystals; medium size plate and stellar crystals; large dendritic and stellar crystals; aggregates of minigraupels and frozen drops; mixtures of large drops (often from melted snowflakes) and snow crystals. Calculations performed for each group of snowflakes include the mean occurrence during long-term sampling and the mean morphological and dimensional parameters. The morphological parameters consist of the mean crystal number in a snowflake, the types of aggregated crystals, and the potential interact ion of rimed crystals. The mean size and maximum size of the specific type of snowflake are among the dimensional parameters mentioned. The size distribution curves of the five first groups of snowflakes are also carefully considered. A simple log-normal distribution is suitable for only one or two groups of snowflakes (e.g. for large dendritic aggregates). Attention was also paid to the simultaneous occurrence and interaction of different types of snowflakes at specific meteorological situations.

Podzimek, J. and E. Robb. Cloud and Aerosol Sciences Laboratory, University of Missouri-Rolla. SNOW CRYSTAL FRAGMENTS COLLECTED AT THE GROUND. The evaluation of 885 snow crystal replica samples collected at Rolla, MO, and at Groveland, IL, during the years 1993-2001 document the importance of the snow elements for the modeling of a self-cleaning atmosphere, the study of visibility, and other phenomena in the atmospheric boundary layer. More than 63% of snow crystal samples--each usually comprised of more than 100 crystals-- contained broken arms of dendritic and stellar crystals. These fragments were divided into one arm, two arm, and three arm elements classified as either dendritic or stellar. For each element, the main dimensional and morphological parameters affecting the motion and aggregation of falling crystals were established. The mean sizes of dendritic one arm fragments were 1.16 mm, of two arm fragments 1.21 mm, and of three arm fragments 1.49 mm. The corresponding fragments of stellar crystals we re 0.80mm, 0.89mm, and 1.02mm. Possible effects of the air temperature, wind speed, riming, and presence of frozen droplets or minigraupels on the crystal fragmentation were investigated. It appears that the temperature and wind speed at the ground does not substantially affect the occurrence of fragments, which is usually related to the formation of very large dendritic and stellar crystals at temperatures around -15[degrees]C in the atmosphere.

Rutter, A.P.,(1) P.D. Whitefield,(1) D.E. Hagen,(1) A.R. Hopkins,(1) and M. Ross.(2) (1) Cloud and Aerosol Sciences Laboratory, University of Missouri-Rolla. (2) Air Force Office of Scientific Research. OBSERVATIONS OF VOLATILE AND NONVOLATILE PARTICULATES IN A HIGH ALTITUDE ROCKET PLUME. Measurements of Non-volatile and volatile (-198[degrees]C) particulates were made at high altitude (16-19km) in the exhaust plume of the Athena II (IKONOS, 9/24/99, Vandenberg Air Force Base) rocket. Size distributions and particulate concentration profiles of total and non-volatile particulates extending over the size range of 8-4000nm were measured. Data in the size range 340-4000nm was collected throughout the flight using laser particle counting techniques. Data in the size range 8-250nm were acquired only during plume incursions using a Grab Tank Sampling system. The NASA WB-57F high altitude research aircraft was used as the experimental platform. Three modes were observed in the particulate size distributions centred at <8nm, 50-60nm and 900-l000nm. The relative mass fractions present in each mode wer e 0.04%[+ or -]0.03, 7.39%+/-3.96, 92.57%[+ or -]4.00. A volatile component was observed for the first time in the Athena II measurement flight. References: [1] Ross, M.N., P.D. Whitefield, D.E. Hagen and R. Hopkins, "In- Situ Measurement of the Aerosol Size Distribution in Stratospheric Solid Rocket Motor Exhaust Plumes", Geophys. Res. Let. 26, 819-822. (1999).

Zacher, C.A. and J.T. Moore. AERO Research and Department of Earth and Atmospheric Sciences, St. Louis University. REVISITING BIPOLAR PATTERNS IDENTIFIED IN TRACKING ELECTRIFIED MESOSCALE CONVECTIVE SYSTEMS. We have called attention in an earlier paper on the synoptic development of cloud-to-ground (CG) lightning discharges tracked across the country, to the relative positional placements of positive (+) and negative (-) flashes within storms (MAS Transaction, 1991). Mapping shown onscreen is produced in near real time by an updated National Lightning Detection Network (NLDN) array of sensors and ultimately communicated via satellite link through the Lightning Data and Information Systems (LDIS). This information transmitted, especially in developing mesoscale convective systems (MCS), a predominance of positive CG flashes appearing in advance of negative CG flashes, contrary to expectations. This would reverse (seemingly) the order of precedence - convective elements with negative discharges initially, leadi ng to stratiform, anvil-like portions with positive discharges subsequently, We will hereby reexamine some factors that might contaminate these data, falling broadly in two areas: 1) Confusion with, or synchronous intracloud flashes, and 2) Instrument detection limitations. The conclusion is that we strongly suspect masking or data contamination by such factors.