Search Results

You are looking at 1 - 10 of 12 items for

  • Author or Editor: Edward Szoke x
  • Refine by Access: All Content x
Clear All Modify Search
Edward J. Szoke

Abstract

Observation of an overcast of low cloudiness on a Denver Cyclone day revealed a circular area of clearing or “eye” at the approximate center or the mesoscale circulation. The eye passed close to an automated observing station, providing a time series of data that shows the relationship of the circulation to the cloud-free area. Conditions present in the lower troposphere are examined and compared to conclusions from recent modeling results that suggest the causes of the Denver Cyclone.

Full access
Edward J. Szoke and Edward J. Zipser

Abstract

This is Part II of a two-part paper describing the vertical profile of radar reflectivity in GATE convective cells. Time-height radar life histories for 42 cells over three GATE days are examined, using data from the Quadra radar with 5-minute resolution. Mean profiles and plots of cell characteristics are generated, and confirm that the mean profiles in Part I are representative of the active portion of the cell lifetime. There are marked differences between the cell life histories of isolated cells and the longer-lived cells associated with mesoscale systems. In contrast to cells sampled in organized systems, the isolated cells are often of very limited vertical extent and must be dominated by the warm rain process. When forcing features exist such as gust fronts and intersecting lines of convection, they appear to dominate the generation of new convection, and isolated strong echoes are not observed.

Composite life histories for typical GATE cells are constructed. The typical radar echo forms first at an altitude of 2.5 km and reaches the surface about 5 minutes later, strongly suggesting early domination by the warm rain process. At the same time the echo top rises and the mid-to-late stages of cell lifetime involve both warm rain and ice processes. The reflectivity profiles of the longer-lived echoes change relatively little in the middle 50% of the life cycle.

Full access
Edward J. Szoke and Raymond H. Brady

Abstract

The case of a tornadic thunderstorm on 26 July 1985 in northeastern Colorado is described from the synoptic to the thunderstorm scale utilizing a number of datasets some of which will become operational in the 1990s. The available data included profilers, Doppler radar, surface mesonet, satellite, and special soundings. Although the synoptic environment did not favor tornadic thunderstorms, strong thunderstorms formed in localized area during a 2-h period in the afternoon and produced an 18-min tornado. A number of events took place to produce the stronger then anticipated development, including interaction among mesoscale outflow and stationary boundaries. Of particular importance was the change in the local environment along a stationary boundary known as the Denver Convergence-Vorticity Zone. Special soundings taken near the stationary boundary revealed a deepening moist layer over time in association with the convergent wind field. Additional forcing from the collision of this boundary with other outflow boundaries was required, to release the increasing convective potential. Of further importance was the coincidence of sunshine over a portion of the boundary where the eventual collision would occur. In association with experiments under way at the time, three groups having various access to the datasets issued probability forecasts in real time. The short-range forecasts and warnings of thunderstorm initiation and severe weather provide a useful evaluation of the problems of predicting the events of this day. Two major problems encountered on this day are commonly faced by forecasters of summertime convection: evaluating the importance and timing of an approaching weak upper-1evel feature, and monitoring low-level mesoscale boundaries. Use of the datasets in real time to diagnose these problems is emphasized. The forecasters using the new datasets demonstrated the ability to focus on the most likely area of thunderstorm initiation, although the timing and intensity of the development still presented forecast problems. Analyses of the special soundings revealed the above-surface temperature and moisture changes resulting from the two features and their importance for this case. Means of addressing such changes operationally with the new datasets are discussed.

Full access
Raymond H. Brady and Edward J. Szoke

Abstract

The evolution of the 26 July 1985 Erie, Colorado tornado is described using data from NCAR's CP-2 Doppler radar. This tornado develops within 20 km of the radar site under weakly forced synoptic conditions and weak tropospheric flow, and is not accompanied by a mesocyclone. The initial circulation forms near the surface at the intersection of two mesoscale boundaries and develops vertically, intensifying into an Fl tornado when it becomes collocated with the intense updrafts of a rapidly developing cumulonimbus.

This tornado appears to be the land equivalent of a waterspout, and comparisons between the two vortices are made. It is speculated that Florida and portions of the western High Plains may be prone to nonmesocyclone tornado development, and that vortex-intensification processes associated with nonmesoscyclone tornadoes may be important in mesocyclone tornadogenesis. Suggestions on how to better forecast these tornadoes are also presented.

Full access
Edward J. Szoke, Edward J. Zipser, and David P. Jorgensen

Abstract

This is Part I of a two-paper describing the structure of the vertical profile of radar reflectivity in convective cells which are of part of mesoscale convective systems in GATE. Earlier work has established that characteristic mean vertical velocities in such convective clouds in GATE were rather weak, <3–5 m s−1. The microphysical implications of the weak updrafts have been proposed to include the rarity of large particles above the freezing level. As a working hypothesis for these papers it is proposed that cells with weak updrafts have characteristic vertical reflectivity profiles exhibiting modest reflectivities at low levels, and decreasing rapidly with height above the freezing level.

Vertical radar reflectivity profiles are compiled from 296 convective cells having at least 40 dBZ surface reflectivity, and it is found that the profiles are consistent with the above reasoning. The, mean profile is of modest strength—45 dBZ echo at the surface and a 20 dBZ echo top of 8.2 km, and the reflectivity decreases rapidly with height above the freezing level. Statistics of surface reflectivity, echo top, and height of the maximum echo aloft all lend support to the above picture of the GATE cell. The GATE radar profiles are similar to mature hurricane radar profiles, consistent with their updraft velocities being weak as well. In contrast the GATE profiles are markedly weaker than profiles from continental thunderstorm cells, consistent with the much higher vertical velocities in the continental cells.

Full access
John F. Weaver, James F. W. Purdom, and Edward J. Szoke

Abstract

During the late afternoon and early evening of 6 June 1990, a series of severe thunderstorms produced nine tornadoes and numerous incidents of large hail on the High Plains of eastern Colorado. While the morning synoptic data clearly indicated a severe threat over the entire eastern half of the state, the severe activity that did occur was much more localized. Significant events were confined to a relatively small geographical region east and southeast of Denver, Colorado, including the small town of Limon some 70 miles to its southeast.

Satellite, radar, surface, and upper-air data are combined in this paper to study some of the mesoscale aspects of the severe storm environment. Results show that thunderstorm outflow from a large mesoscale convective system in Kansas and Nebraska played a crucial role in focusing the severe activity in eastern Colorado. Also, the evolution of convective development during the early part of the day suggested the presence of a sharp moisture gradient along the Front Range of the Rocky Mountains, which further helped to localize the outbreak. Finally, interactions between individual storms appear to have been critical to severe storm evolution.

Full access
Edward J. Szoke, John M. Brown, John A. McGinley, and Dennis Rodgers

Abstract

Stormscale Operational and Research Meteorology-Fronts Experimental Systems Test (STORM-FEST) was held from 1 February to 15 March 1992 in the central United States as a preliminary field systems test for an eventual larger-scale program. One of the systems tested was a remote operations center, located in Boulder, Colorado, which was significantly displaced from the main field concentration of scientists and research aircraft. In concert with the remote operations center test was a test of remote forecasting support, also centered in Boulder. The remote forecasting for STORM-FEST was the first major cooperative effort for the Boulder-Denver Experimental Forecast Facility (EFF), a cooperative effort between operations and research aimed at finding more effective ways of addressing applied meteorological problems. Two other newly formed EFF's, at Norman, Oklahoma, and Kansas City, Missouri, also played key roles in the forecasting/nowcasting support. A description of the design and function of this remote forecasting and nowcasting support is given, followed by an assessment of its utility during STORM-FEST. Although remote forecasting support was deemed plausible based on the STORM-FEST experience, a number of suggestions are given for a more effective way to conduct forecasting experiments and provide forecasting support during a field program.

Full access
Stanley G. Benjamin, Barry E. Schwartz, Steven E. Koch, and Edward J. Szoke
Full access
Stanley G. Benjamin, Barry E. Schwartz, Edward J. Szoke, and Steven E. Koch

An assessment of the value of data from the NOAA Profiler Network (NPN) on weather forecasting is presented. A series of experiments was conducted using the Rapid Update Cycle (RUC) model/assimilation system in which various data sources were denied in order to assess the relative importance of the profiler data for short-range wind forecasts. Average verification statistics from a 13-day cold-season test period indicate that the profiler data have a positive impact on short-range (3–12 h) forecasts over the RUC domain containing the lower 48 United States, which are strongest at the 3-h projection over a central U.S. subdomain that includes most of the profiler sites, as well as downwind of the profiler observations over the eastern United States. Overall, profiler data reduce wind forecast errors at all levels from 850 to 150 hPa, especially below 300 hPa where there are relatively few automated aircraft observations. At night when fewer commercial aircraft are flying, profiler data also contribute strongly to more accurate 3-h forecasts, including near-tropopause maximum wind levels. For the test period, the profiler data contributed up to 20%–30% (at 700 hPa) of the overall reduction of 3-h wind forecast error by all data sources combined. Inclusion of wind profiler data also reduced 3-h errors for height, relative humidity, and temperature by 5%-15%, averaged over different vertical levels. Time series and statistics from large-error events demonstrate that the impact of profiler data may be much larger in peak error situations.

Three data assimilation case studies from cold and warm seasons are presented that illustrate the value of the profiler observations for improving weather forecasts. The first case study indicates that inclusion of profiler data in the RUC model runs for the 3 May 1999 Oklahoma tornado outbreak improved model guidance of convective available potential energy (CAPE), 300-hPa wind, and precipitation in southwestern Oklahoma at the onset of the event. In the second case study, inclusion of profiler data led to better RUC precipitation forecasts associated with a severe snow and ice storm that occurred over the central plains of the United States in February 2001. A third case study describes the effect of profiler data for a tornado event in Oklahoma on 8 May 2003. Summaries of National Weather Service (NWS) forecaster use of profiler data in daily operations, although subjective, support the results from these case studies and the statistical forecast model impact study in the broad sense that profiler data contribute significantly to improved short-range forecasts over the central United States where these observations currently exist.

Full access
William R. Moninger, Stanley G. Benjamin, Brian D. Jamison, Thomas W. Schlatter, Tracy Lorraine Smith, and Edward J. Szoke

Abstract

A multiyear evaluation of a regional aircraft observation system [Tropospheric Aircraft Meteorological Data Reports (TAMDAR)] is presented. TAMDAR observation errors are compared with errors in traditional reports from commercial aircraft [aircraft meteorological data reports (AMDAR)], and the impacts of TAMDAR observations on forecasts from the Rapid Update Cycle (RUC) over a 3-yr period are evaluated. Because of the high vertical resolution of TAMDAR observations near the surface, a novel verification system has been developed and employed that compares RUC forecasts against raobs every 10 hPa; this revealed TAMDAR-related positive impacts on RUC forecasts—particularly for relative humidity forecasts—that were not evident when only raob mandatory levels were considered. In addition, multiple retrospective experiments were performed over two 10-day periods, one in winter and one in summer; these allowed for the assessment of the impacts of various data assimilation strategies and varying data resolutions. TAMDAR’s impacts on 3-h RUC forecasts of temperature, relative humidity, and wind are found to be positive and, for temperature and relative humidity, substantial in the region, altitude, and time range over which TAMDAR-equipped aircraft operated during the studied period of analysis.

Full access