Close Window

Pennsylvania State Univ.: "Studies to improve the forecasting and warning of weakly forced and non-classic severe convective events"

Final Report


1.1 Introduction: Summary of Overall Project Objectives

The combination of reflectivity, velocity, storm relative velocity, and derived product data made available by the WSR-88D radar now allows forecasters to quickly identify classic characteristics of severe and possibly tornadic thunderstorms. Forecasters are even computer-alerted when certain classic severe storm signatures are present (e.g., a mesocyclone signature).

However, the new technology also brings to the attention of the forecaster a variety of other storms not of classic form but containing noteworthy features which bear watching as possible clues to storm severity. Many storms can be seen to contain modest rotation, for example, that does not meet the criteria for a mesocyclone. Much less is known about the circumstances under which--or how frequently--these "borderline" severe thunderstorms produce severe weather. This may be particularly important in regions like the Mid-Atlantic States where classic storm signatures are not as frequent as in the Midwest, and where terrain may influence the frequency and location of occurrence of severe weather. Thus, research is being performed to determine additional factors that can be used to identify which of the borderline storms are sufficiently likely to produce severe weather that warnings are justified, and which should be monitored further prior to issuance of a warning.

Unfortunately, many of the non-classic storms in the Mid-Atlantic States occur over data- and spotter-sparse regions where verification of severe weather is difficult to obtain in real-time, or even after the fact, unless a man-made structure is hit or a road is blocked. Thus, the needed research cannot be done unless it can be ascertained with a high degree of confidence, which of the borderline storms actually did and did not produce severe weather. To attain this degree of confidence, special verification efforts are needed.

1.1.1 Project Objectives.

The project had five main objectives:

a. To make special efforts to obtain positive or negative confirmations of the occurrence of severe weather whenever WSR-88D signatures suggested the possibility of severe weather to the duty forecaster. These efforts included special efforts to solicit public and volunteer observer input as well as damage surveys by NWS and University project participants.

b. Using the set of cases for which the presence or absence of severe weather has been reliably determined, develop statistics of the reliability of the various WSR-88D signatures as severe weather indicators.

c. Conduct case studies to gain further knowledge of the special circumstances under which the WSR-88D signatures are and are not reliable indicators of storm severity. Both radar-detected factors (such as storm structure and evolution history) and environmental factors (low-level winds and moisture thresholds, static stability, vertical wind shear, and helicity) were considered.

d. Conduct case studies to focus on particular convective tendencies of the Appalachian region.

e. Initiate communication with other NWS offices in the region, ultimately leading to workshops on the use of WSR-88D for severe weather detection and warning over the Mid-Atlantic and Northeast region of the United States.

Progress has been made on each of these five objectives, as described in subsequent sections.

1.1.2 Project Participants.

Research assistantships for Matt Pearce and Bob Much (graduate students) were partially supported by Project funds. Greg Forbes (Associate Professor) served as the principal investigator. Forbes participated in all aspects of the project, including damage surveys and detailed examination of radar and companion data sets. Matt Pearce participated through the end of the summer, 1997. Bob Much (graduate student) was supported during spring and summer, 1998. Paul Knight, of the Penn State Department of Meteorology Weather Communications Group, participated by airing requests for storm damage information on the WPSX-TV program "Weather/World" on several occasions. Penn State undergraduate honors student Joel Widenor participated in verification data collection and damage surveys during 1996 as a NWS student trainee and conducted additional studies in conjunction with his Honors Thesis. Nearly every meteorologist in the NWS Office in State College participated in the project in one or more ways. Each forecaster kept logs of severe storm occurrence and noteworthy radar signatures and, in most cases, reviewed the radar for cases they worked once damage information had been obtained. Tom Dunham (WCM), Ernie Ostuno, Rich Grumm (SOO), Bruce Budd (MIC), and Paul Head were the most frequent participants in intensive damage survey efforts, though nearly all meteorologists participated in one or more damage surveys.

Responsibilities for conducting damage surveys and post-storm case studies were typically agreed upon on a case basis. Following a NWS radar replay, more detailed studies were most frequently conducted by Penn State personnel, sometimes in conjunction with NWS meteorologists. The NWS, of course, took full responsibility for issuing real-time severe weather and tornado warnings. The NWS (and student hire Joel Widenor) took responsibility for calls to Emergency Managers, utility companies, and other public officials in soliciting initial information regarding storm damage. The NWS took final responsibility for assimilating storm damage survey information into an account for Storm Data.

1.2 Description of Research Accomplishments

Research was performed on several topics:

  1. Structure and evolution of shallow supercell thunderstorms;

  2. Interaction of supercell thunderstorms with mesoscale boundaries and tornado development;

  3. Development of tornadoes from non-supercell thunderstorms;

  4. Thunderstorms producing hail and wind damage;

  5. Impact of terrain on severe thunderstorms in Central Pennsylvania;

  6. Pennsylvania tornado climatology.

Results of many of these studies have been reported in publications (cited in sections 3.2 and 3.3), and are summarized briefly in the following paragraphs.

Structure and evolution of shallow supercell thunderstorms. Most of the thunderstorms containing mesocyclone signatures over Central Pennsylvania exhibit maximum rotation at the lowest elevation scans, rather than forming aloft and extending downward with time. In some cases mesocyclones were weak prior to development of the first tornado of the family. Pearce (1997) and Pearce, Forbes, and Ostuno (1998), and Grumm (1998) summarized this research. This research is being continued by Ernie Ostuno and Rich Grumm of the State College office of the NWS.

Interaction of supercell thunderstorms with mesoscale boundaries
. Tornado development. During several notable tornado outbreaks, there appeared to be a remarkable tradeoff between (a) the static stability of antecedent cool air masses and (b) the backed low-level winds in these air masses with regard to the potential for tornadic thunderstorms. The presence of the stable air sometimes appeared to be sufficient to inhibit tornado development, despite the presence of strong mesocyclones (e.g., 30 June 1996). The interaction of developing thunderstorms with a mesoscale boundary (and increased helicity of air on the cool side of the boundary) on several occasions seemed to enhance mesocyclone intensity and tornado development (e.g., 19 July 1996). Studies of these phenomena have been reported by Pearce (1997) and Pearce, Forbes, and Nicosia (1998).

Development of tornadoes from non-supercell thunderstorms. Many of the tornadoes affecting Central Pennsylvania during 1996, 1997, and 1998 occurred from shallow supercell thunderstorms (e.g., Widenor, 1997). However, a large number of tornadoes also occurred from non-supercell thunderstorms, such as from bow echoes and echo mergers. Several cases of tornado-producing mini-bow echoes were well documented in 1996 and 1997. These have been reported in Pearce (1997) and Forbes et al (1998). Bob Much (Penn State), Ernie Ostuno, and Rich Grumm continue research on this topic.

Thunderstorms producing hail and wind damage
. During several damage surveys it became obvious from the damage to crops that large hail had impacted the region. These events have been noted in Pearce (1997), Pearce, Forbes and Ostuno (1998) and Pearce, Forbes and Nicosia (1998). Microbursts have been reported by Pearce (1997) and continue to be studied by Ernie Ostuno, Bob Much, and Evan Bookbinder (Penn State). Wind damage from pulse storms does not appear to always be systematically preceded by collapse of reflectivity cores in Central Pennsylvania storms.

Impact of terrain on severe thunderstorms in Central Pennsylvania
. Forbes (1998) has summarized the occurrences of tornadoes in Central Pennsylvania during 1996 and 1997 in relation to terrain. Supercell tornadoes exhibit variations in substructure related to traversal of significant terrain slopes. Most non-supercell tornadoes occurred on the lee or cyclonic-vorticity slopes of terrain features. Bob Much and other Penn State students continue to examine the impact of terrain features on thunderstorm development and intensity variations.

Pennsylvania tornado climatology. A statistical analysis of tornado occurrences was conducted in order to quantify the variations in tornado threat by county in Pennsylvania. Results have been published by Nese and Forbes (1998). It was suggested in this study that there are biases that impact the frequency of reported tornadoes in many counties within the Commonwealth. As also hypothesized by others in other regions, it would appear that distance from a National Weather Service office and population density of a county affect the frequency of reported tornadoes, particularly for weak tornadoes. For strong tornadoes, there is
probably less of a reporting bias, but the distribution is heavily weighted by a few outbreaks in a climatological sampling period that is too short to be statistically valid. An undergraduate student then began to quantitatively assess the impact of population density on reported Pennsylvania tornado density, but that research has not been completed.


a. Real-time Interaction during Severe Weather Events

Greg Forbes and Matt Pearce collaborated with the State College office of the National Weather Service in severe weather situations, offering independent opinions regarding the likelihood of severe weather on various occasions and regarding the interpretation of radar signatures.

Forbes, Pearce and Bob Hart made various observed and model diagnostic products made available to NWS forecasters in real time that would not have otherwise been available to them:

b. Damage Surveys

Penn State and NWS meteorologists conducted dozens of storm surveys of severe weather across central Pennsylvania, jointly or in separate teams. These surveys have improved the ability of the NWS office to categorize damage, provide detailed public information statements about the severe weather across the region, and identify in post-storm mode the radar signatures of severe storms.

c. Workshops

NWS Mid-Atlantic Severe Weather Workshop, State College, PA, May 1, 1996 (PSU/NWS Mid-Atlantic offices).

Session on Mid-Atlantic Severe Storms, 73rd Annual Meeting, Pennsylvania Academy of Science (PAS), Malvern,
PA, 12 April 1997 (PSU/NWS Mid-Atlantic offices).

NWS Severe Weather Workshop, State College, PA, April 28, 1997 (PSU/NWS State College).

Session on PA Severe Weather, 74th Annual Meeting PAS, Monroeville, PA, April 1998

NWS Severe Weather Workshop, State College, PA, May 18, 1998 (PSU/NWS State College).

Session on Meteorology, 75th Annual Meeting of the PAS, April, 1999

d. Joint Penn State-NWS Map Discussion Class

e. Shared Facilities; Exchange of Data and Expertise

f. Educational Exchanges

Several NWS meteorologists are taking courses in the Penn State Department of Meteorology. Some of these courses involve
student presentations, such that the other Penn State students gain insights from the NWS operationally oriented presentations.
NWS meteorologists have given guest lectures in Penn State courses.

g. Award Demonstrating Effectiveness of Exchange

Greg Forbes and Matt Pearce received special service award from the National Weather Service Eastern Region for their
cooperative efforts to improve severe weather forecasting and warnings in Central Pennsylvania (November 1997).


3.1 Presentations on Project Research

a. 73rd Annual Meeting, Pennsylvania Academy of Science, April, 1997

Forbes, G.S., 1997: Challenges in forecasting and warning for weakly forced, non-classic severe weather in PA.

Forbes, G., M. Pearce and Staff, NWS Office, State College, PA, 1997: Radar signatures of the severe storms in
Central Pennsylvania, 1996.

Forbes, G., M. Pearce, T. Dunham, R. Grumm, 1997: Challenges in forecasting and warning for weakly forced,
non-classic severe weather in PA.

Grumm, R.H. and P. Jung, 1997: Memorial Day severe weather outbreak: A Pennsylvania perspective.

Nese, J.M. and G.S. Forbes, 1997: An updated tornado climatology for Pennsylvania.

Pearce, M., G. Forbes, J. Widenor, B. Budd, E. Ostuno, T. Dunham and P. Head, 1997: Pennsylvania tornado
outbreak of 19 July 1996.

Pearce, M. and J. Widenor, 1997: The Shippensburg tornado and other severe weather on 30 July 1996.

Widenor, J., T. Dunham and B. Budd, 1997: Factors that affect the skill of severe weather forecasts and warnings in

b. 74th Annual Meeting, Pennsylvania Academy of Science, April, 1998

Forbes, G.S., E. Ostuno, T. Dunham, 1998: Severe weather from mini-bow echoes over Central PA.

Grumm, R.H and J. LaCorte, 1998: Serial derechos: Cold season severe weather type.

Ostuno, E.J., 1998: Severe weather from non-classic supercell thunderstorms over Central Pennsylvania

c. 16th Conf. Weather Analysis and Forecasting, Phoenix, AZ, January 1998

Presentations of papers listed in section 3.2: Publications

d. 19th Conf. Severe Local Storms, Minneapolis, MN, September 1998

Ostuno, E., R.H. Grumm, G. Forbes, 1998: Doppler radar observations of supercell thunderstorms over central
Pennsylvania. Poster.

Other presentations listed under section 3.2: Publications

e. National Weather Service Severe Weather Workshops, 1997

Greg Forbes: Non-classic severe weather radar signatures across Pennsylvania during the 1996 severe weather
Matt Pearce: A potpourri of severe weather
Matt Pearce, Rich Grumm: Poster session on Pennsylvania severe weather
Ernie Ostuno: Effective warnings and statements
Paul Head: Mesocyclone characteristics and tornadoes

f. National Weather Service Severe Weather Workshops, 1998

Forbes, G.S., 1998: Weakly forced and non-classic severe weather in Central Pennsylvania. Presentations at
National Weather Service Severe Weather Workshops in State College, PA and Mt. Holly, NJ, May 18 and 19, 1998,

g. Miscellaneous

Forbes, G.S., 1996: Severe weather studies in PA. Appearance on WPSX-TV Weather/World, 5 June 1996

Forbes, G.S., 1997: Severe weather studies in PA. Appearance on WPSX-TV Weather/World

Forbes, G.S., 1997: Studies of Pennsylvania Severe Weather. Invited presentation at PSUBAMS (Penn State Student
Chapter of the AMS) meeting, 3 September 1997.

3.2 Publications Supported Directly by Project Funds

Forbes, G.S., M.L. Pearce, T.E. Dunham, R.H. Grumm, 1998: Downbursts and gustnadoes from mini-bow echoes and affiliated mesoscale cyclones over Central Pennsylvania. Preprints, 16th Conf. Wea. Analysis and Forecasting, Phoenix, AZ, Amer. Meteor. Soc., 295-297.

Forbes, G.S., 1998: Topographic influences on tornadoes in Pennsylvania. Preprints, 19th Conf. Severe Local Storms, Minneapolis, MN, Amer. Meteor. Soc., 269-272.

Nese, J.M. and G.S. Forbes, 1998: An updated tornado climatology of Pennsylvania: Methodology and uncertainties. J.
Pennsylvania Academy of Science, 71(3), 113-124.

Pearce, M.L., 1997: Non-classic and weakly forced convective events: A forecasting challenge for the dominant form of severe
weather in the Mid-Atlantic region of the United States. M.S. Thesis, The Pennsylvania State University, 106pp.

Pearce, M., G. Forbes, D. Nicosia, 1998: Mesoscale boundaries, intersections, and storm-relative environmental helicity as factors in weak supercell and tornado formation over Pennsylvania. Preprints, 16th Conf. Wea. Analysis and Forecasting, Phoenix, AZ, Amer. Meteor. Soc., 332-334.

Pearce, M., G. Forbes and E. Ostuno, 1998: Storm-scale aspects of non-classic borderline supercell thunderstorms over
Pennsylvania. Preprints, 16th Conf. Wea. Analysis and Forecasting, Phoenix, AZ, Amer. Meteor. Soc., 109-111.

Widenor, J.: 1997: Evaluating the effect of an intensive severe weather verification program in Central Pennsylvania. B.S. Honors Thesis, The Pennsylvania State University, 112 pp.

3.3 Related Publications (including publications in progress)

Bookbinder, E, 1999: Radar signatures of pulse thunderstorms. B.S. Honors Thesis in progress

Forbes, G. and T. Miner, 1997: Record annual precipitation in State College, 1996. Pennsylvania Academy of Science Newsletter 55(2), 20.

Hart, R.E., G.S. Forbes, R. Grumm, 1998: Using hourly model-generated soundings to forecast convection in the Mid-Atlantic
region. Preprints, 16th Conf. Wea. Analysis and Forecasting, Phoenix, AZ, Amer. Meteor. Soc., 164-166.

Grumm, R.H., 1998: Memorial Day severe weather outbreak. Preprints, 16th Conf. Severe Local Storms, Minneapolis, MN, Amer. Meteor. Soc., 132-135.

Grumm, R.H., D.L. Michaud, K.P. Hlywiak, 1998: Precipitation forecasts along a narrow pre-cold frontal rainband. Preprints, 16th Conf. Wea. Analysis and Forecasting, Phoenix, AZ, Amer. Meteor. Soc., 233-235.

Knight, P., 1996: 1996 - A record weather year. Pennsylvania Academy of Science Newsletter 55(2), 22.

Much, R., 1999: Trends in radar parameters of severe thunderstorms over Central Pennsylvania. M.S. Thesis, The Pennsylvania
State University, in progress.

Pearce, M.L. and G.S. Forbes, 1997: A preliminary investigation of the 10-km Meso-Eta model in the Mid-Atlantic region.
Attachment 1, Semi-Annual Report 2, NOAA Award NA67WA0443, distributed by NWS Eastern Region, 49 pp.

Pearce, M.L. and S.G. Hoffert, 1997: The SNAP Weather Information Integrated Forecast Tool (SWIIFT): Next generation
software for research and operational meteorology. Preprints, 13th Int'l. Conf. Interactive Information and Processing Systems for Meteorology, Oceanography, and Hydrology, Long Beach, CA, Amer. Meteor. Soc., 269-272.


4.1 University's Perspective

Research efforts by University personnel have benefited from the COMET-sponsored NWS collaborative project in many ways. The National Weather Service has provided Penn State faculty, staff, and students with access to radar and numerical model data and to computer display systems for accessing, displaying, and manipulating the data. The NWS made several of its vehicles available for use in damage surveys. NWS meteorologists performed damage surveys that provided "ground-truth" data and allowed more storms to be studied than if Penn State personnel had acted alone. NWS meteorologists offered valuable insights into interpretation of radar and numerical model data and with regard to analysis of severe storm events. The collaboration went very smoothly with no identified problems.

4.2 NWS Perspective

The research efforts involved with this COMET project benefited the NWS Forecast office (WFO) in many ways. Operationally, the office gained expertise and training on how to survey storm damage and how to relate damage to radar signatures. New information gained during the project was rapidly integrated into seasonal workshops. This information proved useful and is probably reflected in the WFO's high POD for all types of severe convective weather. Joint research projects involved many members of the WFO staff. These efforts resulted in web pages showing storm surveys, radar imagery, and supporting meteorological information. This information benefited both the WFO and the emergency management community across the Commonwealth. The research efforts helped reinforce the information, further improving severe weather operations and facilitating rapid integration of new information into operations.

The nearly three-year interaction between the WFO and the Pennsylvania State University (PSU) during this project had many significant and unexpected benefits. Our office's knowledge of computers and computer software to display radar data improved. Jointly with Dr. Forbes and his students, we maintain a robust data archive, which includes archive II data from the WSR-88D and other meteorological data. Using the computer software, including WATADS and SWIIFT, we improved our understanding of the WSR-88D algorithms and their limitations. We learned to maintain two web servers and to use these servers to provide information for training and to facilitate emergency management efforts after significant severe weather events. Due to the efforts of Dr. Forbes, Matt Pearce, and Robert Much, our office routinely produces detailed storm survey maps and makes these data available on the web.

Other benefits include interactions with students who do case studies in the PSU Meteorology 418W course. Over 18 students have been involved in projects involving severe weather events over the mid-Atlantic region over the past 2 years. These interactions foster learning for the students and the staff of the WFO. These interactions foster scientific discussions, which have spilled over into a PSU/NWS seminar course. This course was conducted during the autumn semesters of 1997 and 1998. In this course, case studies of recent weather events are discussed and related to other published events. This course has helped improve the scientific knowledge at the WFO.

This project exemplified what can be achieved when two meteorological organizations work together. The PSU provided insights and scientific information, which have been rapidly integrated into operations. This raised the level of expectation and scientific knowledge in the WFO. Dr. Forbes routinely shared all the key findings with the WFO through presentations at seasonal and regional workshops.