Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, CLIMATE SCIENCE ( (c) Oxford University Press USA, 2016. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 17 July 2018

History of Convective Storm Science

Summary and Keywords

Convective storms are the result of a disequilibrium created by solar heating in the presence of abundant low-level moisture, resulting in the development of buoyancy in ascending air. Buoyancy typically is measured by the Convective Available Potential Energy (CAPE) associated with air parcels. When CAPE is present in an environment with strong vertical wind shear (winds changing speed and/or direction with height), convective storms become increasingly organized and more likely to produce hazardous weather: strong winds, large hail, heavy precipitation, and tornadoes.

Because of their associated hazards and their impact on society, in some nations (notably, the United States), there arose a need to have forecasts of convective storms. Pre-20th-century efforts to forecast the weather were hampered by a lack of timely weather observations and by the mathematical impossibility of direct solution of the equations governing the weather. The first severe convective storm forecaster was J. P. Finley, who was an Army officer, and he was ordered to cease his efforts at forecasting in 1887. Some Europeans like Alfred Wegener studied tornadoes as a research topic, but there was no effort to develop convective storm forecasting.

World War II aircraft observations led to the recognition of limited storm science in the topic of convective storms, leading to a research program called the Thunderstorm Product that concentrated diverse observing systems to learn more about the structure and evolution of convective storms. Two Air Force officers, E. J. Fawbush and R. C. Miller, issued the first tornado forecasts in the modern era, and by 1953 the U.S. Weather Bureau formed a Severe Local Storms forecasting unit (SELS, now designated the Storm Prediction Center of the National Weather Service). From the outset of the forecasting efforts, it was evident that more convective storm research was needed. SELS had an affiliated research unit called the National Severe Storms Project, which became the National Severe Storms Laboratory in 1963. Thus, research and operational forecasting have been partners from the outset of the forecasting efforts in the United States—with major scientific contributions from the late T. T. Fujita (originally from Japan), K. A. Browning (from the United Kingdom), R. A. Maddox, J. M. Fritsch, C. F. Chappell, J. B. Klemp, L. R. Lemon, R. B. Wilhelmson, R. Rotunno, M. Weisman, and numerous others. This has resulted in the growth of considerable scientific understanding about convective storms, feeding back into the improvement in convective storm forecasting since it began in the modern era. In Europe, interest in both convective storm forecasting and research has produced a European Severe Storms Laboratory and an experimental severe convective storm forecasting group.

The development of computers in World War II created the ability to make numerical simulations of convective storms and numerical weather forecast models. These have been major elements in the growth of both understanding and forecast accuracy. This will continue indefinitely.

Keywords: buoyancy, convection, storms, hail, wind, heavy rain, flash flood, tornado, numerical models, forecasting

Introduction—Definition of Convective Storms and Associated Weather

A convective storm is an atmospheric process driven primarily by the force of buoyancy (Doswell, 2001). Buoyancy is an unbalanced vertical perturbation pressure force associated primarily with spatial variations in density. Although the analogy between convective clouds and a hot-air balloon is often used, Doswell and Markowski (2004) have discussed how this is a problematic comparison. Buoyancy is not defined relative to some hypothetical “environment” surrounding a storm, but rather is associated with a departure (i.e., a perturbation) from a state of hydrostatic balance—a balance between the acceleration due to gravity and the vertical pressure gradient force. Perturbation pressures are the sum of a dynamic part (related to the relationship between pressure and airflow) and a static part (due only to density variations); the latter is discussed by Davies-Jones (2003).

Convective storms, like all atmospheric processes, are a response to a disequilibrium; in this case, a departure from hydrostatic balance. The vertical motions that result from this imbalance act to redistribute heat (and momentum) so as to mitigate the imbalance. In the process, various types of weather events are a by-product of the storm’s effort to restore hydrostatic balance. The state of disequilibrium is referred to as a type of instability, although the terminology used in meteorological practice can be confusing (see Schultz, Schumacher, & Doswell, 2000).

Deep convective storms (i.e., storms that extend through much of the troposphere and occasionally reach into the stratosphere) often are referred to collectively as thunderstorms, although they do not always produce lightning. The energy that powers their up- and downdrafts comes from buoyancy that is dependent on the release/absorption of the latent heat of condensation. Thus, the three necessary ingredients for deep, moist convection are as follows: 1. the presence of water vapor in the ascending air of an updraft; 2. a condition known as conditional instability, in which the lapse rate of temperature (i.e., the rate at which temperature decreases with height) is less than or equal to the dry adiabatic value (9.8 deg C per km) but more than the moist adiabatic value (which becomes asymptotic to the dry adiabatic lapse rate with increasing height, but is roughly 6 deg C per km near the surface); and 3. a source of lift to initiate the storm. The sufficiency of ingredients 1 and 2 can be diagnosed for any particular air parcel being lifted by its Convective Available Potential Energy (CAPE). CAPE is determined by the temperature difference between a parcel ascending adiabatically and the temperature profile in the prestorm environment, using a simplified model of convection called Parcel Theory (Textbook references such as Houze, 1993 or Emanuel, 1994). Since vertical motion is not a reliably measured quantity, the sufficiency of ingredient 3 is problematic in practical terms. In general, convective storm initiation requires lift because of the common presence of convective inhibition (CIN) and the fact that parcels must be lifted to the level of free convection (LFC) before the storm can become sustained.

Weather events of significance produced by deep convective storms include tornadoes, hail, strong horizontal winds, and rainfall. The specific criteria that determine whether a particular hail and/or wind event is called “severe” are mostly a matter of arbitrary choices. In the United States, hail must be ≥ 2.5 cm (one inch) and winds must be ≥ 25 m s−1 (50 knots). Historically, the definition of what wind speeds qualify as severe has been the most troublesome (see Galway, 1989). Unlike most countries, in the United States, heavy rainfall is not considered officially to be severe weather, although the impacts of floods and flash floods in the United States are as large or larger than any of the other severe storm events. On some occasions, a storm may produce all of these hazardous events, although it only takes one of them for a storm to be considered severe.

This introduction to convective storms is necessarily abbreviated, in order to be able to discuss the history of convective storm science. A different treatment can be found in Doswell (2007). In what follows, the rather unique character of convective storm science is described in The Unique Character of Convective Storm Science. Subsequent sections will consider the chronological developments in convective storm science chronologically from the 18th century to the present.

The Unique Character of Convective Storm Science

Anyone familiar with any particular branch of science must recognize that in most, if not all, sciences, there is a dichotomy between basic research science and the application of that research to solve real-world problems. Convective storm science certainly has that division into the abstract and the practical. Unlike some sciences, however, convective storm science has its roots deeply embedded in a compelling need for people to anticipate the weather in the future (i.e., forecasting).

Meteorological science is often said to begin with Aristotle and his work Meteorologica, the name of which is inspired by the root word “meteor” that, at the time of Aristotle, referred to anything in the sky, rather than the more restricted use of that word as an astronomical term in modern times. Aristotle’s work was more of a presentation of his own explanations for meteorological phenomena, rather than the content we would consider necessary for a scientific publication today: the proposing and testing of hypotheses regarding events in the natural world. Thus, the work of ancient Greek philosophers really is but a few steps beyond pure mythology. It offers explanations but offers no quantitative testing of them to check for consistency of the explanations with reality.

During the Renaissance, science evolved from this “natural philosophy” into what we recognize today as science, in which the idea of doing more than mere speculation came to be a necessary element. The great physicists Galileo, Copernicus, Kepler, Newton, and others set the stage for how physical science should be done. It is this testing of ideas that not only is crucial for establishing the validity of some hypothesis, but also forms the basis for practical applications of the new understanding. For example, the inspiration for the great pioneer of thermodynamics, Joseph Black (who is credited with the discovery of latent heat), was the need for efficiency in the distillation of Scotch whisky.

Thus, meteorology, along with its subdiscipline convective storm science, owes much of its motivation for learning about the atmosphere to the needs of those dependent on the weather. That basically includes all of humanity, to a greater or lesser degree for specific individuals. The requirements of the military to deal with the weather, the importance of weather for transportation and commerce, the massive dependency on weather for agriculture—all of these have been the source for scientific motivation and, in recent times, for the funding that allows both basic and applied research to move forward. The importance of weather in many human affairs naturally forces the science to develop a practical component, in addition to learning about the atmospheric just for the sake of adding to our knowledge.

In meteorology, unlike some disciplines, such as particle physics, observations are the key to driving the direction of new hypotheses, rather than seeking observations for the purpose of confirming purely theoretical/mathematical notions. Purely mathematical analysis is not unheard of in convective storm research, but is not typically the source for new insights. Part of the reason for this is that the governing equations in meteorology are distinctly nonlinear, making them unsolvable for most reasonably realistic situations. Analytic solutions of the mathematical formulations almost always require drastic simplifications to work around the nonlinear aspects of the problem under consideration. When primarily mathematical analysis is done, it is almost always solved numerically, so computer-based simulations are an important tool for modern convective storm research. For the most part, in atmospheric science, the observations come first and guide the theoretician in using basic principles to provide explanations for the observations. Thus, improvements in observing technology are primary contributors to new understanding. That is especially so in convective storm science, where the scale of the atmospheric processes demands observations with high resolving power often associated with remote sensing, such as radars and spacecraft.

Eighteenth-Century Convective Storm Science

Benjamin Franklin was a pioneer in several different scientific fields; he is most well known, perhaps, for his famous kite experiment, from which he concluded that lightning was an electrical phenomenon. Less well known is his recognition that weather systems affecting the eastern United States typically moved into the region from the west. This notion, so obvious to us today, was not known at the time, so the idea of weather forecasting using that principle was seminal contribution. Even less well known is an early storm chase anecdote, in which Franklin followed a “whirlwind” on horseback in the spring of 1755 (Cox, 2002, p. 9). He also proposed the idea that convective storms required moisture and instability (Cox, 2002, p. 9ff.).

In an age where most weather observations were primarily visual, convective storm science was strongly limited by the nature of the available observations. The idea that meteorology could be cast as a problem in the mathematical physics of fluids had not yet been proposed. Meteorology was not really much of a quantitative science and so was mostly a hobby topic, except for weather-sensitive efforts, such as mariners and farmers. The ability to use knowledge of atmospheric physic to guide decision-making was not yet advanced enough to be of much interest.

Early European settlers in the eastern United States were quite surprised by the ferocity of thunderstorms and tornadoes in comparison to the weather experienced within those nations from which they came. Although many early American tornado events are documented (Ludlum, 1970), at the time, there was little systematic effort to study them scientifically.

Nineteenth-Century Convective Storm Science

In the early decades of the 19th century, the great French evolutionary biologist Jean-Baptiste Lamarck and the English observationalist Luke Howard developed the first cloud taxonomy more or less simultaneously (Cox, 2002, p. 13ff.). The terminology developed by Howard survives in the modern cloud classification scheme (World Meteorological Organization, 1987), whereas Lamarck’s similar effort using French terminology has largely been forgotten.

William C. Redfield and James P. Espy conducted an acrimonious battle over their scientific ideas regarding cyclones (Cox, 2002; Kutzbach, 1979) that began in 1834 and continued for several years. Redfield emphasized the rotation of cyclones whereas Espy promoted the notion of latent heat release contributions to thermal buoyancy in creating radial airflow. Despite the bitter conflict between them, both perspectives contained elements of truth, as well as some erroneous ideas. Elias Loomis (Cox, 2002, p. 41ff.) showed, in his study of what would become known as an extratropical cyclone (ETC), that cyclones on that scale combined elements of both Espy’s and Redfield’s ideas.

The famed Robert FitzRoy, famous for being Captain of the HMS Beagle during Charles Darwin’s voyage of discovery, began to use weather observations to make storm warnings, starting in 1861 (Cox, 2002, p. 75ff.). Unfortunately, the practice was not “scientific” enough to satisfy the existing standards of the Royal Meteorological Society, and following FitzRoy’s suicide in 1865, the storm warning operations were terminated in 1866.

The world’s first severe convective storm forecaster was John Park Finley (Cox, 2002, p. 101ff.; Galway, 1985). Finley was an Army officer deeply interested in tornadoes and was a pioneer who both studied them using whatever resources he had and began an effort to forecast them in 1884. He recruited storm observers around the United States and was the first to begin a systematic collection of storm reports. He developed a verification scheme for his tornado forecasts that many years later was recognized as a major advance in forecast verification methods (Murphy, 1996). However, similar to FitzRoy’s attempts at storm warnings, his activities were not supported by his superiors and in 1887, Finley was ordered to cease his tornado forecasting operation. In 1889, he was relieved of any duties regarding tornado research and forecasting, and he never returned to his efforts. An official U.S. government ban on the use of the word “tornado” in forecasting was to continue until 1948. The end of Finley’s tornado research and the ban on tornado forecasts had a large negative impact on the climatological record of tornadoes in the United States. By not making an extra effort to seek out storm reports, only those tornadoes with a major impact were recorded and so the knowledge of how many tornadoes were occurring was very limited and, as we now know, was underestimating the frequency dramatically.

Convective Storm Science Before World War I

In the early part of the 20th century, there were several European scientists engaged in studying tornadoes. The most well known of these is Alfred Wegener, a German meteorologist most famed for his prescient idea of continental drift, the forerunner of modern plate tectonics. His work (Wegener, 1917) was well advanced for its time and was something of an anomaly, given that his efforts to document tornadoes over much of Europe were unprecedented (Dotzek, 2001). Wegener analyzed individual tornado events and also began to collect records of tornado occurrences in Germany.

Another prominent European tornado researcher of the era was the meteorologist Johannes Letzmann, of German descent but born in Estonia (Peterson, 1992). Mentored by Wegener himself, Letzmann collected data on European tornado occurrences, including poststorm surveys of the damage, and was intensely involved in studying vortices and the effects of tornadic winds in producing characteristic damage patterns. He wrote extensively about his research and published many papers on a wide variety of topics related to tornadoes and convective storms.

Vilhelm Bjerknes published a seminal work in 1904 (Bjerknes, 1904) that considered how to cast the task of weather forecasting as a topic in basic dynamic of fluids (as applied to the atmosphere). At the time, this was a purely theoretical concept, since the mathematical expression of atmospheric dynamics cannot be solved in a mathematically closed form. It would have to wait until the development of computer-based numerical approximations for this framework for understanding the atmosphere to become a useful tool.

Convective Storm Science After World War I Through the End of World War II

During WWI, a British scientist, Lewis F. Richardson, was serving at the front as an ambulance driver and, while doing so, managed to find time to formulate a way to solve the problem posed by Vilhelm Bjerknes. This work was not published, however, until after the war (Richardson, 1922). Curiously, in Richardson’s book, it is evident that the resulting forecast, laboriously carried out by massive amounts of hand calculations, was a dismal failure (see Lynch, 1992). Despite that, Richardson’s idea was indeed the first conceptual basis for modern numerical weather prediction.

Modern meteorology is widely understood to begin with the founding of the so-called “Bergen School of Meteorology” by Vilhelm Bjerknes (in Bergen, Norway) in 1917 (Cox, 2002, p. 147ff.; Shapiro & Grønås, 1999). Numerous meteorologists from several nations contributed to the Bergen group’s observational studies that culminated with the polar front theory of ETCs. Although not directly about convective storms, ETCs play an important role in creating the large-scale setting for convective storms. Sadly, several potential contributors to the work on both sides were killed in combat during WWI. A dynamical basis for understanding ETCs was created when baroclinic instability was shown by Jule Charney (Charney, 1947) and Eric Eady (Eady, 1949) to explain the physical characteristics of midlatitude ETCs.

Following WWII, given the importance of air power during that war, it became apparent to the military that there was a dearth of knowledge about thunderstorms and their effects on aircraft. This challenge provided the impetus and funding for what was called the Thunderstorm Project (Byers & Braham, 1949), which was an intense field observing campaign held in Florida and Ohio in 1946–1947. This effort gave rise to the notion of a thunderstorm system (Figure 1) made up of multiple “cells,” each at various stages in their individual life cycles of about 20–40 minutes (Figure 2), a concept that has survived the test of time. Moreover, this project became the prototype for many subsequent field observation projects studying convective storms and their immediate surroundings.

History of Convective Storm ScienceClick to view larger

Fig. 1. An example a thunderstorm system with several cells in different stages of their life cycles.

Image from Byers and Braham (1949).

History of Convective Storm ScienceClick to view larger

Fig. 2. Schematic of a thunderstorm cell during its mature stage.

Image from Byers and Braham (1949).

A significant milestone was achieved by two U.S. Air Force officers in March of 1948. Following a damaging March 20 tornado that struck Tinker Air Force Base and did major damage to military aircraft, on March 25 Major Ernest J. Fawbush and Captain Robert C. Miller issued the first tornado forecast in the United States since John Park Finley’s efforts (Maddox & Crisp, 1999). It proved astonishingly successful and the two officers began a pursuit of more understanding aimed at improving their forecasts (e.g., Fawbush & Miller, 1954). Public demand based on leaked Air Force tornado forecasts forced a reluctant U.S. Weather Bureau to establish its own severe convective weather forecasting unit in 1952 (see Corfidi, 1999).

Convective Storm Science in the 1950s

With the founding of the Severe Local Storms (SELS) unit of the U.S. Weather Bureau in 1952, its new director, Donald C. House—named to the post in 1954 after the unit moved to Kansas City, MO—understood that more research was needed if SELS was to improve its forecasts. As discussed by Galway (1992) and Doswell (2007), this led to the founding of the National Severe Storms Project (NSSP), as a part of the SELS unit. The NSSP conducted annual spring observation campaigns patterned roughly on the prototype established by the Thunderstorm Project, with networks of enhanced observational capability, including aircraft, extra soundings (vertical profiles of temperature, pressure wind, and humidity; usually obtained via instrumented balloons) beyond those at operational locations (see Field Programs in Convective Storm Science), radars, and a network of dense surface observations.

The creation of SELS affected the reporting of severe weather, which began to increase dramatically after 1953 (Figure 3).

History of Convective Storm ScienceClick to view larger

Fig. 3. Tornado occurrence frequency from 1916 to 2013, showing the raw numbers for each year (red diamonds and dashed lines) as well as the same after being smoothed somewhat.

Although the dramatic increase in reporting that began in 1953 seems to be leveling out, it is virtually certain that many tornado events are still going unreported (see Brooks et al., 2003; Doswell et al., 2005; Kelly et al., 1978, 1985; Doswell, 2001). Any careful examination of tornado and severe thunderstorm reports in the United States shows many nonmeteorological artifacts. There is as yet no better database on severe convective storm occurrence than those in the United States, but the data are far from perfect even in the United States. Many widespread misconceptions about tornado and severe thunderstorm likelihoods continue to be a concern for safety.

In the early1950s, a new research scientist arrived from Japan, having been invited by University of Chicago professor Horace R. Byers. Tetsuya (Ted) Fujita began to analyze data from a mesonetwork of surface observation stations in the U.S. plains operated by the U.S. Weather Bureau, a work (Fujita, 1955) that ultimately provided a new conceptual model of mesoscale convective systems and their life cycles (Figure 4).

History of Convective Storm ScienceClick to view larger

Fig. 4. Fujita’s conceptual model of a mesoscale convective system.

Image from Fujita (1955).

That network had been put in place to validate an idea proposed by Morris Tepper (Tepper, 1950), but Fujita’s work supplanted Tepper’s hypothesis. Thus began a long, productive career of research by Fujita, which includes so many diverse and important contributions, it is difficult to overstate his value to convective storm science. Much of the scientific terminology of convective storms was created by Fujita, including the cloud structures associated with supercells (Fujita, 1960).

Convective Storm Science in the 1960s

By the early 1960s, discord between researchers and operational forecasters culminated in many members of the NSSP moving out of the building in which SELS was housed, taking their research to Norman, OK, and becoming the National Severe Storms Laboratory (NSSL). In 1963, Edwin Kessler was named the new director of NSSL, appointed with the endorsement of Robert H. Simpson (who was then Deputy Director of Research in the U.S. Weather Bureau—see AMS/UCAR transcript of a 1989 interview with R. H. Simpson—available at OpenSky).

Simpson also was involved with the 1965 appointment of Allen D. Pearson to be Director of the National Severe Storms Forecast Center (including the SELS unit), as the replacement for D. C. House, as well. These two appointments were to have an influence on the interaction between convective storm forecasters and researchers for many years (Doswell, 2007).

Another important research scientist came to the United States in the 1960s: Keith A. Browning had received his doctorate under Frank H. Ludlam (1980), and his dissertation was about a powerful storm that happened in 1959 near Wokingham in England. He was invited by the U.S. Air Force’s Cambridge Research Laboratory to participate in the field program that sampled the storms on May 26, 1963, in Oklahoma (Browning, 1965). Browning recognized the significance of such storms, and was the first to label them “supercells” (Browning, 1962) to distinguish them from ordinary thunderstorms. Browning primarily used radar reflectivity to deduce the airflow in supercells (Browning, 1964).

During the early 1960s, Neil Ward was a tornado researcher in Oklahoma who developed a laboratory tornado simulator to explore tornado dynamics (Ward, 1972). This work stimulated others, including at Purdue (Church et al., 1977), and this work has been useful in understanding vortex breakdown and the existence of multivortex behavior. Numerically modeled tornado vortices have provided quantitative understanding to augment the qualitative results from laboratory simulators (e.g., Rotunno, 1977). Most modern work on tornado dynamics is being done with numerical models rather than laboratory models (e.g., Lewellen & Lewellen, 2007).

The 1970s Revolution in Convective Storm Science

As described by Doswell (2007), the 1970s produced a revolution in convective storm science with three new tools for research: 3-dimensional numerical cloud models (Schlesinger, 1973; Klemp & Wilhelmson, 1978), Doppler radar (Brown et al., 1978), and storm chasing (Golden & Morgan, 1972; Moller et al., 1974). These developments allowed an unprecedented observational and dynamical perspective on storms, especially supercells (but was not limited only to supercell storms).

Numerical modeling provided the opportunity to explore the nonlinear consequences of storm dynamics, including the evolution of the observationally difficult problem of the dynamically important perturbation pressure distribution (Newton, 1963; Rotunno & Klemp, 1982) and its effects on storm structure, evolution, and movement. Storm simulations within a numerical model allowed for complete control over all the experimental variables, so that the effect of the prestorm environment on storms could be determined in quantitative terms (e.g., Weisman & Klemp, 1982).

Doppler radars, which were to become fully operational in the late 1980s, quickly became a critical observational tool in understanding airflows and the internal dynamics of convective storms. In fact, Browning (1977) defined supercells as those convective storms with a mesocyclone. With the implementation of a nationwide network of Doppler radars, many more storms could be sampled in detail than ever before.

The notion of a mobile component for observing storms had been limited to aircraft up until the advent of scientific storm chasing in 1972 with the University of Oklahoma/National Severe Storms Laboratory Tornado Intercept Project (Golden & Morgan, 1972). On May 24, 1973, both Doppler radar and mobile storm chase teams documented a tornado that struck Union City, OK (Figure 5).

History of Convective Storm ScienceClick to view larger

Fig. 5. Shrinking stage of the tornado that struck Union City, OK on May 24, 1973.

Photo copyright © C. Doswell. Used by permission.

This event was the first time a Doppler radar circulation signature, the so-called Tornadic Vortex Signature (TVS), was unambiguously tied to an observed tornado in progress in the operational application of Doppler radar for tornado warnings (Burgess et al., 1975).

This combination of new tools influenced Leslie R. Lemon and Charles A. Doswell III (at the time, both were research forecasters with the Techniques Development Unit—a research unit within SELS) to update the Browning supercell conceptual model (Lemon & Doswell, 1979). In particular, they noted that supercells usually produced a rear-flank downdraft and tornadoes typically occurred on the interface between the storm’s updraft core and the rear flank downdraft (Figure 6).

History of Convective Storm ScienceClick to view larger

Fig. 6. A conceptual schematic diagram, showing low-level airflow and the relationship of the tornado to the vertical drafts within the storm. The so-called forward flank downdraft is denoted by FFD, the horseshoe-shaped updraft is denoted UD, and the rear flank downdraft is denoted by RFD. The outline of the radar echo is shown in green and the outflow boundaries produced by the downdrafts are shown in blue. After Lemon and Doswell (1979).

Convective Storm Research in the 1980s

The advent of geosynchronous satellite imagery as an important forecasting tool also had a large impact on the recognition of the significance of mesoscale convective systems (MCSs). Not only are these systems responsible for severe weather, but they also account for a great deal of the beneficial precipitation during the warm season, as well as flash flood–producing heavy rainfall. The NOAA Atmospheric Physics and Chemistry Laboratory (later disbanded) included several convective storm researchers, notably Robert A. Maddox, J. Michael Fritsch, and Charles F. Chappell. They took notice of the large, circular, persistent MCS they called mesoscale convective complexes (MCCs—Figure 7).

History of Convective Storm ScienceClick to view larger

Fig. 7. Water vapor channel satellite image of an MCC in the central United States, with an MCS near the Appalachian Mountains.

Image courtesy of the National Center for Atmospheric Research, funded by the National Science Foundation.

This led Robert A. Maddox (who had been an Air Force severe weather forecaster, working with Robert C. Miller) to a detailed observational description of MCCs and their surroundings (Maddox, 1980). One facet that proved important was the frequent connection between MCCs and low-level warm thermal advection (Maddox & Doswell, 1982).

Although Maddox focused on MCCs as the most persistent, largest, and most circular of all MCSs, those not meeting the MCC criteria developed by Maddox are still important convective storms. Many flash flood events are tied to MCSs and low-level thermal advection (Maddox et al., 1979), which ties back into some early work by Lynn L. Means (Means, 1944). Most of the flash flood events in the United States are the result of MCSs and MCCs.

Another facet of the weather often associated with MCSs is damaging convective wind gusts. In their most dramatic forms, MCSs can produce widespread areas affected by damaging winds and may meet the criteria for being named a “derecho”—a term coined by Robert H. Johns (Johns & Hirt, 1987), who was a SELS forecaster. Derechos are important because of the vast area affected by wind damage, amounting to tens of thousands of square kilometers. Major derecho events are infrequent (a handful of them in North America during any given year) but their impact can be quite large because of the large areas affected.

Convective Storm Science in the 1990s and Beyond

An important development in convective storm science came when the former SELS unit in Kansas City moved from Kansas City and was reunited with its former research arm, the NSSL at Norman, OK, in 1994. SELS had been renamed the Storm Prediction Center (SPC), and they first moved into the NSSL building in Norman. However, plans were already underway to have the NOAA weather units (the SPC, NSSL, the Norman forecast office, the WSR-88D Radar Operations Center, and the Warning Decision Training Branch) to join several units associated with the University of Oklahoma in the so-called National Weather Center on the university campus. This reunion to support a renewed collaboration between convective storm researchers and forecasters was driven by the recognition of the value of such cooperation between operations and research in the past. As has been noted in the preceding historical presentation, many operationally oriented meteorologists have contributed significantly to advancing the science of convective storms. Many working partnerships have been developed, despite the barriers that often arise in having operational practitioners work in partnership with their counterparts in research. The advances in forecast skill that have been documented by the SPC (Hitchens et al., 2013) over its history are closely connected to advances in convective storm science. In addition to severe convective storm research in Oklahoma, there have been a host of contributions to severe storm science by many forecasters and researchers in the United States (e.g., Bunkers & Stoppkotte, 2007; Klemp & Wilhelmson, 1978; Knupp et al., 2014; Moller, 2001; Parker & Johnson, 2004; Przybylinski, 1995; and numerous others).

The importance of numerical modeling in convective storm science has increased as a result of increases in the capabilities of computing systems. When 3-dimensional convective storm simulation models first began, horizontal grid spacings of 1 km over a limited domain were right on the edge of even a supercomputer’s capacity. Now it is possible to simulate storms with 100-m horizontal grid spacing over a moderately large domain, and to include more sophisticated parameterizations for microphysical processes in convective storms, which have advanced considerably (Dawson et al., 2015).

New observing systems are being implemented that promise improved routine observational capability, such as the Oklahoma “Mesonet” of surface observing systems (Brock et al., 1995). There are many new operational observing systems that serve a variety of needs for a variety of users—the issue is gaining access to the data for either research or operational purposes. For example, there has been a big increase in the availability and use of in-flight meteorological data (AMDAR/ACARS systems—see AMDAR Observing System).

But access to the data has been somewhat limited.

Another interesting development has been the growth of interest in convective storm forecasting and research in Europe. After the late-19th-century and early-20th-century interest in severe convective storms and tornadoes waned, the perception grew that these convective storm events were rare in Europe and were mostly a phenomenon of the United States. The involvement of Europeans in convective storm science dwindled for many decades. Recently, however, the European Severe Storms Laboratory has been established and the biannual European Conferences on Severe Storms have been initiated. Some examples of recent publications related to severe convective storm research by European authors include Antonescu and Burcea (2010), Groenemeijer and Kühne (2014), Ramis et al. (2009), Romero et al. (1988), Setvák et al. (2010) and many others. Considerable effort is currently being focused on the climatology of severe convective storm weather in Europe because the climatological records of severe weather in virtually all of Europe have largely been neglected until the European Severe Storm Conferences began in 2002. The neglect is evidently attributable to a widespread erroneous perception that severe storms and, in particular, tornadoes are uniquely confined to the United States (Doswell, 2015).

Poststorm Surveys

Wegener and Letzmann in Europe were among the first to do field surveys of areas hit by tornadoes in order to understand the storms that produced the damage. By analyzing the damage it became possible to evaluate the structure of strong wind fields as they moved over the area. By the mid-1960s, the U.S. Weather Bureau (now called the U.S. National Weather Service) was producing storm disaster surveys, including a notable example after the Palm Sunday Tornado Outbreak of April 11, 1965 (Weather Bureau Survey Team, 1965). Prof. Ted Fujita from the University of Chicago in Illinois did an independent scientific survey of the event (the analysis of the data can be seen in Fujita et al., 1970). Subsequently, Fujita participated in numerous disaster surveys of major tornado events and established many of the procedures for analysis of the data collected, including aerial overflights, photogrammetric analysis, and wind speed estimation by evaluation of the damage. The latter was established in 1971 as the so-called Fujita Scale, which in somewhat modified form is still in use in the early 21st century, albeit not without some controversy (Doswell et al., 2009). Fujita’s death in 1998 has resulted in a dearth of scientific storm surveys since then (see Speheger et al., 2002). The National Weather Service stopped doing scientific disaster surveys in the early 1990s, and replaced them with “Service Assessments” that focus on the operational service provided by users during a severe storm incident, rather than on the science of storms.

Field Programs in Convective Storm Science

Enhanced-resolution field observing efforts typically produce detailed information about a modest sample of storm events. Alternatively, data from much lower resolution routine observations can provide a much large sample of events but with substantially fewer details. Modern severe storm science uses both approaches. Although tornadic supercell storms have many features in common, each individual storm has unique aspects, so many such storms need to be sampled in order to develop a context for understanding their detailed structure and evolution.

Ever since the Thunderstorm Project, enhanced observational capabilities in studying convective storms and their surroundings has been essential to severe storm science. The spatial and temporal density of observations defines the scale of the phenomena that can be resolved. The classic parable of the three blind men and an elephant makes clear a simple fact: the understanding of some phenomenon that is undersampled can never be comprehensive and may be misleading. Using “conventional” observations (i.e., those observations that are obtained daily, mostly for the purpose of weather forecasting) does not resolve many of the physical processes going on within and near convective storms. If we release a single sounding balloon that ascends within an MCS, the data it gathers are generally useless, representing only a tiny portion of the system. The sounding is considered “contaminated” since it reveals virtually nothing of the internal structure of the MCS.

Some routine remote sensing systems, such as radar or satellite imagery, provide a relatively high-resolution sample of convective storms, but they have severe limitations in revealing the quantitative relationships among key atmospheric variables. Only by incorporating remote sensing observations in a complex process called “data assimilation” associated with numerical model simulations (e.g., Tong & Xue, 2005) can the remote sensing data be used to increase the useful resolution in a quantitative fashion.

Before the development of balloon soundings (Figure 8), kites were used occasionally to sample the atmosphere above the surface (e.g., van Everdingen, 1925).

History of Convective Storm ScienceClick to view larger

Fig. 8. A rawinsonde balloon being launched, showing the balloon, with a red parachute directly below to lower the sonde instruments slowly and safely to the ground after the balloon bursts. The instrument package is at the end of the string beneath the balloon.

NOAA photo.

An operational “rawinsonde” is tracked (originally with a radiotheodolite) to provide wind information (which assumes the sonde is moving with the winds at every level), and carries instruments to measure pressure, temperature, and humidity as the balloon ascends. The data are transmitted to a ground receiving station using a radio within the instrument package. Although rawinsonde balloons are now relatively “old” technology (having first appeared as a tool of value in sampling the atmosphere in relatively primitive form in the late 19th century), they began to see increasing use worldwide in the 1930s, and were in widespread use by the 1950s. For many decades, rawinsondes were the only way to obtain quantitative information about atmospheric structure above the surface. Brock and Richardson (2001) provide a more detailed treatment of atmospheric measurement systems. Modern sounding systems now use GPS tracking for wind calculations and have instrument packages that are much smaller and lighter than those developed in the 1950s. A more or less identical system can be employed on a sonde that descends (slowed by a parachute) from an aircraft—called a dropwinsonde (or dropsonde). Dropwinsondes are used extensively in tropical cyclone forecasting and research, but their use over land is complicated by safety concerns.

Most field programs deploy special soundings to augment the routine soundings collected by the National Weather Service twice daily from a number of fixed sites in the United States (Figure 9).

History of Convective Storm ScienceClick to view larger

Fig. 9. Current routine operational sounding sites with their identifiers over North America, as of late 2015. Those in the United States are shown with a colored background. The average spacing between sites in the United States is roughly 400 km. Most sounding sites have been moved, many more than once, over the life of the sounding network.

Image courtesy of the National Center for Atmospheric Research, funded by the National Science Foundation.

For many years, beginning in 1963, the National Severe Storms Laboratory ran spring field programs involving enhanced observing networks. The issue of hailstorms and possible modification of hail hazards was associated with several field campaigns, including the National Hail Research Experiment (NHRE—see Foote & Knight, 1977) and the Alberta Hail Project from 1956 to 1985.

Convective storm winds can be hazardous, and after several fatal airline crashes (e.g., Fujita & Caracena, 1977), The Joint Airport Weather Studies (JAWS) field campaign was conducted in Colorado. Project leaders were Ted Fujita, John McCarthy, and Jim Wilson (see McCarthy et al., 1982). With the identification of microbursts as an important hazard to aviation, airline accidents associated with microbursts have become relatively infrequent—a testament to the operational value of applying convective storm science to solve practical problems. Further, the importance of extensive wind damage associated with derechos (often produced by storms that take on a “bow echo” configuration) (Figure 10) stimulated a field observational campaign called the Bow Echo and MCV (Mesoscale Convective Vortex) Experiment, or BAMEX, with considerable mobile observing capabilities (Davis et al., 2004).

History of Convective Storm ScienceClick to view larger

Fig. 10. A radar reflectivity image of a derecho-producing bow-shaped echo.

Image courtesy of the National Center for Atmospheric Research, funded by the National Science Foundation.

There have been two major field observing campaigns targeting tornadoes and tornadic storms, one in 1994–1995 (Verification of the Origins of Rotation in Tornadoes Experiment—VORTEX; Rasmussen et al., 1994), and one in 2009–2010 (VORTEX2), as well as numerous other field observational campaigns since the prototypical Thunderstorm Project. These programs had predominantly mobile observing systems—particularly, mobile Doppler radars positioned near storms to obtain the best possible resolution of storms, mobile sounding systems, and mobile instrumented vehicles—to increase the number of events that could be sampled in detail. These field programs have greatly increased our understanding of such important topics as tornadogenesis (e.g., Kosiba et al., 2013), storm environments (Parker, 2014; Weiss et al., 2015), and storm structure/evolution (Rasmussen et al., 2006).

Although tornadic supercell storms have many features in common, each individual storm has unique aspects, so many storms need to be sampled in order to develop a context for understanding them. Mostly due to the relative rarity of tornadic storms in any one location, when NSSL operated a fixed network of sites for their field campaigns, it proved difficult to find well-sampled events; perhaps only one or two cases per year or fewer provided an adequate sample. Mobile observing systems greatly enhance the chances to sample many storm events during a campaign. But of course, they still suffer from the fact that in any given year, there may or may not be many opportunities for storms within a practical observing region. Fixed or mobile field observational campaigns always will be at the mercy of the atmosphere.

Bergeron (1959) proposed that progress in meteorology has resulted from the improvement in three elements: observations, analytical tools, and models. Convective storm science exemplifies that proposition quite well. Whenever all three elements are advancing simultaneously, the most rapid and enduring progress is made. The various “schools” Bergeron described made their contributions precisely because at those times and in those places, the interaction and mutual respect between theoreticians and forecasters were substantial. The current era is one in which it might be said that all three elements are advancing rapidly within the subfield of convective storm science. As shown in the preceding, convective storm science has a long tradition of positive interactions between researchers and forecasters, which bodes well for the future.


Antonescu, B., & Burcea, S. (2010). A cloud-to-ground lightning climatology for Romania. Monthly Weather Review, 138, 579–591.Find this resource:

    Bergeron, T. (1959). Weather forecasting: Methods in scientific weather analysis: An outline in the history of ideas and hints at a program. In B. Bolin (Ed.), The atmosphere and the sea in motion (pp. 440–474). New York: The Rockefeller Institute.Find this resource:

      Bjerknes, V. (1904). Das Problem der Wettervorhersage, betrachtet vom Standpunkte der Mechanik und der Physik (The problem of weather forecasting as a problem in mechanics and physics). Meteorologische Zeitung, 21, 1–7.Find this resource:

        Brock, F. V., Crawford, K. C., Elliott, R. L., Cuperus, G. W., Stadler, S. J., Johnson, H. L., & Eilts, M. D. (1995). The Oklahoma Mesonet: A technical overview. Journal of Atmospheric and Oceanic Technology, 12, 5–19.Find this resource:

          Brock, F. V., & Richardson, S. J. (2001). Meteorological measurement systems. Oxford: Oxford University Press.Find this resource:

            Brooks, H. E., Doswell, III, C. A., & Kay, M. P. (2003). Climatological estimates of local daily tornado probability for the United States. Weather and Forecasting, 18, 626–640.Find this resource:

              Brown, R. A., Lemon, L. R., & Burgess, D. W. (1978). Tornado detection by pulsed Doppler radar. Monthly Weather Review, 106, 29–38.Find this resource:

                Browning, K. A. (Ed.), 1965: A family outbreak of severe local storms – A comprehensive study of the storms in Oklahoma on 26 May 1963, Part I. Special Rep. No. 32, AFCRL-65-695(1), 346.pp. [out of print].Find this resource:

                  Browning, K. A. (1962). Cellular structure of convective storms. The Meteorological Magazine, 91, 341–350.Find this resource:

                    Browning, K. A. (1964). Airflow and precipitation trajectories within severe local storms which travel to the right of winds. Journal of the Atmospheric Sciences, 21, 634–639.Find this resource:

                      Browning, K. A. (1977). The structure and mechanism of hailstorms. In G. B. Foote & C. A. Knight (Eds.), Hail: A review of hail science and hail suppression:Meteorological Monographs 38 (pp. 1–39). Boston: American Meteorological Society.Find this resource:

                        Bunkers, M. J., & Stoppkotte, J. W. (2007). Documentation of a rare tornadic left-moving supercell. E-Journal of Severe Storms Meteorology, 2(2), 1–25.Find this resource:

                          Burgess, D. W., Lemon, L. R., & Brown, R. A. (1975). Tornado characteristics revealed by Doppler radar. Geophysical Research Letters, 2, 183–184.Find this resource:

                            Byers, H. R., & Braham, R. R., Jr. (1949). The thunderstorm. Washington, DC: U.S. Department of Commerce.Find this resource:

                              Charney, J. G. (1947). The dynamics of long waves in a baroclinic westerly current. Journal of Meteorology, 4, 135–162.Find this resource:

                                Church, C. R., Snow, J. T., & Agee, E. M. (1977). Tornado vortex simulation at Purdue University. Bulletin of the American Meteorological Society, 58, 900–908.Find this resource:

                                  Corfidi, S. F. (1999). The birth and early years of the Storm Prediction Center. Weather and Forecasting, 14, 507–525.Find this resource:

                                    Cox, J. D. (2002). Storm watchers: The turbulent history of weather predictions from Franklin’s kite to El Niño. Hoboken, NJ: John Wiley & Sons.Find this resource:

                                      Davies-Jones, R. (2003). An expression for effective buoyancy in surroundings with horizontal density gradients. Journal of the Atmospheric Sciences, 60, 2922–2925.Find this resource:

                                        Davis, C., et al. (2004). The bow echo and MCV experiment. Bulletin of the American Meteorological Society, 85, 1075–1093.Find this resource:

                                          Dawson, D. T., II, Xue, M., Milbrandt, J. A., & Shapiro, A. (2015). Sensitivity of real-data simulations of the 3 May 1999 Oklahoma City tornadic supercell and associated tornadoes to multimoment microphysics. Part I: Storm- and tornado-scale numerical forecasts. Monthly Weather Review, 143, 2241–2265.Find this resource:

                                            Doswell, C. A., III. (2001). Severe convective storms: An overview. In C. A. Doswell III (Ed.), Severe convective storms: Meteorological Monographs 28 (pp. 1–26). Boston: American Meteorological Society.Find this resource:

                                              Doswell, C. A., III. (2007). Historical overview of severe convective storms research. E-Journal of Severe Storms Meteorology, 2(1), 1–25.Find this resource:

                                                Doswell, C. A., III. (2015). Severe convective storms in the European societal context. Atmospheric Research, 158–159, 210–215.Find this resource:

                                                  Doswell, C. A., III, Brooks, H. E., & Dotzek, N. (2009). On the implementation of the enhanced Fujita scale in the USA. Atmospheric Research, 93, 554–563.Find this resource:

                                                    Doswell, C. A., III, Brooks, H. E., & Kay, M. P. (2005). Climatological estimates of daily nontornadic severe thunderstorm probability for the United States. Weather and Forecasting, 20, 577–595.Find this resource:

                                                      Doswell, C. A., III, & Markowski, P. M. (2004). Is buoyancy a relative quantity? Monthly Weather Review, 132, 853–863.Find this resource:

                                                        Dotzek, N. (2001). Tornadoes in Germany. Atmospheric Research, 56, 233–251.Find this resource:

                                                          Eady, E. (1949). Long waves and cyclone waves. Tellus, 1, 33–52.Find this resource:

                                                            Emanuel, K. (1994). Atmospheric Convection. Oxford University Press, 580pp.Find this resource:

                                                              Fawbush, E. J., & Miller, R. C. (1954). The types of air masses in which North American tornadoes form. Bulletin of the American Meteorological Society, 35, 154–165.Find this resource:

                                                                Foote, G. B., & Knight, C. A. (Eds.). (1977). Hail: A review of hail science and hail suppression: Meteorological Monographs 38. Boston: American Meteorological Society.Find this resource:

                                                                  Fujita, T. (1955). Results of detailed synoptic studies of squall lines. Tellus, 4, 405–436.Find this resource:

                                                                    Fujita, T. (1960). A detailed analysis of the Fargo tornadoes of June 20, 1957. U.S. Weather Bureau Research Paper No. 42. Washington, DC: U.S. Government Printing Office.Find this resource:

                                                                      Fujita, T. T., Bradbury, D. L., & van Thullenar, C. F. (1970). Palm Sunday tornadoes of April 11, 1965. Monthly Weather Review, 98, 29–69.Find this resource:

                                                                        Fujita, T. T., & Caracena, F. (1977). An analysis of three weather related aircraft accidents. Bulletin of the American Meteorological Society, 58, 1164–1181.Find this resource:

                                                                          Galway, J. G. (1985). J. P. Finley: The first severe storms forecaster. Bulletin of the American Meteorological Society, 66, 1389–1395, 1506–1510.Find this resource:

                                                                            Galway, J. G. (1989). The evolution of severe thunderstorm criteria within the Weather Service. Weather and Forecasting, 4, 585–592.Find this resource:

                                                                              Galway, J. G. (1992). Early severe thunderstorm forecasting and research by the United States Weather Bureau. Weather and Forecasting, 7, 564–587.Find this resource:

                                                                                Golden, J. H., & Morgan, B. J. (1972). The NSSL-Notre Dame tornado intercept program, Spring 1972. Bulletin of the American Meteorological Society, 53, 1178–1180.Find this resource:

                                                                                  Groenemeijer, P., & Kühne, T. (2014). A climatology of tornadoes in Europe: Results from the European Severe Weather Database. Monthly Weather Review, 142, 4775–4790.Find this resource:

                                                                                    Hitchens, N. M., Brooks, H. E., & Kay, M. P. (2013). Objective limits on forecasting skill of rare events. Weather and Forecasting, 28, 525–534.Find this resource:

                                                                                      Houze, R. A., Jr. (1993). Cloud Dynamics. Academic Press, 573pp.Find this resource:

                                                                                        Johns, R. H., & Hirt, W. D. (1987). Derechos: Widespread convectively induced windstorms. Weather and Forecasting, 2, 32–49.Find this resource:

                                                                                          Kelly, D. L., Schaefer, J. T., & Doswell, III, C .A. (1985). Climatology of nontornadic severe thunderstorm events in the United States. Monthly Weather Review, 113, 1997–2014.Find this resource:

                                                                                            Kelly, D. L., Schaefer, J. T., McNulty, R. P., Doswell, III, C. A., & Abbey, Jr., R. F. (1978). An augmented tornado climatology. Monthly Weather Review, 106, 1172–1183.Find this resource:

                                                                                              Klemp, J. B., & Wilhelmson, R. B. (1978). The simulation of three-dimensional convective storm dynamics. Journal of the Atmospheric Sciences, 35, 1070–1096.Find this resource:

                                                                                                Knupp, K. R., Murphy, T. A., Coleman, T. A., Wade, R. A., Mullins, S. A., Schultz, C. J., et al. (2014). Meteorological overview of the devastating 27 April 2011 tornado outbreak. Bulletin of the American Meteorological Society, 95, 1041–1062.Find this resource:

                                                                                                  Kosiba, K. J., Wurman, J., Richardson, Y., Markowski, P., Robinson, P., & Marquis, J. (2013). Genesis of the Goshen County, Wyoming, tornado on 5 June 2009 during VORTEX2. Monthly Weather Review, 114, 1157–1181.Find this resource:

                                                                                                    Kutzbach, G. (1979). The thermal theory of cyclones. Boston: American Meteorological Society.Find this resource:

                                                                                                      Lemon, L. R., & Doswell, III, C. A. (1979). Severe thunderstorm evolution and mesocyclone structure as related to tornadogenesis. Monthly Weather Review, 107, 1184–1197.Find this resource:

                                                                                                        Lewellen, D. C., & Lewellen, W. S. (2007). Near-surface intensification of tornado vortices. Journal of the Atmospheric Sciences, 64, 2176–2194.Find this resource:

                                                                                                          Ludlam, F. H. (1980). Clouds and storms: The behavior and effect of water in the atmosphere. University Park: Pennsylvania State University Press.Find this resource:

                                                                                                            Ludlum, D. M. (1970). Early American tornadoes: 1586–1870. Boston: American Meteorological Society.Find this resource:

                                                                                                              Lynch, P. (1992). Richardson’s barotropic forecast: A reappraisal. Bulletin of the American Meteorological Society, 73, 35–47.Find this resource:

                                                                                                                Maddox, R. A. (1980). Mesoscale convective complexes. Bulletin of the American Meteorological Society, 61, 1374–1387.Find this resource:

                                                                                                                  Maddox, R. A., Chappell, C. F., & Hoxit, L. R. (1979). Synoptic and meso-α‎ scale aspects of flash flood events. Bulletin of the American Meteorological Society, 60, 115–123.Find this resource:

                                                                                                                    Maddox, R. A., & Crisp, C. A. (1999). The Tinker AFB tornadoes of March 1948. Weather and Forecasting, 14, 492–499.Find this resource:

                                                                                                                      Maddox, R. A., & Doswell, III, C. A. (1982). An examination of the jet stream configurations, 500 mb vorticity advection and low-level thermal advection patterns during extended periods of intense convection. Monthly Weather Review, 110, 184–197.Find this resource:

                                                                                                                        McCarthy, J., Wilson, J. W., & Fujita, T. T. (1982). The Joint Airport Weather Studies Project. Bulletin of the American Meteorological Society, 63, 15–22.Find this resource:

                                                                                                                          Means, L. L. (1944). The nocturnal maximum occurrence of thunderstorms in the Midwestern states. Misc. Reports No. 16. Chicago: University of Chicago.Find this resource:

                                                                                                                            Moller, A., Doswell, C., McGinley, J., Tegtmeier, S., & Zipser, R. (1974). Field observations of the Union City tornado in Oklahoma. Weatherwise, 27(4), 68–79.Find this resource:

                                                                                                                              Moller, A. R. (2001). Severe local storms forecasting. In C. A. Doswell III (Ed.), Severe convective storms: Meteorological Monographs 28 (pp. 433–480). Boston: American Meteorological Society.Find this resource:

                                                                                                                                Murphy, A. H. (1996). The Finley affair: A signal event in the history of forecast verification. Weather and Forecasting, 11, 3–20.Find this resource:

                                                                                                                                  Newton, C. W. (1963). Dynamics of Severe Convective Storms. Meteorological Monographs, 5(27), 33–58.Find this resource:

                                                                                                                                    Parker, M. D. (2014). Composite VORTEX2 supercell environments from near-storm soundings. Monthly Weather Review, 142, 508–529.Find this resource:

                                                                                                                                      Parker, M. D., & Johnson, R. H. (2004). Structures and dynamics of quasi-2D mesoscale convective systems. Journal of the Atmospheric Sciences, 61, 545–567.Find this resource:

                                                                                                                                        Peterson, R. E. (1992). Johannes Letzmann: A pioneer in the study of tornadoes. Weather and Forecasting, 7, 166–184.Find this resource:

                                                                                                                                          Przybylinski, R. W. (1995). The bow echo: Observations, numerical simulations, and severe weather detection methods. Weather and Forecasting, 10, 203–218.Find this resource:

                                                                                                                                            Ramis, C., Romero, R., & Homar, V. (2009). The severe thunderstorm of 4th October 2007 in Mallorca: An observational study. Natural Hazards and Earth System Sciences, 10, 1237–1245.Find this resource:

                                                                                                                                              Rasmussen, E. N., Straka, J. M., Davies-Jones, R., Doswell, III, C. A., Carr, F. H., Eilts, M. E., & MacGorman, D. R. (1994). Verification of the origins of rotation in tornadoes experiment: VORTEX. Bulletin of the American Meteorological Society, 75, 994–1006.Find this resource:

                                                                                                                                                Rasmussen, E. N., Straka, J. M., Gilmore, M. S., & Davies-Jones, R. (2006). A preliminary survey of rear-flank descending reflectivity cores in supercell storms. Weather and Forecasting, 21, 923–938.Find this resource:

                                                                                                                                                  Richardson, L. F. (1922). Weather prediction by numerical process. Cambridge, U.K.: Cambridge University Press.Find this resource:

                                                                                                                                                    Romero, R., Ramis, C., & Alonso, S. (1988). Performance of two cumulus convection parameterizations for two heavy precipitation events in the western Mediterranean. Meteorology and Atmospheric Physics, 66, 107–214.Find this resource:

                                                                                                                                                      Rotunno R. (1977). Numerical simulation of a laboratory vortex. Journal of the Atmospheric Sciences, 34, 1942–1956.Find this resource:

                                                                                                                                                        Rotunno, R., & Klemp, J. B. (1982). The influence of the shear-induced pressure gradient on thunderstorm motion. Monthly Weather Review, 110, 136–151.Find this resource:

                                                                                                                                                          Schlesinger, R. E. (1973). A numerical model of deep moist convection. Part I: Comparative experiments for variable ambient moisture and wind shear. Journal of the Atmospheric Sciences, 33, 1021–1041.Find this resource:

                                                                                                                                                            Schultz, D. M., Schumacher, P. N., & Doswell, III, C. A. (2000). The intricacies of instabilities. Monthly Weather Review, 128, 4143–4148.Find this resource:

                                                                                                                                                              Setvák, M., Lindsey, D. T., Novák, P., Wang, P. K., Radová, M., Kerkmann, J., et al. (2010). Satellite-observed cold-ring-shaped features atop deep convective clouds. Atmospheric Research, 97, 80–96.Find this resource:

                                                                                                                                                                Shapiro, M. A., & Grønås, S. (Eds.). (1999). The life cycles of extratropical cyclones. Boston: American Meteorological Society.Find this resource:

                                                                                                                                                                  Speheger, D. A., Doswell, III, C. A., & Stumpf, G. J. (2002). The tornadoes of May 1999: Event verification in central Oklahoma and related issues. Weather and Forecasting, 17, 362–381.Find this resource:

                                                                                                                                                                    Tepper, M. (1950). A proposed mechanism of squall lines: The pressure jump line. Journal of Meteorology, 7, 21–29.Find this resource:

                                                                                                                                                                      Tong, M., & Xue, M. (2005). Ensemble Kalman filter assimilation of Doppler radar data with a compressible nonhydrostatic model: OSS experiments. Monthly Weather Review, 133, 1789–1807.Find this resource:

                                                                                                                                                                        van Everdingen, E. (1925). The cyclone-like whirlwinds of August 10, 1925. Proceedings of the Royal Academy of Amsterdam Section of Science, 25, 871–889.Find this resource:

                                                                                                                                                                          Ward, N. B. (1972). The exploration of certain features of tornado dynamics using a laboratory model. Journal of the Atmospheric Sciences, 29, 1194–1204.Find this resource:

                                                                                                                                                                            Weather Bureau Survey Team. (1965). Report of Palm Sunday tornadoes of 1965. Washington, DC: U.S. Department of Commerce, Weather Bureau.Find this resource:

                                                                                                                                                                              Wegener, A. L. (1917). Wind- und Wasserhosen in Europa [Tornadoes and waterspouts in Europe]. Braunschweig, Germany: Vieweg.Find this resource:

                                                                                                                                                                                Weisman, M. L., & Klemp, J. B. (1982). The dependence of numerically simulated convective storms on vertical wind shear and buoyancy. Monthly Weather Review, 110, 504–520.Find this resource:

                                                                                                                                                                                  Weiss, C. W., Dowell, D. C., Schroeder, J. L., Skinner, P. S., Reinhart, A. E., Markowski, P. M., & Richardson, Y. (2015). A comparison of near-surface buoyancy and baroclinity across three VORTEX2 supercell intercepts. Monthly Weather Review, 143, 2736–2753.Find this resource:

                                                                                                                                                                                    World Meteorological Organization. (1987). International cloud atlas (Vol. 2). Geneva: World Meteorological Organization.Find this resource: