Home Page
Storm Center
  Storm Center
Severe Weather Safety
  Safety Information
FAQ page
  F A Q's
About our work
  About Our Work
Mesoscale Discussions
  Short Term Forecasts
Winter Storm Information
  Winter Storm Center
Contact Info
  Contact Information
Our safety quiz
  Our Safety Quiz


The tornado, perhaps one of the least understood weather phenomena known to man since the earliest of times. Tornadoes have been around since the eighteenth(18th) century, ever earlier, but back in those years, man knew little about them, where they came from, or how they formed. In fact, many people only knew they were a bad windstorm associated with a thunderstorm. It was'nt until 1870 that the Army Signal Corps begin intense research efforts to document and understand severe local storms, like tornadoes. Leading the Corps in this endeavor was Sgt. John P. Finley. In the mid 1880's Finley organized a team of more then 2000 so called "reporters" over the central and eastern states to document tornadoes, and associated weather patterns and conditions. Using this data, Finley produced charts showing charactoristic tornado producing patterns, which were used in tornado "alerts". But by the late 1880's, Finley's forecasts fell out of favor with the Corps, and eventually the "Weather Bureau" because they felt the use of the word "tornado" would provoke panic and undue fear with the public. As a result, a ban was placed on the use of the word tornado in their forecasts. Little progress was made in tornado forecasts into the early 1900's. By 1948, the Weather Bureau along with the Army Navy Corps formed a special unit which would begin forecasting severe local storms. However, our knowledge was still fairly limited in this field, and many forecasters were not trained enough. But, in 1950, that unit moved from Washington D.C. to Kansas City Missouri, and was renamed SELS "Severe Local Storms Warning Center".
From 1950 onward, our knowledge and understanding into some of the general atmospheric conditions conducive to severe local storm development begin to improve. By the early 1960's new IBM computer systems were installed at SELS which became the "National Severe Storms Forecast Center" in June of 1966. These computer systems allowed forecasters to do various analysis of charts which were difficult to compute manually. These computers, along with researchers in the field to gather data on pre storm conditions, would ultimately pave the way for an increased understanding of such destructive storms, and improve warning lead times. Indeed, we have come a long way since the birth of SELS back in 1950 by the WBAN which means "Weather Bureau Army Navy" headquarters in Washington D.C. We have learned extensively about tornadoes, and some of the pre storm environments which may ultimately lead to their formation. With the commissioning of the WSR-88D (Doppler) weather radar and advanced satellites, combined with newer and more advanced computer systems, tornado warning lead times have improved markedly since the 1980's. Yet, despite all these advances in both technology and our knowledge of such storms, we still have much more to learn. For example, we still need to understand how tornadoes may develop from mesocyclones within supercell thunderstorms. We also need to better refine our existing forecast methods so that not only tornado warning lead times can be improved more, but also our ability to more accurately time the development of such thunderstorms. This may one day, allow us to predict where and when a severe thunderstorm or tornado may happen with more precision then today's technology offers.
On this page, you will learn about the basics of tornadoes, what we know about their causes, how scientists and highly skilled meteorologists predict tornadoes, and some of the tools and outside agencies who helped gather much needed data to learn what we know about tornadoes today.


By definition, a tornado is a violent twisting windstorm which is comprised of a funnel shaped pendent that extends from the base of a thunderstorm to the surface. It is not classified as a tornado unless the funnel has contact with the ground. Tornadoes take on many different shapes and sizes, and can form as a single storm, or in groups of twisters. Most tornadoes travel from the southwest to northeast, although they also have come from the northwest to the southeast. The speeds at which tornadoes travel on the ground ranges from stationary to 70 mph. The winds around a tornado funnel have never been measured accurately since the wind instruments never survived a tornado. However, estimates from tornado damage places the speed roughly over 260 mph. A tornado that forms or moves over a body of water is called a "waterspout". Tornadoes can happen at anytime or anywhere, but are most common during the middle to late afternoon and evening hours. Tornadoes have occured in every month of the year. In figure 1 we can see a regional breakdown of tornadoes in our coverage region since 1980. Also shown are number of deaths and injuries.


The thing people really needs to keep in mind is that tornadoes form very suddenly, and often with little or no prior warning. There are several ways to know when tornadoes are possible, and we'll explain that a bit later. Historically, tornadoes have wrecked havoc on many parts of the nation, especially in prone locations like "tornado alley". There have been many documented cases of tornado events which were quite memorable. One such case was the twister which struck Barneveld Wisconsin on June 8th, 1984. For more information on this event, please follow this link. Another case involves the twister which struck Oakfield Wisconsin on July 18th, 1996. For more information on this event, just follow this link. Yet another case involves a strong tornado which struck Door county on August 23rd, 1998. This is regarded as the strongest storm in northeast Wisconsin. For more information on this event, just follow this link. A final case involves destructive tornadoes and very large hail over central and northeast Wisconsin on June 7th, 2007. You can read all about this event here.
These are just a few of the many documented cases of strong tornadoes occuring in Wisconsin. There are many other similar cases reported over the United States and across our coverage region.


While we have learned extensively from recent years and decades from researches, scientists, and storm chasers, and the data collected in the field to further study tornadoes and their causes, we were able to begin predicting the "general" areas most conducive to their development already back in early 1950's. Yet today, even with all we know about tornadoes, the most we can do is predict the general areas where they most likely will occur. The specific elements, or parameters needed to generate tornadoes also remains not fully understood as well. However, ongoing research at the "National Severe Storm Laboratory" in Norman, Oklahoma may provide some additional insight on this within several more years from now. What we can tell you is what we believe may help in generating a tornado. As most people should already know from basic science classes in high school, thunderstorms are really a more forceful (violent) example of convection. For a thunderstorm to develop, three primary ingredients need to be present. Those include:

  • Moisture. This is the energy source for thunderstorm development.
  • Instablilty. In order for any thunderstorm to develop, the atmosphere needs instability.
  • Lifting. A lifting mechanism to get the warm buoyant air rising into the atmosphere.
  • As we know, warm and moist air is less dense, so it typically rises up into the atmosphere. The further it rises, it begins to cool and condense into clouds which tower up into the atmosphere. The rising currents of air are called an "updraft". There are also currents of air which flows downward through a thunderstorm as well. Those currents are called "downdrafts". We can see an example of towering clouds in figure 2 below. For a tornado to possibly evolve, scientists believe that a process called "wind shear" needs to be present in addition to moisture and instability. There are two main types of wind shear we know of presently, directional and speed shear. Directional wind shear basically is wind that changes direction with height. Speed shear is wind speed which changes with height. Let's take a closer look at this. Typically, the winds blow from different directions and speeds as you go higher in the atmosphere. In figure 3 below, we display such a setup known as "cross shearing". Notice how the winds blowing in different directions with height (blue and red arrows), creates a horizontal "tube" like figure of air which spins (light blue arrowed circles) about on a horizontal axis. As the storm strengthens, the updraft currents become stronger inside it. This eventually forces the horizontal spinning air to become stretched upward into a nearly vertical position shown in figure 4. This causes the spin action to transition from a horizontal axis to a vertical axis within the storm. This now vertical axis of wind begins shifting downward through the storm, and in some cases causes the lower portion of the storm to develop a broader circulation, which can extend outward in all directions from the storm itself. This is called a "Mesocyclone". While we do know that tornadoes develop from a mesocyclone, we still don't fully understand how mesocyclones actually contribute to tornado development. As mentioned above in the list of elements needed for thunderstorm development, the first thing we need is a supply of deep rich low level moisture. This is illustrated in figure 2 below.

    FIG. 2

    Next, we need an unstable atmosphere. Instability is measured in many different ways, including the "lifted Index", "CAPE", which stands for Convective Available Potential Energy, wind shear, turbulence, and others. Wind shear is measured in two primary ways, "directional" and "speed" shear. But, a third way to measure shear is referred to a "cross shear". Each of these is equally important in determining how unstable the environment is. To illustrate wind shear, we provide the examples shown below in figures 3 and 4.

    FIG. 3 Directional Shear

    FIG. 4 Speed Shear

    To illustrate how convection works, we have provided the following simplified graphic. As you can clearly see, warm buoyant air near the surface is less dense, so as daytime heating occurs, it rises up into the atmosphere, where it encounters colder and drier air. At the same time, colder and drier air is more dense, so it sinks downward torward the surface. This is how the convective process works, warm moist air upward, and colder drier air downward, as shown in figure 4b below.

    FIG. 4b

    With an unstable environment now established, we next examine how the rising moist parcels of air cool and condense into clouds, and the continued rising of the moist air. This is illustrated in figure 5 below. As the moist air continues rising, it creates an upward current of air known as an "updraft". This helps the clouds to soar higher into the atmosphere. Once the clouds reach their highest point, the tops of the clouds begin to spread out horizontally into a thin layer of cirrus clouds (anvil) by the flow aloft. At the same time, a downward current of air within the storm clouds is established, known as a "downdraft". As positive and negative charges build up in the storm, lightning begins and rain falls to the surface. This is illustrated in figure 6 below. It is when our storm transitions from it's mature stage to the severe stage when destructive winds, hail, tornadoes, and flash floods occur. Finally, the downdraft current inside the storm "cuts off" the updraft current which supplies the needed moisture (fuel) for our storm to survive. Hence, the storm slowly begins to spread out and diminish, then die out completely, shown in figure 7. Depending on how much instability still exists within the surrounding environment, new storms may be developing well after our storm died out. This was a basic explaination of the life cycle of a thunderstorm.

    FIG. 5

    FIG. 6

    FIG. 7

    Up to this point, we have discussed some of the ingredients scientists believe contributes to supercell and tornadic development. There are other parameters looked at and evaluated in determining the potential for tornadoes to develop, but to what degree those parameters actually have in the formation of tornadoes is unknown. Continued evaluations of such parameters as "storm relative helicity", "energy-helicity index", and the "significant tornado parameter" may provide some additional answers. Model simulations from data collected in the field of pre storm environments, as well as those data collected during major tornado outbreaks are helping scientists and meteorologists at the National Severe Storms Laboratory learn more about what happens just prior to a tornado inside the storm, and more about their behavior and other traits. These things could pave the way for an even better warning system then currently in place, as well as vastly improve our severe weather prediction methods.

    Tornadoes can come in an assortment of shapes and sizes, and can occur as a single storm, or multiple tornadoes at the same time. Sometimes, smaller vortices form inside and rotate around the periphery of the main tornado funnel. These vortices were known as suction vortices, but based on field data collected from Project "VORTEX" back in the late 1980's, it was discovered that such vortices nor the funnel itself did any suction. Rather, the violent wind just "blows" the debries everywhere. On occasions, multiple funnels may form in conjunction with a tornado touchdown. As discussed above, these funnels spin around the "core" of a larger and more transparent funnel. Such storms are known as "multi-vortex" tornado. This is illustrated in figure 8 below.

    FIG. 8

    The late Dr. Ted Fujita, who devised the orignal tornado intensity scale back in 1972, also did a study about multiple vortex tornadoes. He concluded that the suction vortices actually pivot around the outermost periphery of the main tornado funnel, but can also move inside the main vortex as well. This was believed to be accurate until field data collected from Project Vortex (mentioned above), disputed Fujita's initial theory about suction vortices, and concluded that the funnel does not create or have any suction. Instead, the winds just blow debries around, and don't suck anything up into it.
    Figure 9 below shows Dr. Fujita's theory about multi-vortex tornadoes.

    FIG. 9

    Again, as cited above, we have made great strides in understanding tornadoes, and their pre storm environments. We have learned more about those elements which leads to the development of a more violent breed of thunderstorms known as "supercells". Most tornadoes occur with supercells, although some have also occurred with multicell or even squall line thunderstorms. Advances in technology have paved the way for a much improved warning system, and better tools for their prediction. However, we are still far from being able to time the development of such events, or where they may occur.


    The prediction of tornadoes as well as related small mesoscale features has always been one of the biggest challenges forecasters have faced for years and decades. In the past 60 years since 1950, forecasters have been able to show the broader areas where the potential is highest for tornado development, and this still holds true today. During the past 30 years since 1980, the National Weather Service was reorganized and equipped with newer, more powerful and advanced tools for severe storm forecasting, including Doppler radar, GOES 12-14 weather satellites, the Advanced Weather Information Processing System (AWIPS for short), and a host of new supercomputers to create better model simulations of coming weather patterns across the conus (that is, continential U.S). So then, just what does a forecaster look for in order to predict severe local storms and tornadoes?

    The task of predicting severe convective weather involves many hours of analyzing and plotting charts, soundings, gridding deterministic model data, analyzing upper level winds and features as well as at the surface, and much more. A forecaster must sort out hundreds of pieces of data and information in order to determine where the atmosphere has become "unstable". When the atmosphere becomes unstable, this means the environment is favorable for severe weather to develop. Once an area(s) are identified as being unstable, the task of forecasting begins! Back in the 1950's and 60's, there were several methods of forecasting severe thunderstorms versus tornadoes which were based largely on different techniques and parameters developed by various forecasters and others involved in the forecasting aspect of these events. Thus, when certain parameters and other conditions existed, the potential favoring severe thunderstorms was mentioned in the outlooks. Then, when many or all parameters and other conditions existed, the potential favoring both severe thunderstorms and tornadoes was mentioned in the outlooks. All this work back then was done at SELS (Severe Local Storm Warning Center), which eventually was renamed the "National Severe Storms Forecast Center" in 1966. Throughout the coming 20 year period, forecasting skills improved markedly with the computers and other available diagnostic systems of the time. In 1974, some researchers and storm chasers from the University of Oklahoma and the "National Severe Storms Laboratory" set out in the field to gather data about pre-storm conditions which leads to tornadic development, and then study how tornadoes form and move. The idea here was to learn more about tornadoes, and the environmental conditions that cause them.

    In figures 10 and 11 below, we get a view of the operations areas of the "National Severe Storms Forecast Center" in Kansas City, Missouri, (formally known as SELS), where all convective outlooks, and severe weather watches originate from. The center can issue two types of watches -- tornado or severe thunderstorm. Watches are issued to provide advance notice to the public of possible severe weather development in their area over the next several hour period, and to review basic severe weather safety. In the meantime, you should keep an eye on the sky, and watch for signs of storm development, but go on with your normal routine. Watches are NOT warnings! Watches are issued to give you extra lead time to prepare for possible severe thunderstorms and tornadoes. Warnings mean that a tornado has been sighted on the ground, or detected on Doppler radar, near your location. This is the time to take cover.

    FIG. 11 NSSFC

    While a lot of data was collected and formed into model simulations for further study, and our knowledge continued to slowly increase, our warning system was very poor at best. The average lead time for a tornado warning was only 3 minutes! That's right, 3 minutes. Continued research work both in the field and at the NSSL would eventually lead to "Project Vortex" being formed in 1996. This group was made up of a number of different people from storm chasers to meteorologists to research scientists, and others. Their mission was to deploy a primary measuring device nicknamed "TOTO" (TOtable Tornado Observatory), into the path of an oncoming tornado. Onboard instruments would measure wind speed, air pressure, humidity, temperature, and the like. The deployment of this device would however, prove unsuccessful owing to weight constraints, and increased lightning strike potential of handling a metal object in open areas. Toto was retired in 1998, and is on display at the NSSL in the National Weather Center building in Norman, Oklahoma. Figure 12 shows Toto as it appeared in 1995 just before use in Project Vortex.

    FIG. 12

    Also during this timeframe (1995-96), the National Weather Service was reorganized with all new buildings, office equipment, radar, satellite, and much more. These advances would allow for significant improvements in forecasts and warnings to the public. With the badly aged and outdated WSR-57c radar and satellite systems replaced by the new WSR-88D or "Doppler Radar" and GOES satellites, the average lead time now for tornado warnings is about 14 minutes, compared to just 3 minutes using the older system. The highly advanced AWIPS system and supercomputer models allow forecasters to "see" the upcoming weather patterns, and determine the severe weather potential. Still, our hats are off to all those who either directly or indirectly participated in the research of tornadoes to help increase our knowledge and understanding of these violent storms and their pre storm environments. This also includes those with Project Vortex, and others as seen in figure 13.

    FIG. 13

    This page has looked at what tornadoes are, and provides some basic charactoristics. It also covered how not only a thunderstorm forms, but the primary factors which often lead to tornadic development inside a supercell several minutes prior to their touchdown on the surface. While we do know that many tornadoes are born out of the mesocyclone of supercell thunderstorms, we still don't understand the direct correlation of how the mesocyclone actually produces a tornado. Finally, we looked at how the forecasting of severe thunderstorms and tornadoes has evolved over the years from 1950 onward. Beginning with SELS, we learned how to recognize and forecast areas where severe weather might form. We developed scientific formulas to calculate different parametes (such as: lifted index, K-index, Showalter index, TT index, and others). As the computer age arose in the early 1960's, we soon were able to create diagnostic charts, and compute many sets of data which were nearly impossible to do manually. Newer computers of the 1970's and 80's paved the way for a wider array of diagnostic and prognostic charts, newer formulas for other parameters (such as: Helicity, EHI or Energy Helicity Index, Supercell Composite Index, Tornado Composite Index, and others). Our forecast methods were also improved with more skill and accuracy. Today, all this, and more is applied to every forecast and watch product issued by the "Storm Prediction Center" in Norman, Oklahoma.

    Web site and all contents Copyright Midwest Weather Service Storm Center 2015, All rights reserved.