A Brief History of the Storm Prediction Center

Stephen Corfidi

The NOAA Storm Prediction Center (SPC) prepares forecasts of hazardous weather affecting the continental United States. The SPC, formerly known as the National Severe Storms Forecast Center, is a component of the National Weather Service's (NWS) National Centers for Environmental Prediction (NCEP).[1] In addition to issuing tornado and severe thunderstorm watches on an as-needed basis, the SPC also produces scheduled severe weather and fire weather outlooks, as well as short-term forecasts for heavy rain and winter storms. The SPC operates around-the-clock, with an on-duty staff that varies between 3 and 5 forecasters depending on the time of day.

Early Severe Weather Forecast Efforts

Although SPC's immediate history dates to the early 1950s, the roots of severe weather prediction in the United States may be traced much further. The development of a centralized weather forecast program by the U. S. Army Signal Corps in 1870 made apparent the need for improved documentation and increased understanding of destructive local storms. Leading the Corps in this effort was Sgt. John P. Finley. In the mid 1880s, Finley organized a team of more than 2000 "reporters" to document tornadoes and their associated weather conditions over the central and eastern United States. Using the data thus collected, Finley assembled maps of characteristic tornado-producing weather patterns that were then used to issue tornado "alerts." Finley's forecasts fell out of favor, however, in the late 1880s as the Corps (and, later, the Weather Bureau, predecessor of the NWS) felt that mention of the word "tornado" provoked undue fear amongst the public.

Little progress was made in the understanding and forecasting of severe local storms in the United States during the first part of the 1900s. Although forecasts occasionally mentioned the potential for severe weather, Weather Bureau policy continued to prohibit use of the word "tornado" in forecasts. Airplane and kite observations sparked renewed interest in severe weather in the 1920s and 1930s. Interest increased with the development of radiosondes and the growth of military aviation in World War II. Nevertheless, even though the ban on the word "tornadoes" was lifted in 1938, very few forecasts made mention of tornadoes during the 1940s.

1948 witnessed the single event that most directly led to the establishment of a centralized severe weather forecast program in the United States. Based on work by Weather Bureau researchers A. K. Showalter, J. R. Fulks, and others --- and on their own investigation of the conditions that produced a damaging tornado at Tinker Air Force Base in Oklahoma City on 20 March 1948 --- Air Force weather officers E. J. Fawbush and R. C. Miller successfully predicted the fortuitous occurrence of another tornado at the base five days later on 25 March. The forecast's accuracy drew considerable attention; soon the officers were responsible for Air Force tornado prediction over much of the central United States. Three years later, the Severe Weather Warning Center, a formal Air Weather Service unit with responsibility for all Air Force sites on the United States mainland, was established under Fawbush and Miller's leadership.

The Birth of SELS

The success of the Air Force tornado program --- along with media pressure to adopt the program for civilian use --- led the Weather Bureau to establish its own severe weather unit on a trial basis at the Weather Bureau-Army-Navy (WBAN) Analysis Center in Washington, DC in March 1952. Fifteen forecasters, including members of the WBAN analysis staff and others from the Bureau's central office and field stations, were selected to staff the unit. Several weeks of techniques development and practice forecasts preceded the release of the unit's first public tornado "bulletin" on 17 March. This forecast mentioned the possibility of tornadoes in parts of Texas, Oklahoma, Arkansas and Louisiana during the late night and early morning of the 17th - 18th. Although two tornadoes did occur in Texas, they were not in the outlook area. The group experienced more success with its second forecast, which was issued on 21 March for parts of east Texas, southern Arkansas, southeast Oklahoma and northern Louisiana. An update extended the forecast into parts of Tennessee, Kentucky and Indiana. Thirty-six tornadoes that began during the afternoon and continued through the night caused 208 deaths in Arkansas, Mississippi, Louisiana, Alabama, Tennessee, Kentucky and Missouri.

The WBAN severe weather operation became permanent on 21 May 1952 when the group was formally recognized as the Weather Bureau Severe Weather Unit (SWU). Forecast responsibility that had been limited to tornadoes was now expanded to include large hail, high winds, and extreme convective turbulence.

Five permanent SWU forecasters were selected during the summer of 1952 to provide continuous shift coverage; temporary staff continued to cover shifts as necessary until the permanent staffing was completed in September. The new forecasters were young; most had been with the Weather Bureau less than ten years and had attended meteorology school with the military during World War II. Comparatively new forecasters were intentionally chosen as it was thought that they would be less likely to harbor preconceived notions about severe storm prediction. Three of the original five permanent SWU forecasters left the group before its move to Kansas City in 1954. Only Joseph Galway, the first forecaster to join the unit and the originator of the well-known atmospheric stability parameter, the "lifted index," remained with the SWU after 1955.

Although the first few severe weather forecasts were issued directly to the public via teletype, tornado forecasts through the remainder of 1952 were released by the affected Weather Bureau district offices --- usually after consultation with the SWU. Consecutively-numbered "Severe Weather Bulletins," the forerunner of today's "watches," were initiated in May 1952. As is the case today, the objective was to keep the threat areas as small as possible, with only as much lead time as believed necessary to allow for adequate public response. These early "watches" were not necessarily parallelograms; some were odd-shaped trapezoids or even circles.

The Severe Weather Unit evolved rapidly in 1953 --- a year that coincidentally produced an unusually large number of tornadoes. In January, an experimental program to issue daily outlooks of the severe weather potential of the upcoming day was initiated. These trial forecasts, called "Severe Weather Discussions," were intended as guidance for selected Weather Bureau district offices for the noon - midnight (CST) time period. They became operational in February and were renamed "Convective Outlooks" when regular transmission began on the "Service A" teletype network in April 1955.

The unit was renamed the Severe Local Storm Warning Center (SELS) on 17 June 1953 --- shortly after death-dealing tornadoes struck Flint, MI, Waco, TX, and Worcester, MA. Devastating storms on 7-9 June alone claimed more than 200 lives. These events tested the endurance of the Center's relatively inexperienced staff. Although the storms on 7- 8 June were well forecast, the Worcester tornado on the 9th caught forecasters by surprise; one forecaster requested (and was granted) a transfer out of the unit. By the end of the year, SELS supervisor Kenneth M. Barnett also had requested a transfer as the group came under increasing scrutiny regarding both the size and accuracy of its forecasts. Because of pressure to issue smaller "bulletins," tornado forecast areas decreased in size from nearly 38,000 square miles in 1952 to 27,000 square miles during the first half of 1953. (By comparison, average tornado and severe thunderstorm watches today cover about 25,000 square miles).

The Move to Kansas City

SELS continued to change in 1954. Staffing increased to include a supervisor, 7 forecasters, 6 chartists, a research forecaster and a research assistant. In addition, a new supervisor, Donald C. House, was selected to replace Barnett. House's enthusiasm for severe weather was immediately apparent: on busy days he often worked the forecast desk. Under his direction, the size of tornado forecasts continued to decrease; average "bulletin" size in 1954 dropped to just 15,000 square miles. House also strived to enhance the scientific integrity of the unit by furthering staff research efforts started under Barnett. A series of contributions by SELS meteorologists Ferdinand Bates, Robert Beebe, James Carr, Donald Foster, Joseph Galway, Bernard Magor, Jean Lee, and others advanced the science of severe weather forecasting beginning in the mid 1950s; many of these studies remain relevant today. House also emphasized the importance of high-level (jet stream) data in forecast preparation, incorporating the recent work of Herbert Riehl. House's emphasis on science, very much supported by Weather Bureau Chief Francis Reichelderfer and Regional Director Clayton Van Thullenar, was to a large extent responsible for the high level of respect that SELS commanded by the late 1950s.

In September 1954, SELS relocated from the WBAN Center in Washington to the Bureau's District Forecast Office on the 9th floor of the Federal Building in downtown Kansas City, Missouri. The move was made, in part, to allay media pressure to locate the office in a region more prone to severe weather. In addition, Kansas City was a major teletype circuit switching center. This allowed for more timely access to nationwide surface observations, and for faster forecast dissemination. But the move also recognized an existing local severe weather operation that had been established in January 1952 when J. R. Lloyd, Meteorologist-In-Charge of the Kansas City office, assembled a small group of forecasters to test the techniques of Fawbush and Miller. Oklahoma at that time was part of the District Office's area of responsibility. Lloyd's effort was the subject of scrutiny as pressure increased to have the Weather Bureau issue tornado forecasts like those of the Air Force. Lloyd intended to use the results of the test group to issue actual forecasts beginning in 1953 or 1954. His efforts were instrumental in hastening the Weather Bureau's decision to issue routine severe weather forecasts in May 1952.

The success of the Air Force and Weather Bureau severe weather programs, in addition to educational efforts that included brochures and presentations on tornado safety, significantly reduced public opposition to tornado forecasts during the mid 1950s. Many in fact praised the forecasts as a means of saving lives. During this period, a typical SELS tornado forecast would read as follows: "...possibility of an isolated tornado along and thirty miles either side of a line from Amarillo, TX to 20 miles north of Gage, OK, from 5:15 to 9:00 PM." Such a forecast would have first been telephoned to the district offices(s) involved. If it were agreed that a public tornado forecast was indeed prudent, the district forecaster would notify the local Weather Bureau offices under his jurisdiction, in addition to the media. If the proposed forecast affected only one district office, that office had final say as to whether or not tornadoes would be mentioned in the public forecast. If, on the other hand, a proposed tornado forecast involved more than one district office, SELS made the final decision. It was not until 1958 that SELS assumed total authority for public tornado and severe thunderstorm forecasts.


Throughout the 1950s and early 1960s, SELS data plotting and analysis were performed by hand. Analysis skills increased significantly with the installation of an IBM 1620 computer in April 1963. The 1620 allowed forecasters to access diagnostic fields of convergence and divergence that were difficult or impossible to manually compute. Automated plotting of surface and upper air observations commenced with the arrival of a CDC 3100 system in November 1965. This computer also was used for data tabulation and research by the District Forecast Office.

In August 1965, Donald House left SELS for a position with the newly-formed Environmental Science Services Administration (predecessor of NOAA) in Washington, and Allen D. Pearson was appointed SELS Director. Early the following year, the entire Weather Bureau Office in Kansas City (including SELS and the District Forecast Office) was renamed the National Severe Storms Forecast Center (NSSFC) to better reflect its national scope. In addition, SELS' tornado and severe thunderstorm forecasts were renamed "watches" to more directly correspond with the suite of products issued by the National Hurricane Center. Shortly thereafter, NSSFC moved to the 17th floor of the new Federal Building at 601 E 12th Street, where it remained until relocating to Norman, Oklahoma in 1997.

A series of computer upgrades significantly enhanced NSSFC's data processing and communication capabilities during the late 1960s and 1970s. But one of the more important developments of the period occurred with the establishment of the Techniques Development Unit (TDU) in April 1976. This group was formed to provide software development and to assist with the evaluation of new forecast techniques. It also provided a link to the severe weather research community. TDU's formation marked the first formal research/development program to be associated with SELS/NSSFC since the National Severe Storms Project, the original research component of SELS, departed Kansas City to become the National Severe Storms Laboratory (NSSL) in Norman in March 1964.

Frederick P. Ostby became the Director of NSSFC in May 1980, shortly after the transfer of Pearson to NWS Central Region Headquarters. Ostby oversaw NSSFC's entry into the age of interactive computing with the arrival of the Centralized Storm Information System (CSIS) in February 1982. This system, developed at the University of Wisconsin, enabled forecasters to overlay objective analyses of conventional surface and upper air data with real-time radar and satellite imagery. Watch areas could be formulated directly on the appropriate radar and satellite displays, and different objective analyses could be simultaneously displayed. Later upgrades allowed the user to "roam" and "zoom" across the entire nation. Mesoscale Discussions, unscheduled products used to describe ongoing convective trends and hazardous weather situations, were instituted in 1986 --- partly in response to the availability of timely analyses on CSIS.

Recent years

As part of a decade-long effort to modernize the nation's weather services around the newly-deployed Doppler radar network, NSSFC was renamed the Storm Prediction Center (SPC) in October 1995, with former TDU chief Joseph T. Schaefer selected to succeed Ostby as Director. The McIDAS-based work stations that had been a mainstay of operations since 1982 gradually were replaced by UNIX-based workstations known as NAWIPS.

Early in 1997 and after more than 40 years of severe weather forecasting in Kansas City, the Center moved to Norman, OK. There, on the site of the former Norman U.S. Naval Air Station (now part of the University of Oklahoma), the SPC re-joined the organization that it had in part given birth to three decades earlier --- the National Severe Storms Laboratory. Shortly thereafter, the Mesoscale Discussion program was expanded to include short-term forecasts of hazardous winter weather and heavy rainfall, and a separate program was instituted to address the meteorological conditions favorable for wild fires in May 2000.

In September 2006 the SPC moved once again --- this time just a few miles south --- to join several other federal, state and Oklahoma University weather organizations in the new National Weather Center (NWC). Located on the University of Oklahoma Research Campus on Jenkins Avenue, the NWC offers opportunities for expanded operations-research collaboration to improve the forecasting and understanding of severe local storms. Improved ensemble forecasts and the development of a nationally-acclaimed operations-research test bed are just two positive effects that already have resulted from the relocation. SPC Science Support Branch (SSB; successor to the TDU) chief Russell S. Schneider was selected to succeed Schaefer as SPC Director in August 2010.

A more detailed history of the Storm Prediction Center is available here.

[1] NCEP also includes the Aviation Weather Center in Kansas City, Missouri, the Climate Prediction, Hydrometeorological Prediction, Ocean Prediction, and Environmental Modeling Centers in Camp Springs, Maryland, the Space Prediction Center in Boulder, Colorado, and the National Hurricane Center in Miami, Florida.

Poster of SPC History

A Time Line of the SELS unit and the SPC

Historical SELS and SPC Images

Commemoration of 50 years of tornado forecasting, 1948-1998

Page last updated: February 13, 2010