SKYWARN is the National Weather Service program that trains volunteers to watch severe storms and report what they see to the forecast office issuing the warnings. Its roots are in the tornado disasters of the 1960s, above all the Palm Sunday outbreak of 1965, and amateur radio operators became one of its most visible reporting channels. This piece traces how the program formed and how ham nets came to carry spotter traffic. It also covers where a report goes after it is sent and what a dispatch center or emergency management office needs to know about the path, which today runs through phones, apps and agency radio as much as through ham repeaters.

Palm Sunday 1965 and the tornado decade: what the record supports

On April 11, 1965, Palm Sunday, a tornado outbreak crossed the Midwest and Great Lakes states. Indiana, Ohio and Michigan took the worst of it, and Illinois, Wisconsin and Iowa were also struck. The outbreak is the event most often named as the catalyst for organized spotter networks and for the warning terminology the Weather Bureau adopted afterward, and it remains one of the deadliest tornado outbreaks in United States history.

The death toll is the first place where retellings go wrong, because they present one number as settled. The figure most often cited is 271, which is the number carried in National Weather Service retrospectives on the outbreak, including material published by NWS offices in the affected region. Other published accounts give lower totals, and the figures I have seen run from the mid-250s upward. I have not found a single source that reconciles the difference, so the honest way to state it is as a range with the NWS figure at the top. Anyone who needs a number for a slide should cite the NWS retrospective by name and say that other counts are lower.

The Weather Bureau sent a survey team to study the outbreak and the warning effort around it. NWS retrospectives credit that survey with recommendations that changed how severe weather information was framed for the public. The best known is the split between a tornado watch, meaning conditions favor tornadoes, and a tornado warning, meaning a tornado has been seen or indicated. The same retrospectives connect the survey to a push for better dissemination and more organized networks of people watching the sky. I am attributing these points to the NWS retrospectives because those documents are where the connection is made explicitly, and readers who want the original findings should go to the survey report itself through the NOAA library.

Palm Sunday was not the only deadly tornado of the decade. Tornadoes struck Jackson, Mississippi, in March 1966, Topeka, Kansas, in June 1966, and Oak Lawn, Illinois, in April 1967, among others. I am leaving out the death tolls for those events because I have not checked them against a primary source for this piece. Together they made the case that forecasts alone were not reaching people in time and that forecasters needed eyes on the ground.

Before SKYWARN: Tinker in 1948, the forecast unit, and early radar

Tornado forecasting in the United States is commonly dated to March 25, 1948, when two Air Force weather officers at Tinker Air Force Base in Oklahoma, Ernest Fawbush and Robert Miller, forecast a tornado at the base and one struck that evening. Before then, the Weather Bureau had a long reputation for avoiding the word “tornado” in public forecasts, a reluctance usually explained by fear of causing panic and by doubt that tornadoes could be forecast at all. Accounts differ on how formal that restriction was, so I describe it as a practice and not as a written rule.

In the early 1950s the Weather Bureau set up a national unit for severe local storm forecasting. That unit moved to Kansas City and decades later became the Storm Prediction Center in Norman, Oklahoma. Its job was the large-area outlook and the watch. That left the local question of whether a tornado was on the ground near a given town to the local office, which in the 1950s had very limited means of answering it.

Radar entered the picture in the same decade. A hook-shaped echo associated with a tornado was documented on a research radar in Illinois in 1953. The Weather Bureau began fielding the WSR-57 radar network in the late 1950s, which gave forecasters reflectivity, meaning a picture of precipitation, but no direct measure of wind. A forecaster could see a storm with a suspicious shape on the scope and still have no way to know whether it was producing a tornado unless someone near it said so.

Accounts of NWS history describe informal and local spotter arrangements in the 1940s and 1950s. Some were tied to military installations, others to civil defense organizations and to police and sheriff’s offices that called the Weather Bureau when a deputy saw a funnel. Amateur radio operators were passing weather observations in some areas during this period as well. I have not found a reliable single date when ham operators first fed reports into a Weather Bureau office, and I would distrust any account that offers one. What the 1960s added was a national program with a name, a training approach and an expectation that every office would build a network.

A Weather Bureau program: what can be dated and what cannot

The National Weather Service describes SKYWARN as a program developed in the 1960s. Accounts outside the NWS put its formal start anywhere from the mid-1960s to the early 1970s, which reflects the fact that the program grew office by office. It did not start with a single announcement. During that span the agency also changed names: the Weather Bureau became the National Weather Service in 1970, when it was placed under the newly formed National Oceanic and Atmospheric Administration. A sentence saying “the National Weather Service created SKYWARN in 1965” is therefore wrong on the agency name even before the date is argued.

The official styling is SKYWARN in capital letters, and the NWS treats the name as a registered mark. Many people write it as SkyWarn or Skywarn, and that is harmless in conversation. In a document that cites the program, though, the official form signals that the writer is describing the NWS program and not a local club.

The design has stayed recognizable since the beginning. Each local forecast office, now called a Weather Forecast Office, recruits volunteers in its county warning area, trains them, and tells them what to report and how. The Warning Coordination Meteorologist at each office is usually the person who runs spotter training and outreach, and that role is the one emergency managers and dispatch supervisors should know by name. Because each office runs its own network, practices differ from one office to the next, including reporting thresholds, whether spotters receive ID numbers, which phone numbers they use, and how closely the office works with local ham groups.

That local variation is why national generalizations about SKYWARN need care. A statement about how spotters report in one office’s area may not describe the office next door, and the office itself is the only reliable source for its current procedure.

Beam height and ground truth: why radar never replaced the spotter

The WSR-88D Doppler radar network, deployed nationally in the 1990s, measures motion toward and away from the radar, so forecasters can see rotation inside a storm. A dual-polarization upgrade completed across the network in 2013 added the ability to detect debris lofted by a tornado in many cases. Each of these advances was described at the time as a step toward warnings based on radar alone, and each left the spotter in place for the same physical reason.

A radar beam travels in a nearly straight line while the earth curves away underneath it. The lowest beam also leaves the antenna at a small upward angle. The farther a storm is from the radar, the higher above the ground the beam passes, and at the distances many counties sit from their nearest radar, the lowest scan is thousands of feet up. Rotation detected at that height suggests that a tornado may be forming or present. It cannot show what is happening at the surface, and the surface is where the damage happens and where the warning has to reach people.

The debris signature has its own built-in limit, because debris in the air means something has already been destroyed. A spotter who sees a funnel touch down, or who reports that a rotating wall cloud is lowering, can give the forecaster information earlier than that and at ground level. Spotters also report things radar estimates poorly or not at all, such as measured hail size, flooding over a specific road, wind damage on a named street, and the difference between rotation that is producing a tornado and rotation that is not.

Spotters have limits too, and NWS training is frank about them. Tornadoes wrapped in heavy rain may not be visible at all, night removes most visual cues, and low ragged clouds called scud are frequently mistaken for funnels by untrained or excited observers. The forecaster weighs a spotter report against radar data, and the two together are stronger than either alone. A report from a trained spotter at a known location carries weight with the warning forecaster because the training is meant to make that report reliable.

Repeaters and net control: how ham radio became a reporting layer

Amateur radio operators fit the spotter role well because of what they already had. A ham with a mobile radio and access to a wide-area repeater can report from a county road with no cell service. The repeater system does not depend on the telephone network, which can be congested or damaged during exactly the storms that matter. Hams also come with some experience of net procedure, where one station controls the frequency and others check in and wait to be called. That habit makes it possible to collect reports from dozens of spotters on one channel without chaos.

The typical arrangement is a SKYWARN net activated on a designated repeater when severe weather threatens, often after the forecast office or an emergency manager requests it. A net control station takes check-ins, logs spotter locations, receives reports, and decides which reports meet the office’s criteria and need to be passed on. The relay to the forecast office happens by telephone, by an online or chat method the office provides, or through an amateur station at the forecast office itself, since many offices host a ham station that volunteers staff during severe weather. The net control operator is therefore acting as a filter, and a good one keeps duplicate and secondhand reports from reaching the forecaster.

The relationship between amateur radio and the NWS has a public face in SKYWARN Recognition Day, developed in 1999 by the National Weather Service and the American Radio Relay League. It is held in early December, when ham operators contact forecast office stations to recognize the volunteer partnership. The event is useful to know about because it shows which offices maintain active amateur stations.

The ham layer also has weaknesses that anyone relying on it should understand. A net with poor discipline turns into a stream of chatter, with spotters reporting rain or calling in what they heard on a scanner, and a net control operator who passes all of it along wastes the forecaster’s attention during the minutes that count. A ham license is required to transmit on the amateur bands, but holding one does not make anyone a trained spotter. The NWS training is separate, and a licensed operator who has not taken it is an untrained observer with a radio.

Not a ham program: who the spotters actually are

A common misconception holds that SKYWARN is an amateur radio program, and that is wrong. SKYWARN is a National Weather Service program, it has never required a radio license, and most reports in many areas arrive by other routes. The NWS SKYWARN page describes the volunteers as including police officers, firefighters, dispatchers, EMS workers, public utility workers and other private citizens. It also encourages people responsible for others, such as staff at hospitals, schools, churches and nursing homes, to train. Hams are one group within that population, often a very active one, but they are not the program.

Reports reach forecast offices through many channels. Trained spotters call reporting numbers that many offices give out in class, and the public and spotters both use online reporting forms on NWS office websites. NOAA’s National Severe Storms Laboratory runs the mPING app, which lets anyone submit precipitation and some severe weather reports from a phone. Many offices watch social media during events. Spotter Network, an independently run system, lets trained spotters share their positions and submit reports that forecasters can view. Law enforcement and fire units in the field report through their dispatch centers, which relay to the office. Emergency managers report directly through partner channels the office maintains.

The NWS SKYWARN page has long given an estimate of 350,000 to 400,000 trained spotters nationwide. That is the agency’s own estimate, and I have not seen the NWS publish a breakdown showing how many of them hold amateur licenses, so any claim about what share of spotters are hams should be treated as unsupported unless someone can point to a source. Check the current NWS page before quoting the figure, since it may be revised.

The ham-only impression has understandable sources. Many local groups use the SKYWARN name for their amateur nets and associations, the nets are easy to hear on a scanner, and the ARRL partnership is well publicized. In some counties the ham net really is the main organized channel. That is a fact about a particular county, and it does not describe the national program.

What, where and when: the report and where it goes next

NWS spotter training teaches a short report built around what was seen, where it was, and when it was seen. The location should be specific enough to plot, such as an intersection, a mile marker, or a distance and direction from a named town, and it should describe the phenomenon’s location, which is not the same as the spotter’s own position. The time should be the time of the observation. A report relayed through a net control station or a dispatcher can lose ten or fifteen minutes on the way, and if the original time is lost, the forecaster may place a moving storm in the wrong spot.

The terms also have to be precise. A tornado is a rotating column in contact with the ground, a funnel cloud is rotation that has not reached the ground, and a wall cloud is a lowered cloud base that may or may not rotate, and calling a wall cloud a tornado over the air produces a false report. Hail should be measured or compared against a known object such as a coin or a ruler. NWS severe thunderstorm criteria use hail of one inch in diameter or larger and wind of 58 miles per hour or greater, so the difference between “pea-sized” and “quarter-sized” matters to the forecaster. Wind speeds should be called estimated unless an instrument measured them. Individual offices set their own reporting thresholds above and below those lines, so spotters should use their local office’s list.

At the forecast office, the report goes to the forecaster working warnings. It can lead to a new warning, an update to an existing one, or a change in wording, and NWS warning text often states the source of the information, such as radar indication or confirmation by spotters. Reports also go into Local Storm Report products the office issues, and later into the Storm Data publication, which serves as the official record of severe weather events and is used to verify warnings after the fact. A false report therefore does two kinds of harm: it can distort a live warning, and it can end up in the historical record and in verification statistics.

Dispatch, fire and EMA: where agencies sit in the spotter path

A dispatch center is part of the spotter network whether or not anyone planned for it. During a warning, the public calls 911 and non-emergency lines to report funnels, hail and damage. Deputies and fire crews already on the road see storms before anyone else in the county. All of that information arrives at the dispatch floor, and someone there has to decide which of it goes to the forecast office, how it gets there, and who makes the call while the phones are also carrying actual emergencies.

In the centers I have worked with, the gap is usually procedure, not willingness. Dispatchers want to pass on a tornado report, but the severe weather SOP may not say who relays it, which number to use, or what information the forecaster needs. A call-taker under load will then pass along “somebody saw a tornado out by the river,” with no time or usable location and no indication of whether the caller was trained. A short written procedure fixes most of this, and the Warning Coordination Meteorologist at the forecast office will help write it, because the office wants better reports.

The radio side needs a decision too. Fire and law units reporting weather on a primary dispatch talkgroup compete with emergency traffic during the busiest minutes of the event. Some agencies send weather observations to a secondary talkgroup, and others have units report by phone or through a supervisor. The right choice depends on the system and the staffing, but it should be made before the storm and written into the communications plan rather than improvised. An emergency operations center with an amateur station should know whether that station can monitor the SKYWARN net and whether it runs on generator power, since the ham net is most valuable when the phone network is failing.

The NWS StormReady recognition program gives emergency managers a structured way to work through these questions with their forecast office. Its guidelines cover how a community receives warnings, how it passes them on, and its connection to spotter training. The specific criteria have been revised over time, so verify current requirements with the Warning Coordination Meteorologist at your forecast office instead of relying on an older checklist.

What to do at your agency

• The emergency management director calls the Warning Coordination Meteorologist at the forecast office that serves the county, gets the date and location of the next spotter training session, and sends it to every fire, EMS and law enforcement training officer in the county before the end of the month.
• The dispatch supervisor writes one paragraph into the existing severe weather SOP stating who relays a tornado, hail or damage report to the forecast office, the number or method used, and the four items required (what, exact location, observation time, and whether the reporter is a trained spotter), and has the Warning Coordination Meteorologist review it.
• The emergency management director asks the forecast office for its current local storm report criteria, including its hail and wind thresholds, and posts a one-page copy at each dispatch position.
• The radio system administrator confirms the frequency and tone of the county SKYWARN net repeater, verifies that the EOC amateur station can monitor it, and runs the station on generator power for one hour, recording the result in the EOC equipment log.
• The fire department training officer adds a fifteen-minute segment on the spotter report format, with hail measured against a coin or ruler and tornado, funnel and wall cloud defined, to an in-service session already on the schedule.
• The communications manager puts one item on the agenda of the next regular communications or operations meeting: which talkgroup or method field units use for weather observations during a warning, and the decision is recorded in the meeting minutes and the radio communications plan.

Takeaways

• SKYWARN is a National Weather Service program that the NWS describes as developed in the 1960s, and it took shape in the period following the April 11, 1965, Palm Sunday tornado outbreak.
• The Palm Sunday death toll is most often given as 271, the figure in NWS retrospectives, while other published accounts give lower totals, so it should be stated as a range with sources named.
• NWS retrospectives credit the Palm Sunday survey with recommendations that led to the distinction between a tornado watch and a tornado warning and to more organized spotter networks.
• Doppler and dual-polarization radar did not make spotters obsolete, because the beam passes high above the ground at distance and cannot confirm what is happening at the surface.
• Amateur radio operators became a major reporting channel through SKYWARN nets on wide-area repeaters, with net control stations filtering reports and relaying them to forecast offices, many of which host ham stations.
• SKYWARN is not exclusively an amateur radio program, since no license is required and spotters include dispatchers, firefighters, police, EMS, utility workers and private citizens who report by phone, online forms and apps.
• A useful spotter report gives what was seen, where the phenomenon was, and the time it was observed, and relays through dispatch or net control must preserve that original time.
• The Warning Coordination Meteorologist at the local forecast office is the person to ask about training, reporting procedures and current StormReady requirements.