What’s changed since the last (and ONLY) August F5/EF5 tornado?

The Plainfield tornado was an elusive event for meteorologists

August 28, 1990, was an unusually hot and humid day across northern Illinois, with temperatures in the low 90s (over 10° Fahrenheit above normal) and dewpoints in the upper 70s, allowing heat indices to surpass 100°F. An approaching cold front from the north in Wisconsin and evapotranspiration of the mature corn crop across northern Illinois caused moisture pooling. While such heat and humidity alone don’t necessarily guarantee severe storms, a cooling mid-level atmosphere from the approaching frontal system created extreme instability. Convective Available Potential Energy (CAPE) levels exceeded 8,000 Joules per kilogram (J/kg), an unprecedented extreme level, indicating strong potential for severe thunderstorms. Although wind profiles were initially unfavorable for tornadoes, mid-level winds changed direction during the afternoon hours, improving the conditions for tornadic development.

Because of the approaching cold front and the extreme instability, the National Severe Storms Forecast Center (NSSFC) issued a Severe Thunderstorm Watch in effect from 1:30 p.m. until 8:00 p.m. local time. The watch mentioned conditions being favorable for the formation of a derecho that would move through the Chicago metro area during the afternoon and evening hours. Around noon, a lone supercell thunderstorm formed in south-central Wisconsin ahead of the southward-moving cold front, producing a brief tornado near Pecatonica, 15 miles west of Rockford, around 1:42 p.m. before moving southeast. A severe thunderstorm warning was issued at 2:32 p.m., more than 45 minutes after the tornado touched down, and made no mention of any tornadoes. Between 2:45 p.m. and 3:15 p.m., the parent supercell spawned four additional short-lived tornadoes in southwest Kane County as it approached Aurora and struck the Aurora Municipal Airport. Some propeller planes flipped over, but no other significant damage was reported.

Around 3:15 p.m., the main tornado touched down near Oswego in Kendall County, quickly intensifying as it entered Will County. In Wheatland Township, the tornado destroyed most homes in the Wheatland Plains subdivision, injuring several people and killing one. It strengthened to F5 intensity over farmland, scouring the ground bare and tossing vehicles, including a 20-ton truck, killing multiple motorists. As the tornado neared Plainfield, it weakened slightly to high-end F4 strength but still caused catastrophic damage. The Chicago National Weather Service issued another severe thunderstorm warning at 3:23 p.m., again with no mention that a destructive tornado was on the ground and in progress. At 3:28 p.m., the tornado struck Plainfield High School, killing three staff members. Quick action by staff saved many students, with the only surviving hallway becoming a shelter. The tornado then destroyed the school district’s administration building, killing one person, and heavily damaged St. Mary Immaculate Church and school, where three more people died. Fifty-five homes were destroyed in Plainfield, along with other structures. The storm continued southeastward, devastating several subdivisions and killing additional residents. It damaged or destroyed schools and dozens of homes in newly built subdivisions. The tornado moved into Crest Hill at 3:38 p.m., causing F3 damage to the Crest Hill Lakes Apartments, killing eight people and destroying multiple buildings. More homes and infrastructure were damaged before the tornado finally lifted in Joliet around 3:42 p.m., but the parent supercell thunderstorm persisted as it crossed into Indiana before finally dissipating around 45 minutes later. A tornado warning was issued at 3:51 p.m., nine minutes after the tornado had dissipated.

In total, the Plainfield tornado caused $165 million in damage (1990 USD), killed 29 people (including later fatalities from injuries and related causes), and injured 353. The Plainfield tornado became the deadliest tornado to affect the Chicago metro area since the April 21, 1967, Oak Lawn F4 tornado. The storm’s northwest-to-southeast movement, an environment that didn’t appear to support strong-to-violent tornadoes, and the lack of visual confirmation of the tornado all make this tornado one of the most distinctive tornadoes in history.

Event failures led to an expedition of meteorological advancements

After the tornado, the National Weather Service’s Chicago office faced heavy criticism for failing to provide any advance warning of the approaching storm. The National Oceanic and Atmospheric Administration (NOAA) Disaster Survey Report faulted the Chicago forecast office for poor forecasting, inadequate coordination with local storm spotter networks, and insufficient preparedness. At the time, the Chicago office was responsible for weather forecasts for the entire state of Illinois, which left it overwhelmed. No warnings were issued until 2:32 p.m. — nearly an hour after the first tornado was sighted southeast of Rockford — and the first warning was only a severe thunderstorm warning. The second severe thunderstorm warning at 3:23 p.m. failed to mention that a tornado was on the ground or the areas it had affected. A tornado warning was not issued until 3:51 p.m., nine minutes after the tornado had dissipated. Contributing to the warning obstacles that day was that the office’s “add-on” Doppler radar, which was installed in 1974, had been disabled by a lightning strike before the event. Finally, the tornado was part of a high-precipitation supercell and rain wrapped while it was moving through the southwest Chicago metropolitan area. There are no known pictures of the tornado during its trek across Kane, Kendall, and Will counties.

In response to these shortcomings, the NOAA and National Weather Service (NWS) responded with two main action items:

• Reducing the Chicago office’s workload by opening new offices in Romeoville (1993) and Lincoln (1995) and delegating forecast responsibilities to surrounding regional offices
• Expediting the development and dissemination of the next generation of radars called the Weather Surveillance Radar —1988 Doppler (WSR-88D), the primary radar system of the Next-Generation Weather Radar (NEXRAD) network, or what we call Doppler radar

The NEXRAD program started its rollout of Doppler radars in 1991 near Norman, Oklahoma, followed by rollouts in Melbourne, Florida, and Sterling, Virginia. After these three locations, the next several radar installation sites were prioritized by their location throughout the region known as “Tornado Alley.” However, due to the failure experienced during the Plainfield tornado and the growing exposure footprint of Chicago’s suburbs, the one of the earliest radars to be installed was located at the Lewis University Airport in Romeoville, neighboring the new NWS office. Between 1991and 2024, nearly 150 Doppler radar locations have been deployed and operational across the contiguous United States.
 

Exposure issue: Yesterday vs. today

The western and southwestern suburbs of the Chicago metro area have grown significantly since that fateful August day in 1990. Many areas that were once vast cornfields have been converted into extensive strip malls, large movie theaters, and sprawling single- and multi-family subdivisions. The three counties in which the Plainfield tornadic supercell traversed — Kane, Kendall, and Will — have all experienced significant population and housing growth since 1990, as shown in Figure 1. These counties are among the top five largest growth counties in Illinois in terms of population and the top six in Illinois for housing units.
 

Figure 1

In the image shown here, the white shaded region represents the damage path of the tornado, the red lines show the streets that were present in 1990, and the green lines show the streets that have been developed since the tornado that would be affected if the tornado were to occur today.

Overall, the three counties experienced a near-doubling of both population and housing stock over the last 30 years. Taking these elements, as well as inflationary factors, into account and applying them to the estimated loss number of $165 million in 1990, it would be possible that an EF5 tornado with a similar track and intensity — due to changes in exposure concentrations, social inflationary factors, position of the tornado swath, and replacement costs being driven by volatility in material and labor expenditures — could have losses that surpass $2 billion, making a tornado of this nature in this location one of the most expensive tornadoes in history. Considering several historical F5/EF5 tornadoes that have ravaged urban regions — the 2011 Joplin, Missouri, tornado and the 1999 and 2013 Moore, Oklahoma tornadoes all come to mind — all have losses that approach or surpass $2 billion (2024 USD), it’s apparent that a tornado of this intensity occurring in a much more densely populated region than Joplin or Moore would approach or surpass the $2 billion loss level as well.
 

Moody’s is taking on the challenge of modeling these rare tornadoes

Major tornado events such as the 1990 F5 Plainfield tornado can create significant hurdles for local economies and insurers alike. Shortages of labor needed for repairs and rebuilding, limited availability of raw materials, short-term spikes in material costs, and disruptions to business operations can all hamper disaster recovery. For insurers, it is essential to proactively prepare for such disasters by having effective claims-handling processes in place, including having enough adjusters to manage the high volume of claims, plan for the potential scale of claim payouts, and maintain adequate claims-paying capacity so policyholders can quickly access the funds required to restore damaged homes and facilities. This approach not only supports rapid recovery but also promotes rebuilding in ways that improve resilience against the severe convective storm sub-perils of hail, straight-line winds, and tornadoes.

Moody’s will soon launch our new Moody’s RMS™ US Severe Convective Storm HD Model for the insurance industry. This model will include large and long-tracked tornadoes like the Plainfield tornado. The model will not only provide a comprehensive understanding of tornado loss severity but will also model  the other two convective storm sub-perils — namely hail and straight-lined winds, including explicit modeling of derechos — to better reflect the types of events being experienced today.  Due to our comprehensive and high-resolution modeling of these perils, our model will allow for more informed decisions on exposure management and capital requirements for this frequent and dynamic hazard.

For more information about the upcoming model, please contact Steve Drews, Moody’s Model and Product Specialist for Severe Convective Storms and Winter Storms, at Steve.Drews@moodys.com. We look forward to continuing to support your risk management needs with our market-leading solutions.