On April 20th, 2004; 16 tornadoes touched down in Illinois during the mid-afternoon and evening hours. A total of 29 tornadoes occurred across Iowa, Illinois & Indiana (a few others occurred in Nebraska and Oklahoma). One of these was an F3 tornado that struck the Village of Utica, Illinois. This tornado proved to be deadly, killing 8 people (plus an indirect fatality that is recognized officially by the village). What makes this tornado so interesting was the fact that it occurred on a day and in an environment where tornadoes weren’t even expected. Thunderstorm development itself wasn’t looking so good let alone the chance for tornadoes. In this post, we will look through the forecast, what occurred, the environment that caused the quick change in conditions and what has changed since the tornadoes of that day to help meteorologists better forecast tornado events.
Before we dive in, there are a few things I would like to say. First, microscale changes in the environment happen all the time. Weather models are used to forecast potential severe weather events but they are never a “slam-dunk” as you’ve heard me say on here (well, sometimes we know that it’ll be bad, but most often, it isn’t that clear). Ingredients could come together for a massive severe weather outbreak, but a microscale change may occur on the day of (say, not enough instability) to cause less favorable conditions. If it can happen this way, then the opposite can also occur. You could have barely a chance of thunderstorms that quickly grow bad because of a microscale change. Even in the time since the Utica tornado, we have seen days where there really aren’t chances of bad weather but things change quickly (the Kokomo/Central Indiana Tornado Outbreak of 2016 comes to mind). However, technology has improved big time (plus the way warnings are issued) to help make these forecasts a bit better and the environmental change can be noticed a lot quicker.
April 20th, 2004 is a day I remember clearly. Back then I knew a lot about the weather but did not know near as much as I do now. We had a chance of thunderstorms that night where I lived, but did not expect anything horrible. I had seen that a “SEE TEXT” area was highlighted for parts of Illinois and into the Ohio Valley but figured nothing would occur based on observations and models. I remember hearing that evening of the tornado striking Utica, which shocked me (I hadn’t been paying attention because I didn’t feel the threat was that big that day). This shock turned to sadness when I found out about the deaths the following day. The reason I wanted to write about this tornado event was because I have studied it just based off the quick change in the environment on a day that things didn’t look threatening. In addition, it comes as a heads-up to people that weather can change quickly and it is a good idea to listen and pay attention, no matter what the forecast says. Plus, I wanted to highlight what has changed since the event to help keep something like this from happening again.
To paint the image of that day, we must first look at what was forecast. Initially, on the Day 2 Convective Outlook, the Storm Prediction Center had a Slight Risk for portions of Missouri, Kansas, Oklahoma, and Texas. The General Thunderstorm outline stretched into the Great Lakes Region, including Central Illinois. The evening before the event, SPC updated the Day 2 Outlook to expand the Slight Risk all the way up to Lake Michigan, including the area that would eventually see the worst of the outbreak (See Photo Below). The Slight Risk remained for portions of Kansas and Oklahoma because of a dry line, however it was much different further north. The expansion to include Illinois was because models were indicating a couple of short wave troughs embedded within a broader scale trough, which was anticipated to provide support for convective development (which was anticipated to be widespread) along the advancing warm front. Hail was anticipated to be the biggest threat in this region due to frontal inversion. It was also apparent that the storms themselves could weaken before strengthening again after dark, pushing into portions of Michigan & Wisconsin.
The next morning, however, the Slight Risk had retreated, barely running through the Utica area before disappearing to include only portions of South Central and Eastern Illinois in the next outlook (see below). A “SEE TEXT” was added for the Ohio Valley and also included this area of Illinois (the “SEE TEXT” was the precursor to the present day Marginal Risk). The reason the “SEE TEXT” was introduced was because some of the ingredients for severe weather were there, but only marginal instability was forecast to occur. In fact, if you look at the Tornado Probabilities from the SPC forecast that day, the tornado threat was kept well east of Illinois to only include the aforementioned dryline in the Plains (see below).
While an environment for severe weather was not evident in the early afternoon, a rapid change occurred in the mid-afternoon hours that set the stage for the supercells that evening. Models had indicated that almost zero instability would occur and local observations even matched those models. However, the warmer conditions in southern Illinois raced northward into Central Illinois. The lower atmosphere was already moistened due to rain earlier in the day, which helped build more rapid instability into the area. The instability across all layers continued to build as the evolving warm front surged northward. An evolving and northward moving warm front is a boundary that is known for the potential to produce tornadoes. If you combine the rotating capabilities and shear with the boundary, the rapidly building instability and the low-cloud bases that came along with the boundary as well as the moistened lower atmosphere, the possibility for significant tornadoes increased rapidly (see graphic below).
This is where we will switch to focus specifically on the Utica F-3. This low-topped supercell formed and had already produced two tornados in Stark as well as Marshall and Bureau Counties in Illinois. The storm continued to cycle and produce rotation as it neared Florid in Putnam County, where the tornado touched down at 4:44 PM Central Time. Evidence that the rotation was strengthening was seen just 20 minutes prior. Shortly after the tornado touched down Northeast of Florid, the tornado moved towards Granville, where the public reported a multiple vortex tornado. At this point, the tornado was riding right along the warm front, which kept feeding the storm. This is what led to its long-track, violent nature as it just rode the boundary for the duration of its lifetime. It grew to be a half-mile wide as it hit Granville, destroying 12 buildings and damaging over 70 others (causing $8 million in damages). Five people were injured in Granville, with an elderly woman suffering a heart attack during the storm and eventually airlifted to Peoria.
The tornado moved into LaSalle County just northeast of Standard. It produced F0 damage in wooded areas as it creeped towards Peru, still riding the boundary. It continued Northeast and had narrowed down to a tornado that was estimated to be about 50 yards wide as it increased to F2 intensity hitting a subdivision near East 250th Road. The tornado creeped just south of Peru & LaSalle, crossing the Illinois River twice as it stayed on its course towards Utica. As it originally moved into the Southwest side of the village, the tornado was producing F0 damage, with only tree and minor roof damage. However, as the tornado moved through the village, passing Johnson Street, it rapidly increased to F3 intensity, growing to 250 yards in width before exiting the village. In the heart of town, several homes were severely damaged with garages destroyed and some old brick buildings collapsed. The damage continued to numerous structures in downtown, including the sandstone Milestone Tap tavern, a turn-of-the-century building that collapsed into the basement, killing eight (seven people were rescued from the basement). As the tornado moved Northeast, it overturned a mobile home, crossed the railroad tracks and dissipated rather quickly on a steep bluff just northeast of town. Debris from the town was found on top of the hill. See map below for damage survey results within the village.
While the tornado dissipated, the supercell continued to ride the boundary to the Northeast, eventually dropping an F2 tornado just east of town. This tornado caused no fatalities but did continue to cause damage, even having an F0 satellite tornado develop near it at one point.
As the conditions became apparent, the Storm Prediction Center issued a Tornado Watch for the area at 5:40 PM Central Time (see photo below). This was accompanied by a Mesoscale Discussion pointing out the continued threat for tornadoes. They moved the Slight Risk to include portions of Illinois, Indiana & Michigan and also put out a Mesoscale Discussion at 8:55 PM Central Time including this region highlighting the chance of tornadoes (see photo below). Severe weather continued well into the night and both the National Weather Service and Storm Prediction Center were on top of it.
While the conditions did develop rapidly, the National Weather Service had their eyes on the storms immediately. In fact, Granville, the first town hit, had 32 minutes of lead time before the tornado hit, leading to the low injury count. As the tornado moved towards populous areas of LaSalle County (Peru, LaSalle & Utica), the massive lead time continued and citizens knew that a tornado was occurring. The fatalities, unfortunately, occurred where people were doing what they were supposed to do, seeking shelter. The old structure of the building buckled in the strong tornadic winds, causing the building to collapse inward on those seeking shelter in the basement.
On April 19th, 2005….364 days following the tornado, a memorial service was held, where the Village of Utica officially recognized a ninth victim. Angela Brown, the Village Clerk, had worked 16-hour days following the tornado and was under an immense amount of stress. At the time, she was also pregnant. Her son was stillborn on June 2nd, 2004. The doctors attributed his death to the stress Ms. Brown had endured following the tornado.
So, what has changed since 2004? Could an event like this happen again? Sadly, the answer is yes (and I mention the 2016 Indiana/Ohio Outbreak again). However, with better technology, we hope that large lead team will continue to limit injuries and fatalities. In addition, with upgrades in radar and forecast modeling, microscale changes are becoming easier to see and are taken into account. Forecasts can still only do so much and it’s important to know, and pay attention to, what is going on in real time.
Since the 2004 tornado, the Storm Prediction Center has overhauled their Convective Outlooks to get rid of the “SEE TEXT” outlook (the new outlooks began on October 22nd, 2014) Outlooks are now worded as follows:
Removing the “SEE TEXT” removes any possible questions regarding possible severe weather. Marginal means it’s possible but it isn’t expected to cover a wide area. It also introduced the Enhanced Category, meaning it isn’t as likely as a Moderate Risk but it definitely deserves more than a Slight Risk. The most important thing to note is the fact that severe weather, while not expected, is possible on even General Thunderstorm days. Why? That constant change in the microscale environments.
Some of the comments following the tornadoes in Central Illinois helped the National Weather Service move to Storm & Impact Based Warnings. The older system in place in 2004 was based on county areas. Entire counties were put in tornado warnings instead of the polygons that are utilized nowadays. The NWS Graphic below best illustrates this.
County Based Warnings help cut down on entire counties having to worry about a thunderstorm. As you can see, despite all of LaSalle County being in a tornado warning, only the southern and eastern portions of the county were in the threat area. This helps focus on specific areas that are in imminent danger. Now, tornado sirens may still sound county wide, but you should always check your local TV station or look at radar to see the warning polygons (and they are really good about listing things even down to neighborhood level). If all else fails, seek shelter just to be on the safe side.
Impact Based Warnings also help with perceived “false alarms”. Many tornado warnings that are issued are due to radar indicated rotation. This means there is no confirmation the tornado has touched down yet, but one could occur at any moment. These should be taken seriously and you should seek shelter just to be safe. The moment a TDS (tornado debris signature) is seen on radar or a tornado is reported on the ground, it becomes a confirmed tornado. The Impact Based Warning will let you know that a tornado (potentially large) is occurring as well as expected damage level. It’ll also outline any other possible hazards and will state what areas are in the most immediate danger. If it is bad enough and approaching a populated area, a tornado emergency will be issued. All of this combined saved lives during the November 17th, 2013 outbreak that impacted Central Illinois.
The tornadoes that hit Illinois on that day in 2004 passed within just two miles of Illinois Valley Community College, Kankakee Community College and Joliet Junior College. This specific event lead to the NWS Chicago forecast office to include colleges in warnings if there is a significant threat. Eventually, many other forecast offices followed suit. Now, if there is a major threat, schools may be included (even major events and fairgrounds have begun to be included as well).
These enhancements, certainly, will continue to save lives well into the future. Sadly, however, even with these in place, they probably would not have saved the lives that were lost on that day. Unfortunately, everyone was doing what they were supposed to do. They listened and seeked shelter, but the building collapsed in on them. With that said, if you pay attention to the warnings, listen, and seek shelter, you’re chances of surviving a strong tornado increase ten-fold.
This blog post is dedicated to those that lost their lives on April 20th, 2004 in the basement of the Milestone tavern:
Wayne Ball
Michael Miller, Jr.
Helen Mahnke
Carol Schultheis
Larry Ventrice
Marian Ventrice
Jay Vezain
Beverly Wood
It is also dedicated to Sean Kennedy Brown, Angela Brown’s stillborn son.
Adventure Brett (Forecaster Brett)
Sources (and many of the photos came from these sources):
https://www.weather.gov/media/lot/events/2004Apr20/StormData.pdf
https://www.weather.gov/lot/uticatornado
https://www.spc.noaa.gov/exper/archive/event.php?date=20040420
Here’s a photo of the Severe Weather reports from April 20th, 2004. As you can see, many of the tornadoes (red dots) occurred in that “SEE TEXT” area and were directly influenced by the northward advancing warm front.
Photo included at the top of the post is Courtesy of the National Weather Service Office in Chicago and is of damage in Utica.
Adventure Brett (Forecaster Brett)
Should you be interested in reading my post from yesterday regarding tornado warnings and why TV stations interrupt regular scheduled programming (and why people should be happy they do), make sure you do. And also subscribe by entering your email to the right. You’ll be notified when more content drops!