The Surprise Tornado Outbreak of August 24th

By: Joshua Steiner

On Wednesday, August 24th, 2016, I was walking back from my atmospheric thermodynamics class (ironic isn’t it?) around 4:00 p.m. when I heard the news that a tornado had touched down in Kokomo, Indiana. A few minutes later, two tornado warnings were issued for the northern section of Indianapolis. One of the tornado-warned cells was close to the Butler University campus. For the next thirty minutes, I was tracking the movement and development of these storms across east and north-central Indiana. I was surprised at how quickly these storms were developing and how unexpected these storms even were. Earlier in the morning, the Storm Prediction Center in Norman, Oklahoma had only issued a marginal risk outlook for the Ohio Valley and lower Great Lakes region. A few days before, forecast models were only depicting a weak warm front sitting across the region bringing a few scattered showers and thunderstorms. I knew something was different about this event when I heard the news about the Starbucks that had collapsed in Kokomo from a powerful EF-3 tornado.

Around 4:40, I met a friend of mine at the Panera Bread on High Street in Columbus, OH for dinner. By this time, I wasn’t thinking much of the severe weather situation out west. I reasoned that conditions weren’t altogether ripe for a much bigger outbreak. It was too cloudy and the atmosphere was hardly unstable. Shear wasn’t particularly impressive and neither were other parameters normally associated with the development of tornadoes. After a few minutes of talking, I kept getting notifications from my phone about developing severe weather. I first heard about the tornado warning for Allen County sometime after five o’clock. When I heard the news, I quickly opened my radar app (specifically Radar Scope, which is an amazing app and I would recommend it to anyone– it can be purchased from the app store for about $10) to see what was going on. I saw the rapidly rotating storm that was forming over northeastern Allen County. I had never seen such a strong couplet on a radar image in our area before and I knew something was about to happen. It was just a few minutes later that the National Weather Service announced that there was a confirmed tornado on the ground near Doty Road and Brush College Road north of Woodburn. Even more concerning was the fact that the supercell was very well developed and moving east-northeast. I knew that if the storm even made a slight shift to the right of its projected track, something that is common among strengthening supercells, it would impact the town of Antwerp.

Fortunately, the storm would shift in a northeasterly direction, impacting homes several miles north of town. The town was spared, but, unfortunately, many homes were not. The storm continued to cycle through its development stages and dropped another tornado to the north of Defiance and later Napoleon, where the National Weather Service issued the first tornado emergency in the state of Ohio since 2012. This was not the only long-lasting supercell that impacted northwest Ohio that evening. The same storm cell that had impacted Kokomo eventually began tracking northeast through Adams, Van Wert and Putnam Counties, dropping at least two or more weaker tornadoes. The surprising number of storms that were tornado-warned for this event was unexpected. The number of tornadoes that formed from this outbreak was also incredible, given the time of year. August isn’t generally known for its tornado outbreaks in our area.

What this event showed me and many meteorologists in the area is that our knowledge of severe weather and especially tornadoes is rather limited. Furthermore, our conceptual models of what conditions allow a tornado outbreak to form were severely challenged by this event. The lack of decent sunshine, rain and storms in the morning, and a weakly sheared/weakly unstable atmosphere made many believe, including SPC forecasters, that any severe weather that formed later in the afternoon would be severely limited, if not non-existence. The fact that tornadoes did form and caused damage across many parts of the area showed us that there factors we had missed in our analysis of the event.

The showers and thunderstorms that had impacted parts of northeast Indiana and northwest Ohio earlier in the day had caused the weak warm front to the south of the area to stall out instead of surging northward. Not only this, but the cooling from the rain generated a weak boundary that would allow for the development of localized shear and vertical ‘helicity’, both elements that are necessary for the development of tornadoes from severe thunderstorms. This air-cooled “outflow boundary” (as meteorologists usually call it) would set up an environment more favorable for tornadoes later in the day. The only other issue was the lack of sufficient atmospheric instability. The leftover clouds from the showers and thunderstorms prevented any major heating from occurring. By mid afternoon, most areas were only in the low 80’s and some areas only made it to the upper 70’s. However, the warm-front was lifting northward, bringing a surge of moisture and stronger atmospheric instability. Instability went from being nearly non-existent in the late morning hours to at least a modest level by mid and late afternoon. And that’s all that’s necessary in an environment with low level shear and helicity already in place. As soon as storms were able to fire along the boundary, they quickly became tornadic, rather unexpectedly, leading to the worst tornado outbreak the area has seen in at least several years.

The most important thing that we can learn from this event is that we should never let our guard down. For several years now, we have seen very little severe weather. The last major severe weather outbreak that the area was a part of was in November 2013. Since then, we’ve seen sporadic severe weather, but nothing substantial.

One of the positive things about this event is that it showed how prepared the area was for violent weather. It can be easy to be dismissive of tornado warnings after so many false alarms, but it’s always beneficial to know what to do when an actual tornado comes along and does real damage. It was a miracle that there were few injuries, if any, even from the stronger tornadoes that impacted Allen, Paulding and Defiance Counties, as well as Howard County (Kokomo). The more research and the more effort we put into tracking tornadoes, studying them and warning the public about them, the more we can reduce the likelihood of deaths and injuries from future tornado outbreaks.