by Ian Livingston and Katie Wheatley
(This post was last updated on September 14)
Hurricane Harvey may be best remembered as a lingerer, a punisher. Like a deranged energizer bunny, it just kept going. What seemed like never-ending tornado activity occurred on eight days with Harvey.
The storm spawned dozens of tornadoes, including strong tornadoes which are somewhat less common in tropical cyclones than their mid-latitude cousins. Some places were under assault from tornado warnings for days. Those warnings number well past 300.
Gulf of Mexico systems are notorious tornado producers, but they don’t all amount to a lot. Much of the story has to do with how long a system spins over land. As we all know, hanging around is Harvey’s specialty.
Given the mass flooding, the windblown coastline, and crippled infrastructure, getting a handle on tornado numbers is even more trying than normal. The hardest hit area, NWS Houston, has thus far not released any information on surveys or plans for them. That’s understandable given everything else. For now, we’ll focus on what we can piece together.
Tornado warnings: 327
Tornado watches: 10
Filtered tornado reports logged with NWS: 86
Estimated confirmed tornadoes (so far): 45-50
How it stacks up…
Tornado events are hard to get a handle on quickly. Even under normal circumstances, surveys take time, especially when twisters keep coming for days on end. Add in a vast and in some areas heavily flooded landscape — well… ¯\_(ツ)_/¯
It is certainly true that the number of tornado warnings was prolific, particularly for the Houston region. Records were set by that office, where more warnings were issued than the past few years combined. The overall numbers rival warning numbers from the largest of outbreaks. As comparison, the big tornado outbreak of April 2012 had 161 tornado warnings, but also well over 100 tornadoes confirmed.
Huge numbers of warnings are one thing, but tornado warnings typically include plenty of false alarms, and in general confirmed reports are still well below warning numbers. That said, Houston has now checked in with at least 29, with more likely to come from them.
Using filtered reports as something of a proxy for now, 86 is impressive, and keeps the door open to Harvey ranking among the largest tropical tornado events on record. Roger Edwards of SPC has done excellent work on the subject. Below is his top list for the modern tornado record (check out the full article).
You’ll note that right behind the top event, Ivan (2004), is a storm called Beulah. It hit Texas in 1967, and dropped over 100 tornadoes in the state, mostly just to the southwest of where Harvey has impacted.
While we are still waiting for final tallies, totals are now likely in the 45 to 50 range, or so. If that’s the case, we’re cracking the list above.
Significant tornadoes…
Past studies have found that tropical cyclones tend to produce a smaller percentage of the worst tornadoes compared to systems that originate in the mid-latitude westerly flow that moves across the United States. In tornado-watcher parlance these are significant tornadoes, or those rated F/EF2 and higher.
While we watched an EF2 tear across western Alabama on Thursday Aug. 31, at least one significant tornado had already happened.
Supercell in west/central Alabama has a history of producing tornadoes. pic.twitter.com/lbE3qUy9PH
— U.S. Tornadoes (@USTornadoes) August 31, 2017
Although it may not have come from such classic supercell like the one above, the first confirmed significant tornado of this event ripped through parts of Louisiana on Tuesday the 29th.
Between Lake Charles and Lafayette, a tornado that actually at one point crossed Interstate 10 was rated EF2, with winds of at least 115 mph. Add in one or more tornadoes associated with the long-tracked supercell yesterday, and Harvey looks even more impressive.
Ultimately, at least three EF2s have been confirmed to date.
Some of the tornadoes captured…
https://twitter.com/JackRoyer/status/903405367027269632
Picture from Toriano Blanks of the tornado near Petal earlier this morning. pic.twitter.com/Ui3dcYOB3O
— NWS Jackson MS (@NWSJacksonMS) August 30, 2017
Great video captures #tornado forming & touching down in Crowley. Video credit: Sierra Lejeune. h/t @KPLC7News #LAwx pic.twitter.com/MOVHbeYzKr
— Steve Caparotta, Ph.D. (@SteveWAFB) August 29, 2017
VIRAL VIDEO – Watch as a possible #tornado blows by this police dashcam on Saturday afternoon #Harvey #txwx #WeatherNation pic.twitter.com/cauDCJfWRt
— WeatherNation (@WeatherNation) August 27, 2017
https://twitter.com/kendall_haynes/status/901555866054406146
We will update this post as more and better information is available.
U.S. Tornadoes
Latest posts by U.S. Tornadoes (see all)
- March 21-23, 2022 tornado outbreak - March 21, 2022
- Tornado events of 2020 - February 17, 2020
- The tornado outbreak sequence of May 2019 - May 30, 2019
Thanks for the post. Tragic event I hope we rarely see.
Couple of questions:
1. Other than originating from a hurricane, does tornado genesis differ in any way from springtime supercell convection/ setup?
2. Do the intensity (F scale) of these tornadoes deviate from the mean for all tornadoes? (strong, weaker)
In general, tropical tornadoes are weak. I have not compared to the mean of all tornadoes but I would imagine it is at least a little weaker overall. That said, there are a ton of weak tornadoes every year.. they make up a majority of events.
I’d say tropical tornadoes are somewhat like a typical low cape high shear event, but I would not claim to be an authority on that. Main difference might be that the focus of development is generally lower in the atmosphere so supercells are small/low topped in general. Lots of low-level shear usually. Longer it is over land the more it can act like a typical mid-latitude system with more dry air entrainment etc.
Appreciate your quick response. Your answer leads me to another question. Tornadoes are categorized ex post facto (intensity, shape, storm characteristics), so I am wondering if there are predictive categories for tornado genesis? More specifically, we use a general range of factors (MLCAPE, shear, boundaries, dewpoint, etc..) to get a general probability. Moving towards the formulaic, is there a tighter set of thresholds/ ratios/ combos (per setup/ event type?) per storm that can be applied universally for springtime setups, hurricane spawned cells, waterspouts,etc.. WTF am I asking for? Indulge me. 🙂
Interesting to note that of all of the top 10 except for Beulah (though Andrew could be an exception as well) are from the NEXRAD days. Given the weak nature of these tornadoes, and not always having a clear hook echo, this is not surprising. It’s probably not unreasonable to think there were other prolific tornado producers that we simply didn’t know about.
Yes, I think this is true. Believe author primarily considers it a 1995+ database but Beulah had a better record than many in that time frame.