Your Top 10 Strongest Tornadoes Since 1950

[JoMo]

Ironically, Tim Marshall said the same thing about Joplin in a talk he did with our local AMS group. He said most damage was EF-3 to EF-4 and they had to look quite hard to find any sort of damage that would be characteristic of an EF-5 tornado.

That was due to the multiple-vortex nature of the tornado. The larger envelope of the tornado had winds of EF-3 to EF-4 primarily. The areas that received the EF-5 portions were in intense suction vortices.

The video on Minnesota Ave was south and east of the below picture by about 2-3 blocks but closer to the actual center of the tornado. They probably sustained EF-3 tornado damage at the place the video was shot while just to their NW sustained EF-5, probably due to a suction vortex.

The famous Fastrip video also provides evidence that they were perhaps hit by an intense suction vortex as well.

Just found this link:

http://extremeplanet...nter-of-an-ef5/

[tornadotony]

That was due to the multiple-vortex nature of the tornado. The larger envelope of the tornado had winds of EF-3 to EF-4 primarily. The areas that received the EF-5 portions were in intense suction vortices.

The video on Minnesota Ave was south and east of the below picture by about 2-3 blocks but closer to the actual center of the tornado. They probably sustained EF-3 tornado damage at the place the video was shot while just to their NW sustained EF-5, probably due to a suction vortex.

The famous Fastrip video also provides evidence that they were perhaps hit by an intense suction vortex as well.

Just found this link:

http://extremeplanet...nter-of-an-ef5/

I don't want this to be an unfair question, but do you happen to know by any chance what side of the street those swept away homes are on?

[andyhb]

I don't want this to be an unfair question, but do you happen to know by any chance what side of the street those swept away homes are on?

Going by the Spencer data possibly?

[tornadotony]

Going by the Spencer data possibly?

The Spencer data?

[andyhb]

The Spencer data?

That homes/buildings on the southern side of the circulation of that tornado suffered much more complete destruction than the northern side.

[tornadotony]

That homes/buildings on the southern side of the circulation of that tornado suffered much more complete destruction than the northern side.

No. And that's not "the Spencer data." That's just typical storm-relative dynamics of a circulation. Translation of a vortex must be factored in when trying to determine where maximum damage is going to be. It's the same principle behind why the right side of a hurricane is the most intense.

But no, my question has nothing to do with that concept.

[andyhb]

No. And that's not "the Spencer data." That's just typical storm-relative dynamics of a circulation. Translation of a vortex must be factored in when trying to determine where maximum damage is going to be. It's the same principle behind why the right side of a hurricane is the most intense.

But no, my question has nothing to do with that concept.

Alright then, I was just reading Wurman and Alexander's report on the Spencer event and thought you might be referring to that in some fashion.

[AtticaFanatica]

No. And that's not "the Spencer data." That's just typical storm-relative dynamics of a circulation. Translation of a vortex must be factored in when trying to determine where maximum damage is going to be. It's the same principle behind why the right side of a hurricane is the most intense.

But no, my question has nothing to do with that concept.

Alright then, I was just reading Wurman and Alexander's report on the Spencer event and thought you might be referring to that in some fashion.

To be fair, Wurman and Alexander (2005) is the most detailed comparison of estimated tornado wind speeds and damage in the literature, and they found the effect was larger than many previously thought.

[tornadotony]

To be fair, Wurman and Alexander (2005) is the most detailed comparison of estimated tornado wind speeds and damage in the literature, and they found the effect was larger than many previously thought.

Right, but that has nothing to do with why I asked the original question. Given the debris pattern, it appears as though the mean wind was almost perpendicular to the rows of houses at this locations. Now note that in the lower left corner, that row of homes only sustained what appears to be EF3 level damage, whereas the homes behind them sustained what appears to be EF5 damage. My question regarding the side of the street is in reference to the thought that the more severely damaged homes appear to be downwind of the lesser damaged homes, which would allow for increased localized debris loading and a possible explanation for the more intense damage, since the debris pattern and pattern of damage, if the "EF5," if you will, houses, are on the west side of a north-south street, would tend to argue against purely multiple-vortex effects.

[JoMo]

I don't want this to be an unfair question, but do you happen to know by any chance what side of the street those swept away homes are on?

They were on the west side of the street. Across the street is a soccer field and baseball field. (the high school) The street pictured is Iowa street, around 22nd to 24th.

According to the track map, the center of the tornado crossed just south of 24th and Iowa but went nearly directly over where the video was taken for the Minnesota Ave vid. Would the NE side have been that much stronger?

On that image: East = top right. West = Bottom Left. North = Up. South = Down

I've been trying to find out where this one was located but seeing pieces of houses it's difficult to locate. Either way, appears it was high end EF-4 or low end EF-5 from time to time.

[tornadotony]

They were on the west side of the street. Across the street is a soccer field and baseball field. (the high school) The street pictured is Iowa street, around 22nd to 24th.

According to the track map, the center of the tornado crossed just south of 24th and Iowa but went nearly directly over where the video was taken for the Minnesota Ave vid. Would the NE side have been that much stronger?

On that image: East = top right. West = Bottom Left. North = Up. South = Down

I've been trying to find out where this one was located but seeing pieces of houses it's difficult to locate. Either way, appears it was high end EF-4 or low end EF-5 from time to time.

What would be interesting is to do a comparison between the amount of EF4/EF5 damage on the west side of N-S streets in Joplin vs. on the east side. Theoretically, there is a lot more E-W space between houses across a street than for those that share backyards, and given the vertical motion in a tornado, that may make the difference in the amount of debris loading experienced by structures and, in turn, the severity of damage experienced.

[AtticaFanatica]

Right, but that has nothing to do with why I asked the original question. Given the debris pattern, it appears as though the mean wind was almost perpendicular to the rows of houses at this locations. Now note that in the lower left corner, that row of homes only sustained what appears to be EF3 level damage, whereas the homes behind them sustained what appears to be EF5 damage. My question regarding the side of the street is in reference to the thought that the more severely damaged homes appear to be downwind of the lesser damaged homes, which would allow for increased localized debris loading and a possible explanation for the more intense damage, since the debris pattern and pattern of damage, if the "EF5," if you will, houses, are on the west side of a north-south street, would tend to argue against purely multiple-vortex effects.

Yeah, I was just saying the reference to the Spencer paper was completely justified.

[tornadotony]

Yeah, I was just saying the reference to the Spencer paper was completely justified.

Understandable. Given the train of thought I was on, it didn't click to me that he was making that connection.

[JoMo]

What would be interesting is to do a comparison between the amount of EF4/EF5 damage on the west side of N-S streets in Joplin vs. on the east side. Theoretically, there is a lot more E-W space between houses across a street than for those that share backyards, and given the vertical motion in a tornado, that may make the difference in the amount of debris loading experienced by structures and, in turn, the severity of damage experienced.

The pics were taken 10 days after so it's possible that the houses had been cleaned up and weren't swept clean like that image depicts. My ex-gf did say her flooring (carpet/linoleum) was ripped up. This is a shot that is farther west than the last shot (by one street). It looks like everything is destroyed. There were a lot of fatalities in this neighborhood.

http://xpda.com/jopl...do/P1170192.jpg

What made me bring up the multiple-vortex thing was the odd locations of damage and how a house survived when all the houses around it were destroyed. In this pic, it's the one house left standing that looks to still have shingles. I witnessed this in the EF-4 area over by my house as well. A house in the middle of two other houses that suffered EF-3 to EF-4 damage had very little damage other than the windows blown out and part of the roof peeled up.

http://xpda.com/jopl...do/P1170191.jpg

There's always oddities.

I remember Tim Marshall mentioning these poles in one of his talks. He thought the way the poles bent was weird. Any ideas?

http://brittoglesby..../0/XL/50-XL.jpg

[tornadotony]

The pics were taken 10 days after so it's possible that the houses had been cleaned up and weren't swept clean like that image depicts. My ex-gf did say her flooring (carpet/linoleum) was ripped up. This is a shot that is farther west than the last shot (by one street). It looks like everything is destroyed. There were a lot of fatalities in this neighborhood.

http://xpda.com/jopl...do/P1170192.jpg

What made me bring up the multiple-vortex thing was the odd locations of damage and how a house survived when all the houses around it were destroyed. In this pic, it's the one house left standing that looks to still have shingles.

http://xpda.com/jopl...do/P1170191.jpg

There's always oddities.

I remember Tim Marshall mentioning these poles in one of his talks. He thought the way the poles bent was weird. Any ideas?

http://brittoglesby..../0/XL/50-XL.jpg

Off the top of my head? I would think separate debris impacts associated with a suction vortex. One pole with debris coming from one direction, one the other, and they just so happened to be 180 degrees difference. To be honest though, there are probably a myriad of explanations one could devise.

[ExtremePlanet]

This is my first time posting on here, but I've read americanwx and been on stormtrack for years.

My blog is the one cited a couple times earlier. In reference to the El Reno radar reading (125m/s or 280mph), I assure you that is accurate. Without seeming all "they'll kill me if I tell you how I know!", I'll just say I am 100% certain that was the reading they recorded.

As for the images of the Joplin storm damage, I just posted a new entry that details the street names of the places, including the one replied to above (which is from just east of Joplin High School on Indiana Ave, if I remember correctly).

The aerial damage pictures are ten days old, but the geodetic survey images (taken right after the tornado and before any clean-up) verify the homes were swept away in apparent F5 fashion. I know an empty foundation is not clear evidence of EF5 damage, but with such urban density, I think its safe to say the Joplin tornado caused THE most intense tornado damage ever in an urban area. Worse than Moore, worse than Tuscaloosa, worse than Greensburg. Even worse than Parkersburg.

Of course, this is all just my opinion.

[andyhb]

This is my first time posting on here, but I've read americanwx and been on stormtrack for years.

My blog is the one cited a couple times earlier. In reference to the El Reno radar reading (125m/s or 280mph), I assure you that is accurate. Without seeming all "they'll kill me if I tell you how I know!", I'll just say I am 100% certain that was the reading they recorded.

As for the images of the Joplin storm damage, I just posted a new entry that details the street names of the places, including the one replied to above (which is from just east of Joplin High School on Indiana Ave, if I remember correctly).

The aerial damage pictures are ten days old, but the geodetic survey images (taken right after the tornado and before any clean-up) verify the homes were swept away in apparent F5 fashion. I know an empty foundation is not clear evidence of EF5 damage, but with such urban density, I think its safe to say the Joplin tornado caused THE most intense tornado damage ever in an urban area. Worse than Moore, worse than Tuscaloosa, worse than Greensburg. Even worse than Parkersburg.

Of course, this is all just my opinion.

I'm not sure I would call Parkersburg or Greensburg "urban" areas per say (at least to the degree that Joplin, Tuscaloosa or Moore are)...they are more in the style of Hackleburg, Phil Campbell or Tanner, AL.

[Ensō]

Well, what do we mean by "urban"? Not trying to be pedantic, just not sure what the criteria would be.

[andyhb]

Well, what do we mean by "urban"? Not trying to be pedantic, just not sure what the criteria would be.

I'm basing it more on population density and the size of cities like Joplin and Tuscaloosa compared to Greensburg, Hackleburg or Parkersburg.

Also what in the world is with all these "burg" towns getting hit by EF5 tornadoes? There's more as well...Wheelersburg, OH/Vicksburg, MS/Brandenburg, KY....

[Ensō]

I guess since "---burg" is one of the most common town names in the US the chances are higher than normal, but that's still pretty strange.

By the way, I found this video I had in my bookmarks. It's a discussion of the Joplin survey by Tim Marshall from a symposium in Indiana, figured it'd be relevant to this thread for those who haven't seen it. Not really anything new but, if nothing else, he does a good job of explaining just how terrible most of our building codes are.

[AtticaFanatica]

It's probably best to wait for the peer review process to play out on the El Reno measurements. Highly publicizing a measurement only to find out there's an issue would be pretty embarrassing.

[thewxmann]

This is my first time posting on here, but I've read americanwx and been on stormtrack for years.

My blog is the one cited a couple times earlier. In reference to the El Reno radar reading (125m/s or 280mph), I assure you that is accurate. Without seeming all "they'll kill me if I tell you how I know!", I'll just say I am 100% certain that was the reading they recorded.

As for the images of the Joplin storm damage, I just posted a new entry that details the street names of the places, including the one replied to above (which is from just east of Joplin High School on Indiana Ave, if I remember correctly).

The aerial damage pictures are ten days old, but the geodetic survey images (taken right after the tornado and before any clean-up) verify the homes were swept away in apparent F5 fashion. I know an empty foundation is not clear evidence of EF5 damage, but with such urban density, I think its safe to say the Joplin tornado caused THE most intense tornado damage ever in an urban area. Worse than Moore, worse than Tuscaloosa, worse than Greensburg. Even worse than Parkersburg.

Of course, this is all just my opinion.

Side note, but I read through your entire blog in one sitting when I discovered it, and I learned a lot. Great work on your posts.

[ExtremePlanet]

To me, "urban" is an area with a population density greater than 2,000 per square mile. On second look, forget my inclusion of Parkersburg, but Greensburg still stands. Hackleburg and Phil Campbell would not be included, but Tuscaloosa and Concord would.

I believe the Joplin tornado is very, very underrated. I, too, felt the same way at first. I thought it was just an over-glorified F4 tornado that just happened to enter an extremely populated area. But after getting my hands on several thousand aerial damage shots, my perspective has changed.

And as for the El Reno reading, I think plenty of people (including some people who were on the team) are getting a little annoyed at the reluctance of some to release the figure. Its been over a year now.

I don't take the doppler readings that seriously anyway, as they are not "measurements" but instead just statistical bell curves with standard deviations.

I don't think 280mph is that wild of a figure.

I would just love to have seen an accurate radar velocity for the Smithville tornado as it ripped through town. Of the tens of thousands of damage photographs I have looked at, the Smithville tornado is among the most impressive. I bet winds were near 300mph right above the surface, but thats just my thought.

[andyhb]

I'd like to see wind measurements from all 4 of the EF5 tornadoes (and frankly several of the EF4s) from 4/27, but of course in that area the terrain doesn't support that kind of research like it does in the Plain states.

Perhaps Tony could chime in on this, but I'm wondering if the University of Alabama in Huntsville was developing some sort of radar system similar to what Bluestein/Wurman have used in the Plains on storms such as the Spencer, Bridge Creek, El Reno and Red Rock tornadoes, since Dixie Alley does have such a high concentration of strong/violent tornadoes. I know some of their research has had funding cuts in the past couple of years (I heard Dr. Tim Coleman talking about in the Weatherbrains episode following 4/27).

[tornadotony]

I'd like to see wind measurements from all 4 of the EF5 tornadoes (and frankly several of the EF4s) from 4/27, but of course in that area the terrain doesn't support that kind of research like it does in the Plain states.

Perhaps Tony could chime in on this, but I'm wondering if the University of Alabama in Huntsville was developing some sort of radar system similar to what Bluestein/Wurman have used in the Plains on storms such as the Spencer, Bridge Creek, El Reno and Red Rock tornadoes, since Dixie Alley does have such a high concentration of strong/violent tornadoes. I know some of their research has had funding cuts in the past couple of years (I heard Dr. Tim Coleman talking about in the Weatherbrains episode following 4/27).

We have the MAX radar (Mobile Alabama X-band), which is a 3-cm wavelength, dual-polarization radar we can take out for observational work. The problem we have in northern Alabama is terrain. With all the hills, mountains, and trees, there are only a few places we can actually take the radar to get good data.

More on UAH systems: http://vortex.nsstc.uah.edu/mips/

[AtticaFanatica]

And as for the El Reno reading, I think plenty of people (including some people who were on the team) are getting a little annoyed at the reluctance of some to release the figure. Its been over a year now.

There's two ways to go about this: 1) Scream to the high heavens about the observation with no qc and no outside check of the work or 2) Be careful with the observation, qc everything, let peer review have a say, and then release a scientifically accurate information. The process can be annoying, but it's also the responsible thing to do.

I don't take the doppler readings that seriously anyway, as they are not "measurements" but instead just statistical bell curves with standard deviations.

This is a really bad reason not to take radar measurements seriously. The radial velocity you see is a mean value (first moment) of the power-weighted distribution of radial velocities in a resolution volume. So while it's not one measurement, the value you see is based on a weighted distribution of measurements.

In terms of the bell curve, there are a few ways to estimate the moment value. Assuming you're discussing the autocovariance method, then yes, a Gaussian model is assumed (this not an issue in the spectral method). However, most work done has shown that the biases in estimates of the spectrum first moment resulting from non-Gaussian distributions using the autocovariance method are small to moderate.

Do the radial velocity values have error bars? Definitely (assuming the Gaussian model is accurate, spectrum width gives an idea as to the distribution of the radial velocities in the resolution volume), but the errors are not close to the point where the values shouldn't be taken seriously.

[JoMo]

This is my first time posting on here, but I've read americanwx and been on stormtrack for years.

My blog is the one cited a couple times earlier. In reference to the El Reno radar reading (125m/s or 280mph), I assure you that is accurate. Without seeming all "they'll kill me if I tell you how I know!", I'll just say I am 100% certain that was the reading they recorded.

As for the images of the Joplin storm damage, I just posted a new entry that details the street names of the places, including the one replied to above (which is from just east of Joplin High School on Indiana Ave, if I remember correctly).

The aerial damage pictures are ten days old, but the geodetic survey images (taken right after the tornado and before any clean-up) verify the homes were swept away in apparent F5 fashion. I know an empty foundation is not clear evidence of EF5 damage, but with such urban density, I think its safe to say the Joplin tornado caused THE most intense tornado damage ever in an urban area. Worse than Moore, worse than Tuscaloosa, worse than Greensburg. Even worse than Parkersburg.

Of course, this is all just my opinion.

Enjoy your blog that I found the other day. Out of the new pictures you posted, it's interesting to note that the most damage (swept down to the foundation) occurred on Iowa street (facing the open field to the east) and on Indiana street (facing the open field to the west). The picture of the swept homes on W 26th were also facing a large 5 lane street to the north and a field to their south (St. Johns walking track)

It makes me curious just how much all the debris and friction created by the structures, which were placed close together, slowed the winds down.

[ExtremePlanet]

There's two ways to go about this: 1) Scream to the high heavens about the observation with no qc and no outside check of the work or 2) Be careful with the observation, qc everything, let peer review have a say, and then release a scientifically accurate information. The process can be annoying, but it's also the responsible thing to do.

This is a really bad reason not to take radar measurements seriously. The radial velocity you see is a mean value (first moment) of the power-weighted distribution of radial velocities in a resolution volume. So while it's not one measurement, the value you see is based on a weighted distribution of measurements.

In terms of the bell curve, there are a few ways to estimate the moment value. Assuming you're discussing the autocovariance method, then yes, a Gaussian model is assumed (this not an issue in the spectral method). However, most work done has shown that the biases in estimates of the spectrum first moment resulting from non-Gaussian distributions using the autocovariance method are small to moderate.

Do the radial velocity values have error bars? Definitely (assuming the Gaussian model is accurate, spectrum width gives an idea as to the distribution of the radial velocities in the resolution volume), but the errors are not close to the point where the values shouldn't be taken seriously.

Well, you must agree it is a bit of a stretch to refer to my blog post, 14-months after the fact, as "screaming to the high heavens."

The figure is already known by many. My blog is no more official than the stormtrack pages where it is already mentioned.

I'm a storm fanatic above all else. If people think that means I'm a yellow journalist, or unprofessional, or just an amateur, then what can I do?

[andyhb]

We have the MAX radar (Mobile Alabama X-band), which is a 3-cm wavelength, dual-polarization radar we can take out for observational work. The problem we have in northern Alabama is terrain. With all the hills, mountains, and trees, there are only a few places we can actually take the radar to get good data.

More on UAH systems: http://vortex.nsstc.uah.edu/mips/

Nice, thanks.

[ExtremePlanet]

Enjoy your blog that I found the other day. Out of the new pictures you posted, it's interesting to note that the most damage (swept down to the foundation) occurred on Iowa street (facing the open field to the east) and on Indiana street (facing the open field to the west). The picture of the swept homes on W 26th were also facing a large 5 lane street to the north and a field to their south (St. Johns walking track)

It makes me curious just how much all the debris and friction created by the structures, which were placed close together, slowed the winds down.

I can post some more pictures of the area between Pennsylvania and Iowa Ave. The emptiest foundations were definitely adjacent to either a wide road or a large field. But there was a trail of empty foundations that led to Iowa Ave, it just came at a bit of an angle from the southwest. So in some close up pictures it appears all the homes to the west are less damaged but, in fact, there is some continuity to it. The empty foundations stopped a few blocks east of Joplin High, likely because the tornado weakened a little.