What’s Going Wrong?

This year has had a truly terrible human toll from tornadoes. The current fatality count, approximately 500, is an order of magnitude larger than the average for an entire year. It is to be hoped that this number does not substantially rise, but we do have June, the tropical season (which is expected—and I agree—to be quite active and have a higher than average risk of U. S. hits), and the cool-season secondary severe weather peak. However, this figure is horrible enough even if these periods of higher risk produce absolutely no damaging or fatal tornadoes. Something went very wrong this year, something that has not gone this wrong in almost 60 years, and it is important to determine what it was.

Having followed the stories in a fair amount of depth and from multiple sources, I have developed some suspicions of my own about what some of the problems were. These are problems that either appeared in multiple situations or that appeared in sites where high concentrations of deaths occurred. Other people may form their own opinions, but in whatever analysis of 2011 that takes place (and you can just about guarantee that something of the kind will be done), I hope that the following issues are seriously examined.

1. Safety recommendations for urban residents.
One ugly lesson we have had forced on us this year is that, contrary to long-standing myth about cities being safer than rural areas in tornadoes (perhaps because of the idea that buildings will provide resistance?), a city may well be the worst place one can be in a violent tornado. Even in the age of high-resolution Doppler radar, real-time chaser and spotter reports, live coverage, and 20-minute lead times for warnings, we now know that an EF-5 tornado striking an urban center can result in a triple-digit death toll, as happened in Joplin, MO. It is easy, in retrospect, to understand why a densely packed urban area may be the worst possible place to be. Other than high-rise office buildings, there is no safe place to be. High-rises, according to the EF-scale, will not be demolished even in an EF-5; the maximum expected damage is “permanent structural deformation.” However, directing everyone to the nearest tall office building is a ridiculous “safety” recommendation, needless to say.

What are some other problems with urban areas? Many buildings in the central business district, like fast food restaurants and small businesses, are not constructed to withstand anything like a tornado, and they are simply not safe places to be. Big box retailers will contain very heavy stock that is piled high, providing plenty of potentially deadly missiles. Vehicles are everywhere, and they will become airborne. Designated tornado shelters, which some communities do have, would be useful only if people flocked to these sites well in advance of an actual tornado, because congestion on the roads could result in mass fatalities. Storm cellars would be all but nonexistent, and basements are limited in spatial extent and would be directly beneath the houses, which puts anyone taking shelter therein at risk of exposure to tornadic winds and suction if the house is removed. There is no easy way to get out of danger; traffic congestion will occur if people try to evacuate en masse, putting people in possibly even greater danger than they would have been if they had stayed put. And, of course, cities will have more debris than any other type of community.

We need to seriously consider what kind of safety recommendations can reasonably be given to people who live in town—if any. It is highly uncommon for cities to be struck by EF-4 and EF-5 tornadoes, but it can happen, and the buildings do not provide friction-based wind resistance that would mitigate the effect of violent winds. The situation needs to be looked at to see if any sort of useful specialized safety advice can be given to urbanites.

2. Vulnerability of cell phone networks.
Sometimes when tornado victims are quoted in the news as saying “there was no warning,” what they mean is that they, personally, were not aware of the situation. However, every fatal tornado this year except for one occurred within a tornado watch (link), and that one, an EF-3, occurred within a severe thunderstorm watch. I do not want to sound callous, but there is a responsibility to be weather-aware, which includes awareness of tornado watches and warnings. The outbreaks were all extremely well forecast. Most warnings this year had lead times of 20 minutes or so. In these situations, it is simply inaccurate to say that there was “no warning.” It is passive-aggressively blaming the Weather Service for one’s own failure to be aware.

However, in some cases, there was a legitimate lack of warning, though this is not the failure of the Weather Service. These instances involved the failure of the power grid and the cell phone network, taking down any means by which one might receive weather warnings other than a battery-powered or hand-crank radio. This occurred in some of the small towns that were overrun by the extremely violent EF-5 tornado that traversed northwest Alabama on April 27. I think it also occurred in one of the EF-4 tornadoes of the same day. This is a real problem. Most new phones have the ability to function as handheld PDA, music player, portable gaming device, organizer, Rolodex, e-mail, Twitter, Facebook, and even web browser. With this kind of capability—and none of it dependent on a steady AC/DC power supply—it is easy to understand why people would be reluctant to buy a portable radio. They depend on their phones, and with good reason. However, the cell phone network is clearly vulnerable. If there is significant damage to a tower, down the network goes. This goes for other disasters, including terrorism. The problem needs to be examined to see if these towers can be made more robust. Severe weather outbreaks usually have more than one round, and areas that saw their power and communications knocked out by one (even relatively benign) event can be extremely vulnerable later on.

3. The DTV conversion.
I had a bad feeling about this as soon as the DTV changeover began back in 2009. With the old analog signals, one could have a fuzzy screen—even no visuals at all—and still have audio. With the digital signal, the broadcast becomes choppy, and before long it goes completely black and silent. You have either a near-perfect picture or you have nothing. I have heard more than one anecdote of people in tornado-struck areas who did not lose power until after the tornado hit, but who could not hear the warnings because their TV signal had gone out. I don’t know how many fatalities, if any, were caused by this, but it is a problem that, in my opinion, was severely underrated when the rush to DTV was taking place.

4. Dangerous amateur videography.
There are a lot of videos out there of this year’s tornadoes, and a great many of them were not taken by experienced storm chasers or always-running security cameras. They are also not all taken by people who were a safe distance away from the tornado. One tornado video from Alabama was shot by someone who had a car accident while taking the video! This is a major problem. There is not one thing that can legally be done to stop people from taking video of approaching tornadoes if that is what they want to do, but it is a sad reflection on our society. I don’t blame storm chasers for this. Responsible storm chasers and weather spotters have provided a lot of on-site reports, helping newscasters and people following the situation online know when there is actually a confirmed tornado. Real-time reports of a tornado on the ground helped me decide to evacuate in advance of an EF-5 tornado. Some storm chasers behave highly irresponsibly on the road, and they should be condemned by the rest of the community for it, but overall these people take their hobby very seriously. The videos I speak of are taken by people who just happened to see the tornado and decided it would be a cool idea to get video of it, and the videographers clearly have no knowledge of how far away they should be or where to go if the tornado shifts its path. This particular trend is not the fault of storm chasers; it was produced by social media culture.

5. Lack of shelter from violent tornadoes.
The overwhelming majority of this year’s fatalities have occurred in violent tornadoes, those rated EF-4 and EF-5. This is because these tornadoes will utterly demolish well-built houses, leaving only a pile of debris over a foundation (EF-4) or a bare foundation altogether (EF-5). The Hackleburg, AL EF-5 tornado even buckled the concrete slab foundation of one structure, and the Neshoba County, MS EF-5 (“my” tornado) dug up dirt two feet deep. I think that the odds of survival in these tornadoes are still better than 50-50, but it is easy to see how this kind of situation is incredibly dangerous. The unfortunate fact is that a majority of houses in the South and Midwest do not have basements or storm cellars. My position is that storm cellars are preferable to basements, especially if they have a “fallout shelter” design in which the entrance is not directly above the main room, but is horizontally removed from it. I have read enough accounts of people who took shelter in their basements and were sucked out that (while I agree that basements are clearly preferable to any above-ground shelter) I cannot equate basements and storm cellars. I would also recommend firmly anchored handrails in the main room, in case the door was torn away. The opening sequence of Twister is not myth.

I definitely do not equate above-ground saferooms with underground shelter. These structures are engineered, yes, but they are highly vulnerable on two counts. One, if they are undermined from below, they will roll. Two, the engineering is based on a typical flying missile the size of a 2×4 and a typical flying missile speed of 100 mph. EF-5 tornadoes have wind speeds upwards of 200 mph and have even been clocked as high as 300 mph, though it is a matter of debate whether a large object would travel at these extreme speeds. However, these large objects do travel. In Smithville, MS, the town water tower was dented 120 feet above the ground by a car that became airborne. This is known for a fact by matching paint from the tower and the car. There is video on the Internet of a Canadian F5 tornado in which a whole house is clearly airborne at a great height before it disintegrates.

It’s easy to say, of course, that everyone should have an underground shelter. It is quite another to bring that about. I am opposed on principle to any government mandate to protect people from themselves if there is no risk to other people. This must be a matter of personal responsibility. However, I am in favor of rewarding the decision to install a storm shelter with a tax rebate or credit.  Such credits have been offered in the past, usually to specific regions after particularly high-profile and destructive weather events; I argue that they should be permanent and universal.

These are my suggestions about what should be looked into when the year 2011 becomes part of history, or when people begin to examine what has gone wrong with severe weather preparedness, whichever comes first. Undoubtedly other people will focus on other things. One thing is for certain: We need to know whether this year’s atrocious human toll was in any way preventable, because if it was, it must not be allowed to happen again.

Running from an EF-5: Part Three

I hope everyone had a good Mother’s Day, or just a good Sunday. This is Part Three of the series about the EF-5 tornado that went through central Mississippi on April 27, 2011. Part One and Part Two are linked.

Days later, after the terrible human toll becomes widely known, the scientific damage surveys come in.  I’d been keeping an eye on them, not really expecting this storm to be high on the priority list because there were fatalities in other areas such as Smithville (it is not known at the time that “my” storm was the same tornado that killed 3 women in Kemper County), but knowing that eventually it would be examined.  One day, the piece of news I have been interested in for personal reasons comes in; the tornado has been rated EF-4 with 180 mph winds.  It began around Philadelphia and lifted outside of Mashulaville.  The Jackson office of the Weather Service helpfully put a Google Earth file on their website that showed the tracks of the tornado and the supercell that spawned it; I download this file and look at it.  The mesocyclone—the strongly rotating column within the thunderstorm—did indeed pass directly over my house.  The tornado itself would have continued straight into downtown Macon if it had stayed on the ground, most likely missing my house, though barely.  This fits with what I heard the day it happened, but now I know just how bad it really could have been for the town.

Like, I suspect, a great many people in the South, I have graphic nightmares about the outbreak for several days.  When some comparatively mild thunderstorms come through a few days later, my nerves treacherously ignore what my meteorology-educated mind is saying, that there is nothing to worry about except lightning.  I dare not suggest that I have post-traumatic stress disorder when I did not actually experience a life-threatening trauma, or when so many who experienced the tornadoes directly undoubtedly do now suffer with this condition, but everything must have degrees, and I am clearly experiencing some degree of being traumatized.  It’s hard not to experience something like this when you leave home with the completely justified expectation that you won’t be coming back except to ruins, even when that turns out not to be the case.  In addition, there is the knowledge of what might have been, with a significant helping of meteorological education and a vivid reading-influenced imagination thrown in for good measure.  There is knowing that the tornado I had run from was every bit as bad as some of the worst beasts of the outbreak—except that it did not last long enough to make a direct hit on a closely populated site.

And then a week after the outbreak, the tornado’s rating is changed to EF-5, the highest on the scale.  This is the kind of tornado that leaves slab foundations swept clean, the kind that reduces every smallish building in its path to rubble, the kind that obliterates small towns.  In the case of this specific tornado, it’s the kind that, by the force of the wind and probably some microscale debris, pulls up blades of grass and digs out sections of the ground two feet deep.  I recall reading a comment by some meteorologist, I have no idea whom, to the effect that he would not believe an F6 tornado could exist (this was in the days when the old Fujita Scale was used) until he saw coffins pulled out of the ground.  Well, there is no such thing as an EF-6, and “F6” was never put into practice because there were no official damage criteria for it, but four more feet and this one would have been capable of that one man’s stated standard.  And yet I think the Hackleburg/rural Alabama EF-5 was still more violent.  The damage survey for that one is responsible for one of my nightmares.

Three women in Kemper County, MS lost their lives in this tornado that I ran from.  They lived in a mobile home.  There were surely others in the South who tried to take shelter in these structures and did not survive.  Trailers are not safe!  Granted, little will stand up to an EF-4 or EF-5, but a trailer won’t even stand up to an EF-2.  And there are a lot more of those than the 4s and 5s.  The Weather Service guideline of leaving a trailer is spot-on.  And even a constructed house isn’t necessarily safe, though the type of tornado that would create uncertainty about survival in these structures is mercifully rare.  However, such tornadoes do happen.  They happened that day in April.  I wish that more people and communities in the South had storm shelters—and underground ones.  Above-ground concrete bunkers may be all well and good, but houses in the Hackleburg and Phil Campbell area had their concrete block foundations destroyed by the EF-5 that went through there.  Also, something capable of digging up dirt two feet deep is quite possibly capable of undermining a slab foundation by the same process and ripping it from the ground by an extreme wind-tunnel effect under the now hollow space.  (That would be beyond anything I have ever read about, but the possibility has been theorized, and this is how it would happen.)  Anyone who can afford it should build a storm cellar—and it bothers me that more people in this region cannot afford it.  There should be a tax credit for it.

I have never left the house before over a tornado warning, or even a suspected tornado.  This was a decision that I made based on the information that was available to me at the time:  my knowledge of the off-the-charts atmospheric parameters that supported violent tornado formation, my experience driving in supercells, the extremely threatening hooked radar signature, the probable debris ball that is usually seen only in intense tornadoes, the path that would have taken it almost directly over my house, and the report from chasers and spotters of a confirmed large tornado with damage and debris.  I decided that the probability of this being a tornado that my house could not stand up to was unacceptably high.  It turned out that the tornado that I fled from was an EF-5, which seemingly justifies the action, and yet I can’t endorse the choice I made as a general public rule.  It happened to be a good decision based on the fact that I had time and I knew what direction to drive, but in general it is a risky decision, and risky decisions should be made only if there is extensive knowledge to support them.  Blind, panicked “I have to get out of here and it doesn’t matter where I go” driving is not something we need.  Anyone who doubts this should take a good look at some of the damage pictures that involve vehicles.  As a matter of fact, extreme vehicular damage was one of the criteria that the Jackson NWS office used to upgrade “my” tornado to EF-5.

Smithville, MS.  Hackleburg, Phil Campbell, Rainsville, Oak Ridge, and so many other small towns in Alabama.  Tuscaloosa and Pleasant Grove, AL.  Ringgold, GA.  And almost, but for the grace of God, Macon, MS.  The nonchalance that at least some people seemingly had comes back to my mind.  I hope it was the exception.  I hope that, after seeing what happened to their neighbors in small towns just like Macon, they are reflecting on their own close call.  Do they know what a close call they had?  Do they realize just how out-of-the-ordinary the tornado that was barreling straight for them truly was?  Do they realize that, if the tornado had not lifted, there would probably be another small town on the dreadful list of “leveled by an EF-5 tornado”?  And yet, there are so many uncertainties.  Would the tornado have maintained that strength if it had stayed on the ground?  Sometimes they don’t.  Or would it possibly have strengthened even more, as the horrific rural Alabama EF-5 apparently did as it tracked north?  No one can know.  But we can make sure that, if something like this should happen again in our lifetimes, we have a plan of action.

Running from an EF-5: Part Two

This is Part Two of a series of three about my experience with the EF-5 tornado that went through four counties in Mississippi on April 27, 2011. Part One can be found here, and Part Three, an account of the aftermath, will be forthcoming.

Wednesday, April 27, 2011.

Bzzzzzzz! My cell phone is still in vibrate mode.  I’ve forgotten to turn the ringtone on.  I pull my eyes away from the TV and answer it.  It is my father, who is at work.  “Are you—”

“I know,” I say.  “I’m going to take cover.”

“I’ll call after it passes,” he says.  His voice is clearly nervous.  We hang up.

I take another look at the radar that the weatherman is talking about.  That sure looks like a debris ball, I think, as the menacing supercell enters southwestern Noxubee County.  Then bzzzzz! The phone buzzes again.  This time it is one of my sisters.

“Erin, do you know what’s going on?”  Of course.  I have been following it on the local news, which, unfortunately, is swamped at this point with several simultaneous tornadic supercells, and have just checked the Internet to see if anyone has reported a tornado with this one.  “Well, they are saying in the tornado warning that they’ve got a confirmed tornado—a big one.”

“Does that look like a debris ball on radar to you?”  She says that it does.  “All right,” I say, arriving at a decision instantly, as the crawl-space foundation of this old house flashes before my mind’s eye.  “I’m getting the cat and getting out of here.”  She agrees.

The tornado safety guidelines put out by the National Weather Service do not endorse leaving a house in a vehicle.  I understand why.  In general, a house can be regarded as a comparatively safe place to be in a tornado, whereas a vehicle cannot.  Moreover, it’s possible to get on the road and drive directly into a different tornado or an area of high winds.  When I tell my sister that I’m planning to leave, I know full well that I am going against this advice, and for all these reasons, I don’t recommend that to people in general—certainly not when there is not even a confirmed tornado, and in most cases, not even when there is one.  However, as a meteorology student, I have closely monitored the extreme atmospheric conditions that would be in play for this event.  I am aware that, under these circumstances, tornadoes that form are far more likely than normal to become “violent”—to reach an intensity at which even well-constructed homes are definitely not safe to be in because every wall in them is blown down.  I am aware of what to expect if I choose to drive through the precipitating part of a supercell.  (I was close to the wall cloud of one a week ago, after all!)  I am aware that there is a clear spot north of Noxubee County, and there is nothing that will enter that area in the immediate future.  And, most importantly, I have enough time to get away.

But only just enough.  There is no time to lose.  The storm is moving quickly, and at the angle it’s coming, it will be upon me in 20 minutes.  I grab my laptop, leaving behind even the power cords.  I reflect for a moment on the irony of this; I had recently seen my first AC adapter go out and had to get this one over the Internet.  Well, there is no time to waste by crawling under my desk and unplugging the cord.  I grab my purse.  I shove my protesting cat into the cat carrier.  Carrying only these things, I run into the vehicle, hoping that the lightly falling rain does not penetrate the laptop case, and apparently (so I discover later) leave a rut in the yard in my rush to get out of there.  I head north.

I would not leave my cat at home, but the delay in grabbing up these things has cost me a few more minutes.  Meanwhile, the tornado has not waited.  It’s best not to say what speed I am driving, but no one else heading north is driving any differently.  I wonder how many of them are on the road for the same reason that I am.  The rain slacks off.  I never run into any hail on this trip.

It is between Macon and Brooksville that I start seeing small pieces of branches fall from the sky.  They are not large enough to slam to Earth with violence, so there is something almost graceful in it.  I’ve never seen anything like this before.  These are not being blown about horizontally by winds; they are falling like soft rain from the storm itself.  The movement is vertical.  The branches have been sucked into the mesocyclone, which tilts southwest to northeast; the part of the storm that I am under is nowhere close to the tornado!  Seeing debris brings everything to mind that I have pushed out in my single-minded focus on getting away.  It occurs to me that people somewhere may see debris from my house later on.  Well, I’m safe, and the cat is safe, I think to myself.  There’s nothing more I can do.

Almost mockingly, the sunlight breaks out as I leave Noxubee County.  I hear the buzzing of my phone once more.  It is my father, who has tried to call me several times since my sister called him and told him that I had decided to leave the house.

“You’re fine in that part of Lowndes County,” my father tells me over the phone.  That fits with what I had seen on the radar; I knew that there was a dangerous supercell in Monroe County (the Smithville tornado, it turned out).  I also know that, though nothing tornadic is coming for Starkville and Columbus at the time, I do not want to be stuck in one of these cities if that changed in an hour or so, as it often does during tornado outbreaks.  I also don’t like the idea of pulling off the road indefinitely.  I decide to stop at the house of friends in rural Lowndes County, and there I stay for an hour or so.

I am fully expecting that I will not have a house to go back to, or my house will be damaged beyond repair, or the town will be destroyed.  I’ve read a lot of personal accounts of extreme weather events, and now it seems that I am about to live that awful aftermath.  It is truly amazing how we are able to push thoughts like this out of our minds when we are focused on something critical, such as (possibly) survival itself.  Now that this is not an imminent concern, the ugly realities of a tornadic strike hit me.  I don’t know exactly what will be damaged, or how much, but there is absolutely nothing I can do about it.  All I can do is wait to hear some news.  It is a hideous wait, and yet, I am focusing more on the animal confined in her carrier next to me, and the fact that no one else was at home, than on the home itself.  No one wants to lose a house, but when all is said and done, it’s just a house.

Over the course of my visit, my hosts learn that Macon has not been hit.  The storm apparently passed over with rotation still apparent, but no tornado anymore.  To my astonishment, there was apparently some nonchalance about the whole event in at least some parts of Macon.  Finally I decide to return home, since I still indeed have one.  I get there in time to settle in and watch with amazed horror as live footage of the tornado in Tuscaloosa airs.  Later, I see video of the Noxubee County tornado.  I find out through the TV news and Twitter—Macon, amazingly, has power—that many people in the Southeast are not so fortunate as I have been this afternoon.  My own experience is pushed back to a different corner of my mind as the hideous extent of the destruction and suffering becomes known.  I have not suffered loss.  I focus on those who have.

Running from an EF-5: Part One

Note: This is something a little bit different from the usual fare for this blog. Nothing particularly notable is occurring in the South except for the river flooding, and at this point that is a matter of concern for the engineers and hydrologists more than meteorologists. (I could rant about the Corps of Engineers, but that would be better suited for my other blog.) Since this is the case, and since I have felt that I needed to write about my experience on the tornado outbreak of April 27, I’m going to do that. This will be a three-part piece about my experience involving the EF-5 tornado that went through Neshoba, Winston, Kemper, and Noxubee Counties that day. The first part does not even occur on the 27th; it is from a week earlier. I’ve come to the conclusion, though, that it is part of the same story and an important factor in why I chose what I did that dark day.

Wednesday, April 20, 2011. I sit parked on the side of Highway 45, somewhat north of Crawford, in a line of other vehicles whose drivers have made the same decision.  We are all being pelted by hail that I estimate is up to an inch in diameter, mixed in with some rain.  No one dares to drive any farther.  This is, I believe, the third time that hail in this storm has forced me to pull off the road in my drive south to Macon from Starkville that afternoon.  I’ve almost lost count, and that’s not even counting the other storm that produced some hail before I even left Starkville.

I may be in the minority these days in having a cell phone that does not have Internet capability.  Ordinarily that is not a problem for me, but right at this particular moment, I find myself really wishing that I had a radar picture in front of me.  I call my sister and ask her to pull up a radar image on the Internet.  I have some urgent questions for her.

I obviously don’t want the car to be damaged by the hail, but what is currently happening to me is not my primary fear.  See, I have been at Mississippi State University as a graduate student in meteorology, I am well aware that the storm I am under is a supercell thunderstorm, and most likely, I am in what is called the “core” of the storm—the area of the heaviest precipitation.  Supercell thunderstorms are the ones that are most likely to produce tornadoes, and if there is a tornado, the core of the storm is located to the northeast of it.  And the storm is moving northeast.

In this particular situation, I could easily be in far more danger after the hail ends, and that is what I want my sister to tell me about.  I give her my approximate location.  “I think you’re in the hook,” she says uneasily.  (A hook echo on radar is an indication of strong rotation and possibly a tornado.)  My nerves tense at this, but if she can see a hook, it means that there is a slot of low to no precipitation, which I am most definitely not in.  “No, I’m not in the hook,” I say.  “I’m in the core.  I’m getting hammered.  But you are saying there is a hook with this storm?  Is there a tornado warning?”  It turns out that there is, and it is radar- rather than sighting-based.  That’s no surprise, and it certainly doesn’t mean that there isn’t a tornado.  The perils of high-precipitation supercells strike again.  I ask some more questions.  It seems that I am on the north edge of the core, and that based on the motion of the storm, I would be best off staying put until the whole thing is east of me.  The hook should pass south of me if I do that.  I just want to get out of this.  I’m not opposed at all to storm chasing, but it’s really not something I’m inclined to do without another person in the car or live radar available.

Once I am on the road again, I get a glimpse of the wall cloud off to my east with the aid of the nearly constant cloud-top positive lightning flashes that the storm would put out as it intensified.  I recall from my thesis reading that intensifying supercells often do produce a great deal of positive lightning.  This particular supercell would go on to produce larger hail and threaten Columbus.  I take several hours to come down from the adrenaline rush.  One week later, I would be very glad I had this experience.

Part Two of this will be about what happened to me personally on April 27. Part Three will be the aftermath of that event.

Dangerous Severe Weather Situation for Tennessee Valley

So far the severe weather outbreak of late April 2011 has brought devastation to many different areas in the southern Plains and Southeast. A tornado that was probably of a violent intensity hit a small town in Arkansas last night, reportedly removing pavement from the ground, and tornadoes have touched down from Texas through Arkansas and Louisiana into Mississippi, Tennessee, and Kentucky today. The severe weather is going to continue through Wednesday as the second of two low pressure centers intensifies over the mid-South.

A powerful mid-level jet has formed, with winds screaming at 70 knots or more:


Winds and pressure at 700 mb

This will provide powerful upward forcing. Additionally, strong thermal advection feeding into the developing low pressure center (which has become quite strong today, checking in at 989 mb as of this writing) will promote instability as well as additional vertical uplift:


Temperature and wind at 850 mb

For specific tornado threat regions, the NAM model has been fixing on an area centered over the Tennessee Valley. Yesterday the bull’s eye appeared to be right over the state of Tennessee, but today, it has shifted somewhat south.

The 18 UTC run of the NAM had CAPE over 4000 J/kg in central Mississippi at peak (mid-afternoon Wednesday), with a local maximum roughly over the Golden Triangle region and Tupelo, MS:

This run also had 3 km energy helicity values literally off the charts over the same area:

The 00 UTC run has CAPE peaking around the same value as before.

It has, however, moved, the helicity indices to extreme northwest Alabama:

Bottom line? I think that at this stage, such deviations are just model noise. Observations will be needed to determine exactly what area will have the highest risk of severe weather tomorrow. However, it is important to note that the model has been very consistent with an elevated tornado threat Wednesday afternoon in the Tennessee Valley area. With that in mind, I offer up my own subjective severe weather threat map for this event. I am not assigning probabilities to any of the color regions of this map; they are present only to indicate the areas that I think are in increasingly higher danger of seeing powerful severe weather, including long-tracked tornadoes.

Another Multi-Day Tornado Outbreak Expected

The year is off to a record start in tornadic activity, and more is on the way. Another three-day tornado outbreak is currently underway, with day 1 having brought close to 40 tornado reports and over 250 hail and wind reports.


The system is expected to bring more severe weather to the eastern part of the U.S. through Tuesday and Wednesday. At the time of this writing, a squall line has developed in the Mississippi River Delta that is expected to push east overnight, bringing strong winds, rain, lightning, and the risk of embedded tornadoes.


This scenario is more complicated than the setup for the last big outbreak, which had a single powerful system to generate the intense weather. A closed low located over Arkansas is responsible for the day 1 activity. This low formed today from a shortwave kink in an upper trough. This low is expected to be blocked by a strong high pressure system off the Atlantic coast, causing it to move north and eventually northeast to die out over the Great Lakes area. However, a second shortwave kink is expected to enter the mid-South and undergo cyclogenesis on Tuesday afternoon or evening.


Surface low at 36 hours


The combination of the current cyclone’s development, the upper-atmospheric jet that is causing all this shortwave activity (see below), and the next cyclone’s appearance on the scene will result in there being significant sources of uplift.


250 mb jet stream


700 mb upward vertical velocities


The highest values of instability in the event are currently prognosticated by the models to occur around midday tomorrow. The NAM and GFS generally agree on the areas of high CAPE, with each model forecasting at least 3000 J/kg (and it should be noted that models do not do well with CAPE and have a tendency to underforecast. Keep an eye on observations such as soundings).


Surface-based CAPE at 18 hours, NAM


Surface-based CAPE at 18 hours, GFS

For Wednesday’s event, the highest CAPE values are expected to be over Mississippi. The NAM and GFS agree on the maximum values but have the location and orientation of the high CAPE axis different.


Surface-based CAPE at 45 hours, NAM


Surface-based CAPE at 45 hours, GFS


The Energy Helicity Index (EHI) values for 18 hours and 45 hours (midday Tuesday and afternoon Wednesday) indicate the areas that the models forecast are most conducive for tornadic supercell development. Here is what the NAM indicates for the two times:


EHI at 18 hours, NAM


EHI at 45 hours, NAM

I am especially concerned about the middle part of Tennessee on Wednesday if that is accurate. Those EHI values are almost off the scale, and they coincide with an area of at least 3000 J/kg CAPE. That area has a history of tornadoes, and between the strong instability, powerful uplift, and helical pattern to the winds, I think it is quite likely that the Tennessee valley may be in the worst part of this outbreak.


The Storm Prediction Center has already put up a Moderate risk for Tuesday and Wednesday. It is thought highly likely that at least one, if not both, of these days will see that risk upgraded to High.

Tornado and Severe Weather Outbreak in OK/AR/MO Thursday

I am expecting a tornado and severe weather outbreak to occur tomorrow in what I would call the heart of Tornado Alley.

However, that’s a worthless statement from any perspective—forecasting, meteorological, geographical—without a qualification of the term “Tornado Alley.” I have long thought that the true “heart” of Tornado Alley is not over Oklahoma City at all, but is instead farther east. In truth, where that heart is depends on what one chooses to look at. (For long-tracked violent tornadoes, the bull’s eye is, beyond a shadow of a doubt, over Mississippi.) Some meteorologists developed the idea of three distinct tornado alleys: Plains Alley, the traditional storm-chasing zone in the Plains states that is often referred to with the general appellative of “Tornado Alley”; Dixie Alley, a broad region encompassing much of the Southeast but not including parts of Texas that are climatologically and geologically more similar to Plains Alley; and Hoosier Alley, which is essentially the Midwest. However, if you really take a good look at it, I think you’ll find that the entire eastern region of the country between the two mountain ranges is Tornado Alley. There is no distinct border where you can say, “There are not nearly as many tornadoes in this small region as there are in either ‘Alley’ on each side of it,” and in the absence of such a thing, the only reasonable thing to do is to say it’s all one Alley. (Besides, “Hoosier Alley”? Really?) With the whole of the U.S. between the mountain ranges as Tornado Alley, it seems reasonable to expect the core somewhere east of Oklahoma City, and the meteorological data bear this fact out.

Now that that’s out of the way, let’s take a look at the science behind this and see why there will be severe weather, as well as where it should occur.

The weather-maker will be a low pressure center forming east of the Rocky Mountains. A low that is currently analyzed at 1009 mb (which is to say, not much of a low) is located over Kansas right now. This cyclone is forecast by the NAM to open up as it weakens and become more of a broad trough. This should be occurring right now, in fact. After that low dies out, a strong upper-level low (700 mb) is set to close off its circulation over the TX/OK panhandles on Thursday. Upper-level cyclogenesis should have occurred by 9:00 to 10:00 tomorrow morning (CDT). The usual vertical tilt of these kinds of cyclones will result in the surface low being over central Oklahoma at this time, and it should be at about 1000 mb by midday tomorrow.

Thermal advections will be strong, with the greatest warm-air advection (WAA) occurring in an area just west of Tulsa, OK to north TX to the AR/LA border. Temperature advection drops off in northeast Arkansas.

Surface dewpoint temperatures will be quite high, approaching 65-70 from Tulsa south along the OK/AR border in a narrow swath. A larger area of 60-65 F dewpoints will cover the area from Oklahoma City through a diagonal line bisecting Arkansas NW-SE.

However, despite all this moisture around the surface, a dry socket should be present at the critical 700 mb level in the atmosphere. This shows up on a synoptic model forecast map as a medium brownish-gray blotch of low relative humidity.

All this will lead to mixed-layer CAPE values approaching 2500 J/kg.

The instability will definitely be present, and between the thermal advection and the diffluence-based forcing from the low, lift should be present over the target area. Decent upward vertical velocities are forecast by the models to be present over the area of interest.

A cap of CIN will be in place early in the day, but as the surface heats up, the surface temperature should remove the temperature inversion, making it easier for moist parcels of air to rise.

The disappearance of the capping inversion also shows up in forecast soundings for the area bounded by Tulsa, Fort Smith, and Fayetteville. They indicate that a substantial warm surface layer will have developed by late afternoon, and with CAPE values as high as they are expected to be, this will set off convective growth.

Helicity values are also supposed to be quite high, with 0-3 km at 300-500 over the target area, at a time coinciding with the high CAPE values and the eroded cap, which will support supercellular development instead of linear. This increases the risk of tornadoes and large hail.

The Storm Prediction Center has already put up a Moderate risk for this general area, which I agree with 100%. I wouldn’t be overly surprised to see them go with a High by tomorrow morning, though this is a small enough area that they may not do it on the basis of its size. In any case, I am growing quite confident of a risk of supercells tomorrow for eastern Oklahoma, western Arkansas, and southwest Missouri. If I were to pick one city that I think has the highest chance of a significant impact, I’d go with Tulsa. However, a slight shift in any direction of the low’s movement would result in a shifting of the severe weather risk, and these events are never points, but areas. It is looking like a potentially dangerous day tomorrow for this region of the country, so people living there should make sure they keep aware of what is going on in the afternoon and be ready to act.

Severe Weather Season Starts Off With a Bang

Having lived in the Southeast for most of my life, I find it a bit dishonest to say that there is a defined “severe weather season” for this region. We, after all, are the ones who rang in the new year with a tornado outbreak. We’ve had tornadoes a couple days before Christmas, in January, during hurricanes, and occasionally even during the summer months. We have certainly had powerful thunderstorms in the summer months that have brought extreme wind, hail, and lightning. However, it is fair to say that there are periods of the year when the most destructive types of severe weather are more common than others, and perhaps these are what should be defined as the “severe weather seasons” for the South. If this is the case, then we are entering the spring severe weather season. We’ve already had a taste of it when the powerful squall line came through last week on Thursday and Friday, generating several tornadoes and 100-mile-an-hour winds in Tennessee. That, however, was just a teaser, and an event is currently in the process of unfolding that promises to be more significant. Tonight’s event has a couple of parameters in place that last week’s did not, at least not to a great extent, plus all the parameters that last week’s event did have. I am using images from the RUC model, which is a short-range model containing live observations that is run every hour.

First of all, there is a strong jet over the region.

This jet will have uplift occurring on its northeast side. Since it is a 120-knot jet, this will provide significant uplift.

Secondly, there is a decent thermal gradient and fairly strong winds at the 850 mb level.

This will provide fuel to the strengthening low pressure center that is the source of the upcoming bad weather. The advection of warm air is also a source of uplift.

There is also an area of vorticity (essentially, rotational motion of wind). Vorticity is not the source of tornadoes, despite the suggestive name; in fact, uplift associated with vorticity is not directly under the strongest area, but is instead downstream of it. The vorticity is being advected in that direction by the motion of the air.

Incidentally, that accounts for three parameters providing favorable severe weather conditions over pretty much the same spot: Memphis/St. Louis/NE Arkansas.

Now, the parameters that last week’s event did not have in place relate specifically to supercell formation. Last week’s event had plenty of uplift and wind shear, just not a lot of atmospheric instability and helicity. Tonight’s expected outbreak is a different story.

Here’s atmospheric instability, as measured by CAPE (convective available potential energy):

There are values approaching 2000 J/kg expected over, yes, the same area where a lot of uplift is supposed to be occurring. This isn’t a stupendous value, but it’s pretty significant for an early-season outbreak. Some of the highest values of CAPE I’ve ever heard of were on the order of 6000 J/kg for events that produced F5 tornadoes. Where other favorable parameters exist, CAPE is directly correlated with the formation of thunderstorms.

Here’s helicity.

The peak swath of high helicity is a little off-center compared to the rest of the favorable parameters, but a pretty significant area of it is nonetheless centered right over the area in the lower Midwest/upper South where the rest of the favorable parameters are converging.

I am expecting a tornado outbreak over this area tonight, pushing eastward into Tennessee in the early morning hours. This is not good, as nighttime tornadoes are easily the deadliest of all. (See my previous blog entry about the tornado outbreak of February 5, 2008 for more about Southern tornadoes.) I completely agree with the Storm Prediction Center’s decision to issue a moderate risk for this region tonight and tomorrow. I think it is possible that tornadoes may occur a bit farther south than currently indicated, mainly because of that high helicity index, but I am not expecting there to be a very great threat much south of Highway 82.

Lessons of the Super Tuesday Tornado Outbreak of 2008

2008 OutbreakAt the moment, Arkansas is apparently under a blanket of snow in many areas, and more cold weather and possible winter precipitation is apparently expected. This is interesting to me because at this time not too long ago, Arkansas was essentially “ground zero” for an entirely different kind of destructive meteorological phenomenon, which had also occurred in a La Nina winter. Three years ago, a violent, extensive tornado outbreak began in the South and lower Midwest. Because the outbreak began on “Super Tuesday,” the day when the largest number of state presidential primaries were held, the outbreak was dubbed the Super Tuesday Outbreak. It left in its wake 87 confirmed tornadoes, which killed 57 people—a death toll unheard of for almost 25 years.

The system that produced the outbreak, a low pressure center with a deep warm sector and powerful trailing cold front, moved from Texas to Missouri during the outbreak. Upper-level jets provided decent support, with a jet approaching 120 knots located over eastern Kansas, Arkansas, Missouri, and Oklahoma. Instability was sufficient to promote severe weather in the form of supercells, with 1000 J/kg of CAPE observed. This is not particularly spectacular for the usual tornado season, but it is not often found in cold-season outbreaks. High shear as a result of the jet further fueled the outbreak. The tornadoes occurred on the east side of the cyclone, as is usually the case.

Though it was not as intense as the Super Outbreak of 1974 that produced either 6 or 7 F5 tornadoes (depending on what source you believe), this outbreak was typical of large-scale tornadic outbreaks in that it produced a lot of violent and strong twisters, which normally occur in comparative isolation. In this outbreak, five EF4 tornadoes were observed, including a long-tracked tornado in Arkansas that had a path length of over 120 miles. This tornado alone was responsible for 13 of the fatalities. Another long-tracked tornado, this one an EF3 in Tennessee with a 51-mile path, was responsible for 22 deaths.

This relates intimately to a point that I have been making for several years, and what I think is the principal lesson of this outbreak and its calamitous death toll (for modern times). The South, particularly AR, LA, MS, TN, and AL, is subject to far more violent long-tracked tornadoes than any other region of the country. I mean it isn’t even close. Here is a PDF of a paper by two professional meteorologists about the statistical patterns indicating the real “tornado alley,” at least as far as F3-F5 long-tracked tornadoes are concerned. If you are frightened of statistics, this is a graphic from Jackson, MS that puts up the raw data, the paths of all long-tracked tornadoes rated 4 or 5. It’s not exactly the same data being looked at in the graphic and the paper, but nonetheless, the point that is coming through loud and clear is that the Deep South is the primary breeding ground for the most dangerous, destructive, and deadly kinds of tornadoes that the atmosphere can produce: strong-to-violent long-tracked twisters.

Furthermore, the South is more prone than any other region of the country to have substandard housing that is particularly vulnerable to tornadoes. There is no reason for anyone in a house to die in a weak tornado (an EF0 or EF1), but it still happens, and poor construction or deterioration is to blame for a lot of it. Mobile homes do not help the problem, and it is sickening to me (I guess I still have fragments of a social activist in me) that one’s survival in a tornado, in the United States in the 21st century, could depend on one’s financial status, but this is how it is. Also, most buildings in the South do not have tornado shelters. Even large venues in the South often have no severe weather plan. I’m going to pick on one, the Georgia Dome, which was brushed by an EF2 tornado while a basketball game was going on later in 2008. There was little official word given to the spectators about what was happening until the tornado was actually upon them, making the ceiling shake! There is an infamous YouTube video from inside the stadium as the tornado was passing by, so you can see for yourself what I mean. The only reason that didn’t result in a major disaster was that the game went into overtime, keeping the fans inside the stadium. And finally, unlike the other parts of Tornado Alley (“Dixie Alley” and “Hoosier Alley,” the South and Midwest respectively, are not really separate from Plains Alley), the South does not, in many locations, have an effective local warning system. I don’t mean the issuance of warnings. The weather office in Jackson is very effective at producing warnings. If anything, they overwarn! There is not a problem with the warnings being generated. The problem, rather, is with dissemination to the public during times when the public is most likely not able to know what is going on.

Some areas are better than others at issuing tornado alarms. I can think of one time when Oktibbeha County, Mississippi sounded its alarms for apparently the whole county when the tornado in question was on a far corner and headed out. This is how it happens more often than not when a county has tornado sirens. This kind of thing, in my opinion, is overwarning, and it can lead to complacency, but it is preferable to the alternative. I witnessed that on the New Year’s Eve 2010/2011 tornado outbreak when a very strongly rotating supercell entered Noxubee County, MS around 2:00 AM. It was quickly warned by the National Weather Service. However, the county did not sound its alarms until the tornado was actually on the Mississippi/Alabama border! It was an EF3 tornado, and by that time, it had already done its damage in the rural part of the county. I’ll repeat this: 2:25 in the morning on New Year’s Day, and the tornado alarm was not sounded until the tornado had already passed. This is inexcusable. It is fortunate that no one was killed. Some other areas do not have tornado alarms at all. The major problem in the South, which contributes to the higher death tolls from tornadoes in the South than in any other region of the country (including the other two parts of Tornado Alley), is that of public awareness.

However, in general I do not mean awareness of the event as it is ongoing. Unless it is a nighttime event or is occurring when people are likely to be occupied with something else (such as watching a basketball game or being drunk, as the case may be), I think most people are aware if there is a tornado warning for their county. The local media are still the primary source of people’s information about weather warnings (most people are not plugged into the Internet 24/7, and as horrifying as it may seem to some of you, most of the South is still on dial-up). Most TV stations here interrupt regular programming to provide information. WCBI in the Golden Triangle region of Mississippi is good at providing detailed information about what is happening (though, in my opinion, there is a slight urban bias in what events warrant cuts in programming). When I say that there is a public awareness problem, it is referring more to long-term awareness. Awareness of the risk, which (as the paper and the graphic I linked above indicate) is much higher than they seem to realize. This is the sort of awareness that might lead people to build a tornado shelter underground, designate a specific room in the house to be shored up (which, if done properly, generally provides shelter against up to EF4 winds), and have a plan.

The affordability of shelter is a separate problem, and it is one that, as I said, positively sickens me. Back in the day, after the F5 tornado of May 3, 1999 that hit near Oklahoma City, there was a government rebate for people who purchased tornado shelters. It’s difficult in this day and age to make the case that the government should spend money on ordinary people, it seems, but it really wouldn’t hurt that much for it to at least be a tax write-off.

The Super Tuesday Outbreak demonstrated that, even in the modern era of NEXRAD, real-time weather forecasts, high-speed Internet, cell phones, and PDAs, a tornado outbreak lasting essentially a day and a half can still kill over 50 people. A single tornado, one that was not even rated “violent,” can kill over 20 people. Meteorologists for several years have been saying that there is a tornado awareness problem in the South, which is manifest in public complacency, poor shelter options, and ineffective tornado siren policies. Given that the Deep South is, quite clearly, the bull’s eye for the worst kind of tornado that the atmosphere can produce—long-tracked violent ones—this state of affairs is unacceptable. The Super Tuesday Outbreak will be repeated at some point—if not in the exact same geographical location in the South, then somewhere else in the region, and it will have an aftermath that is the same or worse unless something is done about the problem.

Another Winter Storm Coming Very Soon

Here in Mississippi, we are on the verge of getting yet another round of wintry weather. We missed the worst parts of the Groundhog Day Blizzard and even the severe weather was, for the most part, a whimper rather than a roar. However, our luck is about to run out. A system from the Gulf of Mexico is entering the area, and after vacillating for days, the forecasters (both human and computer) have reached a certain amount of consensus about what type of precipitation it will produce and for where.

The writing was actually on the wall last night, in my opinion, when both the 00Z NAM and the 00Z GFS came in with very similar QPF forecasts, both in amount, in type, and in geographic location. What they predicted was not good: up to 3/4 inch of freezing rain, based on the soundings, plus snow, sleet, and rain mixed in (what they call a “wintry mix”). I am not completely sold on 3/4 of an inch of freezing rain now, because it is a bit colder today than expected. This observation ties in with my general philosophy for using the computer models. I don’t pay that much attention to what the big flagship models say for short-term forecasts less than a day out, because they do not accommodate actual observations in their forecasts. Models exist that do, and if you are going to use any models for “nowcasting,” these should be the ones rather than models like the GFS and NAM. So whatever those two are saying today, I’m not paying attention to it. Instead, I am looking at what is happening, and what is happening is that there is more precipitation showing up on radar than the models “believed,” and the temperatures are colder. Not noticeably colder, mind, but just cold enough that it could, I hope, mitigate the amount of freezing rain by changing over to sleet and snow instead.

I think that areas north of I-20 are all going to get something that is frozen. Unlike the last snow event that we had in the South, the one of January 9-11, I do not expect that north Mississippi—the areas that are in the Memphis National Weather Service office’s region—will hit the jackpot. Instead, all indications are that QPF and therefore accumulated precipitation will drop off sharply north of Highway 82 (perhaps Highway 78). The belt of maximum precipitation with this system should be a bit farther south than the January storm.

There will be freezing rain for pretty much every area that gets anything frozen. Fortunately, I don’t think it’s going to be as bad as it looked last night, and it should be mixed in with sleet and snow. The atmosphere is hovering around freezing at several levels, and this will continue until it warms up Friday morning and the precipitation shifts to plain rain. The rain may help to melt whatever falls tonight and early Friday morning. Most of what happens on Friday will be a tapering off, especially since the observed QPF with this system is coming in stronger and a bit earlier than the computers predicted.

There is actually the possibility of yet another round of something in the middle of next week, after a brief “warm” spell this weekend (don’t hold your breath waiting for springlike conditions like what we had last weekend, however; the warmest high will be about 52 on Sunday). It’s too early to say just what this will be, especially since the safe thing to do with respect to such forecasts in the South is to be conservative and forecast cold rain at first, and then modify the forecast as the event approaches—if conditions warrant. Right now, a high of 38F on Wednesday does not warrant modifying the long-term forecast, as utterly miserable as rain would be in that. We’ll have to wait a bit longer to see how next week’s system shakes out, so let’s take this one winter storm at a time for now.