- Posts: 474
- Joined: Sun May 01, 2016 6:43 pm
Pecos Hank – Published Nov, 22 2017
Scientists place instruments inside a tornado to learn why one supercell produces tornadoes and another does not. Atmospheric Scientist Dr. Leigh Orf takes a different approach by growing storms that produce EF5 tornadoes in a supercomputer. / Rare EF5 tornadoes produce wind speeds estimated over 200 mph. Many people ignore tornado warnings because of too many false alarms. Others don’t respond because they don’t see a tornado.
On May 23rd 2011, near El Reno Oklahoma, a powerful rotating thunderstorm--a supercell—could produce a tornado at any moment. Tornado sirens screamed through town, but no tornado touches down. It was another false alarm. About the same time the next day, again near El Reno, another powerful supercell is in progress and a tornado warning is issued. This time a killer is on the ground.
<radio>”. . . your lives may depend on it. Once again, tornado warnings are in effect, a dangerous, deadly tornado outbreak is underway.”
With wind speeds estimated at over 200 miles per hour, this tornado earns the highest rating on the Enhanced Fujita scale, a rare EF 5.
<video> “Everything's destroyed, everything's killed.”
For over an hour this tornado will carve a 63 mile path across rural Oklahoma. Despite early warnings nine people lost their lives and over a hundred and fifty were injured.
Many people ignore tornado warnings because of too many false alarms. Others don't respond because they don’t see a tornado.
<video> “There it is. Coming right this way! Those people gotta get out of here.”
When they finally can it's often too late. The National Weather Service has the difficult task of trying to warn the public. But because we still don't know why one supercell produces a tornado and why another does not they have to warn with the majority of supercells. The National Weather Service is damned if they do and damned if they don't. To reduce the cried wolf effect we're going to have to figure out why one supercell produces violent tornadoes and why another does not. Easier said than done.
Every spring scientists risk their lives hoping to collect the data needed. Scanning killer storms with radar point-blank. And by attempting to place instruments in the path of tornadoes. And with all the probing and scanning the picture is still vague at best. Atmospheric scientist Dr. Leigh Orf has taken a different approach by successfully growing super storms that produce EF5 tornadoes in a supercomputer. After decades of wandering in the dark this technology has brought an alien anatomy to light. We now have the tools to see through the skin of the storm, into individual organs and an organized system of currents. Like a child opening a machine for the first time, we can now observe the components that make it tick. And what the machine is telling us is that many of our previous theories are dead wrong. But is this super storm growing and living inside a computer a true representation of nature? One of the ways to help validate these incredible simulations is to compare them with actual storm footage.
Hank: We are now here with atmospheric scientist at the University of Wisconsin Dr. Leigh Orf. Hello Leigh, greetings from Texas.
Leigh: Hi Hank. It's great to hear from you up here in Wisconsin.
Hank: Could you tell us what you do and what the main objective of your work is?
Leigh: Sure I'm an atmospheric scientist. I study supercell thunderstorms and tornadoes. But I use supercomputers to simulate supercell thunderstorms at pretty much the highest resolution possible with today's hardware. I have access to the Bluewater supercomputer at the University of Illinois and it is able to crunch up to 10 quintillion calculations per second. And it allows us to simulate super cells at resolutions that capture things like tornado Genesis and a lot of the different smaller scale vortical features that occur in super cells that lead up to tornadoes forming and probably involve the tornadoes’ lifecycle itself.
Hank: “What I think the public needs to understand, if I understand correctly, is that you are programming the laws of physics into a computer and then just hitting go and then the computer grows the thunderstorm.”
Leigh: All of those models are obeying the laws of physics as we understand them as human beings. Once they go there's no human interaction with the model until it makes its forecast. The initial conditions for the model that we took were from a pretty faithful sampling of the air that would have been along the right flank for a specific storm on May 24th 2011. And the result was we got a storm that has some very similar traits to the actual storm that occurred on May 24th 2011. So we've got these big storms that have actually occurred. We're gonna try to sort of bring them to life within the computer so that we can better understand what's going on.
Hank: In the past, I would have thought that a tornado is composed mostly of warm inflow.
Hank: And your models are suggesting that the cold pool is what is feeding the tornado.
Leigh: Yes, absolutely. So let me ask you first, Hank, Why did you think that in the first place? What was your thinking?
Hank: Because warm air is more buoyant. So you just assume that warm air is gonna be more easily lifted.
Hank: Also when you're in the path of the tornado, more often than not you find yourself in the warm in flow that is streaming in, you know, and around.
Hank: Then, also, when a storm becomes outflow dominant there's no more tornado.
Leigh: Right. Exactly. First of all I assumed the same darn thing before the simulation. It is counterintuitive to think that the cooler air is being lifted at over, you know, 150 miles an hour, half a kilometer above the ground. Which our simulations show. Which is mind-blowing. And yes, our simulations of this particular storm (May 24th 2011) in all cases, when you trace the air with trajectories (when you drop in these little parcels) and you put them in front of the storm, they'll enter the updraft but they won't go into the tornado (vortice). In contrast, the parcels that you drop in the cold pool (especially certain regions of it) the take a beeline right for the tornado and get pulled up into the tornado (vortice) circulation. The fact that we're not seeing any air coming from the warm side to feed the tornado is an interesting result. And if it turns out to be true I think it's an important result.
Hank: Near a supercell there are many fascinating things to observe that often overshadow the inflow band. Your animations are suggesting there's a lot more to the inflow tale than just a thin cloud streaming into the mesocyclone.
Leigh: Yes. First of all, the fact that it produces a tail cloud is good because tail clouds are common in super cells, along the forward flank. They are not always there. But that interface is where a bunch of interesting things can happen. Let me talk about a couple of the things we've found in our simulation that I don't think has been seen in Mother Nature. We have found very interesting organization of vorticity in the cold pool of our simulated storm. And one of the features we've identified, or at least we've give it a name, is called a stream-wise vorticity current (SVC). And this is what you could call a helically flowing tube of air. It's sort of lifted, tilted into the storms updraft where it becomes rotating cyclonically (ie. in the same direction as the air in the mesocyclone in any subsequent cyclonic tornado). So this feature is one that shows up in our simulations of the May 24th 2011 environment and in all the simulations we've done. So one of the tools that we have used with our simulation is to place these in air parcels—sort of like chaff. You just release them every second or so in a certain region of the storm to see where that air will go. Also, we’ll follow the temperature, the humidity, the precipitation, and the pressure along the path of the parcel and the forces acting on it. And that's a very powerful method for trying to untangle the physics of what's going on.
Hank: We see dozens and dozens of these little vortices along the forward flank that merge together and that sometimes get assimilated into the cyclonic flow of the main tornado. If they're anti-cyclonic they get twisted around the tornadoes periphery and spun and stretched and tilted—and all sorts of cool stuff. I know that some observational studies by Bluestein and Wurman and others have shown that these vortices are out there. You can't see them with the naked eye. Sometimes they'll spin up some dust if you're lucky. But oftentimes there's all these vortices in the air that we cannot see with the naked eye. I'm thinking of creative ways to use videography—like what you produce, Hank—are there ways for us to get some more information from the video. If we're lucky enough to get something kicked up by some dust, and all that, then we can start to look at the full three-dimensional wind field.
Hank: When you're underneath the storm—and let's say it starts off and it's got kind of a linear mode—and then you see that RFD (Rear Flank Downdraft) start to march around, it's usually ‘go time’.
Hank: And, that's when I take position.
Hank: Because it won't be long before the show really starts, which is why we generally think RFD as a mechanism in tornadogenesis. But it might be all of these things working together that are also driving the RFD, it could be the SVC, like you say, that's driving the RFD in sync coming around, perhaps.
Leigh: Perhaps. You raise a very good point. It's a matter of correlation versus causation. I mean, this theory is still the leading theory of tornadogenesis—that a downdraft impinges the ground, it spreads out like a downburst does, it forces convergence of existing vorticity, and once you get it stretched—boom—there's your tornado. Now even if the RFD is an important source of tornadogenesis, I don't see how down drafts in the rear flank can maintain a tornado. Because the rear flank is a very sporadic place. It's not like there's this constant stream of downward air going down, nice and smooth. It's much more transient. And these down drafts happen all over the place in the rear flank. So, perhaps, there's something else going on that is causing both the RFD and the tornado. So, the RFD is more of a symptom than a cause. And again, I don't know that I'm right.
Hank: When I look at your animation, there's a view where we start from the stratosphere and we come down through the tropopause, we see the overshooting top. I've been observing these storms now for 20 years. I've seen eight EF4s and one EF5. In my opinion (pointing to the animation) that's a supercell. All the features move and behave like those I've seen in the field: the rolling plumes of the updraft; the tumbling the mammatis in the back shear region;
Hank: (continued) the wall cloud appearance; the tail clouds streaming in along the forward flank; the laminar winds sculpted region of the mesocyclone; its demarcation from the turbulent region; the tornado; it's multiple vortices; the horizontal vortices riding up the tornado; the way the RFD wraps around the tornado and sometimes clears and wraps again; even the anticyclonic tornadoes in the vicinity, I've observed in numerous supercells.
Leigh: It looks very good, as you say, all the different components that you see in the field show up in the simulation. But as far as calling it a day, in science we have to write journal articles that get into the nitty gritty physics—the quantitative analysis. And that is what we haven't done yet. But it's hard to argue against some of the animations when we see the correlation to things we see in the field. Like you say, the flanking line looks right; the laminar transition to turbulence looks right—it's pretty compelling. Just to look at that and say okay well that happened in a model simulation so in a model simulation contains all this physics and such so it's likely that the simulation is not broken you know. The worst fear I have as a scientist is that we discover a bug in the model and that invalidates our results. Well, I don't see that happening. There's no tricks. There's no Hollywood's graphics going on. It's all manifestations of the data that’s on my hard drive.
Hank: How can actual storm footage help your cause?
Leigh: By validating what the model shows; because some of the most valuable information, for me personally, is looking at the storm structure, and looking at how it changes. Because I can put your time-lapse next to my animation and I can compare. So that is crucial. The best way to help people like us—who are trying to do real storm research—is to put it on a tripod (please don't shake the camera around, it just drives me nuts, it's hard to see things.) If you see little tiny vortices anywhere outside of the tornado I'm very interested. Who knows. I'm not saying that it's definitely going to advance the field. I'm saying that it already has helped me understand more about real storms, by looking at footage like yours and other people, because then I can compare the model with what is being seen in the field—in the one area where we overlap, and that's in the cloud.
Hank: Well not that I need it but I'm definitely more inspired to get out there and get some of these angles and and send them your way.
Leigh: Absolutely and you know (we talked about this earlier) maybe I can take an angle on my simulation that is what you took in your video. And we can do some comparisons
Hank: To collaborate with you is a whole lot of fun and excitement for me. I mean, I'm by no means a scientist, but contributing to the field is extremely rewarding for me.
Leigh: Oh, I can't tell you how much I appreciate this. And I take this seriously too. I take this as seriously as any collaboration.
Hank: If you'd like to know more about Dr. Leigh Orf’s fascinating research you can visit him at ORF dot media (orf.media) or check out his YouTube channel.
James McGinn / Solving Tornadoes
- Posts: 474
- Joined: Sun May 01, 2016 6:43 pm
JMcG: Very interesting conversation. You guys are inching up on an understanding that may, eventually, allow you to break away from the current paradigm. There is nothing trivial about achieving such. However, there are some really big conceptual leaps of faith you will have to make in order to do this. Hopefully some of my comments below will nudge both of you in that direction.
Leigh: The worst fear I have as a scientist is that we discover a bug in the model and that invalidates our results. Well, I don't see that happening. There's no tricks.
JMcG: I think the “bugs” in your model may prove to the the most valuable parts of your model. For example, in a sense you already have a bug in your model in that you show cold, drier air traveling at 150 mph up the funnel tube of the tornado in your model. This is a sharp contradiction to traditional storm theory that assumes “lighter”, warm, moist air provides the power to storms.
Hank: In the past, I would have thought that a tornado is composed mostly of warm inflow. And your models are suggesting that the cold pool is what is feeding the tornado.
Leigh: Why did you think that in the first place? What was your thinking?
Hank: Because warm air is more buoyant. So you just assume that warm air is gonna be more easily lifted.
JMcG: Traditional theory—based on convection—is fundamentally flawed. Actually, warm, moist air is actually heavier (not lighter) than cool, dry air. Moreover, very little of the observed flow of earth’s atmosphere is consistent with the simple up/down motion of convection—ie. jet streams.
Leigh: I assumed the same darn thing before the simulation. It is counterintuitive to think that the cooler air is being lifted at over 150 miles an hour, half a kilometer above the ground. Which our simulations show. Which is mind-blowing. The fact that we're not seeing any air coming from the warm side to feed the tornado is an interesting result. And if it turns out to be true I think it's an important result.
JMcG: It is both true and important. (But don’t expect anybody else in meteorology to even acknowledge it—political factors being what they are in this discipline.) The notion that moist air is lighter than dry air has never actually been tested and it’s based on an absurdity, that the moisture in clear moist air has (magically) turned to steam (gaseous H2O) at temperatures far below the boiling temperature/pressure of H2O.
Additionally, according to my model, the energy of storms comes from above, in the form of low pressure. It does not come from below. The source of the energy of storms is a jet stream (and/or a tributary to a jet stream) and it is delivered by vortices that grow from a jet stream (and/or tributary to a jet stream).
Hank: What I think the public needs to understand, if I understand correctly, is that you are programming the laws of physics into a computer and then just hitting go and then the computer grows the thunderstorm.
Leigh: All of those models are obeying the laws of physics as we understand them as human beings.
JMcG: Yes, they are based on the laws of physics AS WE UNDERSTAND THEM. Obviously they are not based on physics that we haven’t discovered, that are poorly understood and/or not widely distributed. Did you know that currently science (specifically, scientists who deal with the physical chemistry of H2O) recognize over 70 anomalies of H2O. Think about that. That’s 70 observations about H2O that currrent theory is unable to predict or explain! (Amongst these observations is H2O surface tension, which plays a huge role in the atmosphere.)
Along these lines, the notion that a vortice can possess the structural ability to deliver the low pressure energy of storms is an extremely advanced notion and a notion that is completely foreign to meteorology. It requires an advances understanding of H2O polarity and hydrogen bonding as a prerequisite—and this understanding is not yet available to the public. It is too complex to describe here and now. For the time being just be aware that H2O has surface tension and this becomes maximized along high speed wind shear boundaries, producing a kind of plasma substance that has genuine structural integrity. It is this spinning plasma that serves as the basis for the sheath of conduits of high speed, focused airflow—powered by low pressure—that are most commonly observable as the cone or vortice of tornadoes. And it is these vortices—usually unseen because they are, almost always, perfectly clear—that deliver the low pressure energy of all storms.
So, Leigh, consider this (above) in the context of what you stated previously, which I will repeat here:
Leigh (repeat): It is counterintuitive to think that the cooler air is being lifted at over 150 miles an hour, half a kilometer above the ground. Which our simulations show. Which is mind-blowing.
JMcG: So, Leigh, it may seem counterintuitive to you. But to me it is completely intuitive. According to my model vortices grow on energetic, wind shear boundaries (boundaries between bodies of warm moist [below] air and cool dry air [above]--the most prominent of which being the tropopause). The warm, moist air is the resource for the formation of the sheath and the cool dry air is what actually travels through the sheath. And this fast-moving cool, dry air is also the source of the energy that maintains the plasma that comprises the vortice conduit’s sheath. In order for a tornado to reach the ground its associated boundary layers must extend to the ground (or, at least, close to the ground). This is recognizable as a storm front and is plainly apparent in your simulation.
JMcG: A concept that is missing from our understanding of tornadoes is leverage. Without leverage the transference of the low pressure energy (which can potentially travel as fast as the speed of sound) from the jet streams will not be maximized. Tornado touchdown on land provides a solid surface for maximum leverage (*). And so, assuming the upper reaches of a tornadic vortice is still connected to the jet stream, tornado touch down allows for a potentially huge amount of low pressure energy to be delivered. This is what underlies the larger, more destructive tornadoes. But all storms require leverage. The more common way a storm gains leverage is through vortice growth. And so, longer lasting non-tornadic storms will involve multiple vortices growing along boundaries of moist/dry air, pulling in more moisture, retracting as they lose leverage, growing again, continuously.
(* This is the reason water spouts are so benign in comparison to land-based tornadoes—water cannot provide the degree of leverage that a solid surface can provide.)
(Also, I think hurricanes and tropical storm involve multiple vortices that collude to create their own leverage.)
JMcG: I think I can explain RFD and why it anticipates the appearance of the tornado vortice. But, frankly, you guys are already on the right track, in my opinion:
Hank: When you see that RFD (Rear Flank Downdraft) start to march around, it's usually ‘go time’.
. . . we generally think RFD as a mechanism in tornadogenesis. But it might be all of these things working together . . .
Leigh: Perhaps it's a matter of correlation versus causation. I mean, this theory is still the leading theory of tornadogenesis—that a downdraft impinges the ground, it spreads out like a downburst does, it forces convergence of existing vorticity, and once you get it stretched—boom—there's your tornado. Now, even if the RFD is an important source of tornadogenesis, I don't see how down drafts in the rear flank can maintain a tornado. Because the rear flank is a very sporadic place. So, the RFD is more of a symptom than a cause.
JMcG: To understand my explanation of RFD you have to keep in mind that in my model the low-pressure energy of storms is delivered from above by vortices that grow down into the moist air that is the resource for their growth. Accordingly, before the highly coherent tornado vortice forms there is a less coherent delivery of low pressure energy at higher altitude. This causes a more general updraft of moist air. (It is also what creates the wall cloud and, essentially, the whole supercell.) Wben the coherent vortice begins to form and grow down this more general updraft stops as the more general delivery of low pressure is now focused in the descending cone of the vortice growing toward the ground. And, therefore, the precipitation that was previously suspended by the general updraft is no longer suspended, causing very heavy precipitation. But, to be accurate, RFD really isn’t a downdraft. It’s just heavy rain (but I get the sense that you already realize this).
JMcG: Another “bug” in your model/simulation. And, again, one that I think will prove to be very valuable, involves how the different gases of your model maintain separation from one another under high energy conditions. In nature—according to our current understanding of physical principles—gases can't maintain separation, especially under high energy conditions. Or, at least, they’re not supposed to--according to our current understanding of physical principles. So the fact that they do maintain this distinctness in nature—allowing for the existence of vortices and other weather related phenomena—is a genuine mystery. (As I will further explicate below, in my model I am able to explain this separation. But my model uses physical principles [ie. maximization of H2O surface tension] that only I understand.) Therefore, you must have some kind of hidden assumption in your model—some artifice of one kind or another—that achieves/maintains this separation. (Your model/simulation would not be capable of laminar flow under high energy conditions [ie. vortices] if it did not have this hidden assumption.) What exactly this hidden assumption is I wouldn’t pretend to know. It might be something as simple as maintaining a constant distinction between the cool, dry air and the warm, moist air of your model. Or it might be something more complex, like a variable distinction between the cool, dry air of your model and the warm, moist air of your model that takes place only under high energy conditions. I don't know. Whatever the case, the fact that your model does such a great job of painting an accurate picture of what actually happens in a storm (as explained so thoroughly by Hank) and the fact that—I assert—your model/simulation could not possibly paint the accurate picture it paints without this artifice is evidence that points to a missing physical principle—a physical principle that I have already discovered (I am sure).
Specifically, with my model the resistance to mixing only occurs with high energy wind shear and, even then, only when the initial conditions involve smooth boundaries. (Vortices can only emerge and grow with smooth, distinct boundaries, which themselves only form under calm conditions.) Most importantly, in my model the separation is maintained by way of the plasma that emerges on wind shear boundaries. This theoretical plasma also provides for a reduction in friction, lending to an explanation as to how high wind speeds are achieved over long distance—also something that current theory cannot explain. And so, I kind of have selfish reasons for valuing this bug in your model in that it confirms—or will eventually, I suspect—my theory.
Hank: When I look at your animation, all the features move and behave like those I've seen in the field.
Leigh: It looks very good, as you say, all the different components that you see in the field show up in the simulation. But as far as calling it a day, in science we have to write journal articles that get into the nitty gritty physics—the quantitative analysis. And that is what we haven't done yet.
JMcG: Tornadogenesis (and/or the study of severe weather) is a failed paradigm. Members of failed paradigms will be mostly concerned with making sure nobody else shows them up. Peer review is just a weapon they use to achieve that end. Don’t fall for it. What you’ve achieved is already head and shoulders above what anybody else in this field has achieved. Your simulations are the kind of quantitative analysis that is threatening to them, because it can be validated by non-scientists (like Hank).
Leigh: By validating what the model shows; because some of the most valuable information, for me personally, is looking at the storm structure, and looking at how it changes.
JMcG: I agree. And if anybody tells you that your methods aren't valid and then expects you to jump through some kind of hoop ask them if they have jumped through the same hoop with their model. They will not know how to respond to this.
JMcG: Lastly, I just want to thank you guys for making a really exciting and interesting video and having the temerity to broach the subjects you have broached. Its a refreshing departure from the brain-dead orthodoxy that dominates this discipline.
James McGinn / Solving Tornadoes
Who is online
Users browsing this forum: Bing [Bot] and 1 guest