A 10-m resolution quarter trillion gridpoint tornadic supercell simulation
Vložit
- čas přidán 19. 01. 2020
- AMS page for this talk: ams.confex.com/ams/2020Annual...
Leigh Orf's talk at the 100th American Meteorological Society annual meeting, part of the Severe Local Storms symposium, on January 14, 2020 in Boston, MA. Downloads of this video without audio but with slowed-down animations can be found at orf.media/AMS2020 - Věda a technologie
Here's what I want to know: If you run this program multiple times does it always evolve in exactly the same way every time?
An excellent question, near and dear to my heart! It's also the same question I asked the engineers working on the Blue Waters supercomputer before it was fully built - I wanted to know, and their own answer was they didn't know!
The short answer to your question about bit-perfect reproducibility is "yes" but only under specific circumstances. The CM1 model I use is compiled into a binary executable from source code (like all models that run on supercomputers). If I use the same exact binary I should get the same exact results, down to the very last bit, so bit-identical. I have verified this. However, a while back I tried to reproduce a simulation using a newly compiled version of the model and immediately saw that the solution was very slightly different, and the differences got larger over time (this is expected to happen with models like CM1 which simulate highly nonlinear phenomena). So what happened? Updates to the compiler and updates to the libraries on the supercomputer. The source code did not change, but the results were different in the least significant bits of data, and these small changes amplified over time, like the "butterfly flaps its wings" scenario.
In science, reproducibility is considered monumentally important. You may have heard news stories where social science or medical studies cannot be reproduced (usually they involve human subjects). This has been called the "reproducibility crisis." In my world of cloud modeling, I am no longer so worried about this because any given simulation is just one out of a nearly infinite number of outcomes. I can always create a new simulation, and often times I get a very similar result (i.e., a big fat EF5 tornado that lasts a long time) even when I recompile the model or run the model on a different machine. I am always careful to point these things out, and the atmospheric science community I think generally understands this. It's also why the focus from specific individual simulations has changed to *ensembles* of simulations where you look at a bunch of simulations in slightly different environments/configurations and do statistics on them. In ensembles, it is hoped that the statistics will not change much even if each individual run may have changed. But it is especially challenging to do ensembles at the scale I typically run my simulations at, since my whole approach is to push the hardware very hard. So I kind of consider each simulation to be similar to an actual storm chase - with a storm chase you only get one shot on a storm. For super big simulations (like the 10 meter El Reno 2011 simulation I did on Blue Waters) I will never get that bit-perfect result if I run it again because the machine that I ran it on no longer exists, and there will be very slight differences in mathematical operations on a different machine.
@@LeighOrfsThunderstormResearch - Thanks for your reply. Alas, it doesn't *_unequivocally_* answer the question, for at least two reasons. One of them you mentioned-namely, you haven't had the opportunity to run the simulation on the same hardware twice. If the way the program runs is dependent on idiosyncrasies that vary from one machine to another, then it would be impossible to answer the question with any certainty.
Let me back up and tell you why I've asked the question. I'm not concerned about a "reproducibility crisis". That's a concern only for physicists and others who deal with simple systems. You're dealing with a complex system; the Newtonian paradigm doesn't apply. The question is, have you truly succeeded in modeling a complex system?
Now, if you run the program multiple times on the *_same_* hardware and you get exactly the same result every time, you haven't created a successful model of a complex system. From my perspective, it's a "boring" result. That's not a criticism, and it certainly isn't in any way a deprecation of the massive amount of ingenious work that you've obviously done to create these awesome and unique simulations. Nor is an assertion that these sims are not a giant leap forward in understanding tornadogenesis. I rather expect the opposite is true...
@@LeighOrfsThunderstormResearch I might be wrong on this, but I don't suppose these kinds of models are great candidates for being run on GPU based supercomputers? I would guess there's too much interdependence between different data points for that to be efficient, but maybe I'm wrong on that.
Do you think newer systems are moving in the direction that would let you realistically run multiple iterations of the same model (maybe with slight noise added to the input data or model constants) in the next few years?
Wow Wow wow!!!
HELLO!!! WHEN IS YOUR TORNADOES OF 2019 VIDEO GONNA COME OUT!? :D
Yeah no kidding
I could fall asleep to these models they're relaxing as well as scientifically intriguing
Hypothetically, if a team were to place out smoke canisters on the front of an approaching super cell do you think we could use smoke to detect these invisible vortices that are forming the tornado on the ground before we see the vapor wrap from the sky down? Would it make a difference if the smoke was chilled like the dry ice effect before releasing it into the storm? Your work is incredible Pecos Hank, and of course Dr. Orf and his team!
@@jaycalhoun8714 you would need a lot of smoke like ALOT and it would need to linger and not just dissipate when the air moves it... that and without color differential it would be hard to see the different votices
Reed Timmer recorded a tornado at greenfield recently (it shows some wind turbines going down) but it shows the sub vortices talked about at 13 mins really well, and they look just like the model
Yup, you're not the first person to make the connection! Very cool stuff!
You beat me to this comment by an hour lol. Skip talbot told me to check this video out after I noted the unique structure. I recall watching it a while back, but it was cool to revisit after being mesmerized by reeds footage. Being over an open dust field created a perfect real life example of the simulation! Great stuff!
I was about to say the same thing and realised I'm 8 days late to the party! Was watching Ryan Hall's stream when it happened and had no idea how visibly striking the tornado actually was until the next day when Timmer uploaded the footage. Watching this video for the first time today having already seen Reed's video and the parallels are astonishing!
“I had so many hardware failures… I have a couple hundred terabytes of data, and that’s compressed”. I don’t know a better example of pushing computer power to its limits than your simulations. Brilliant work!
I've always been into stuff that pushes computers. I was into calculating digits of pi for a while. Then I wanted to get into weather simulations like this but it wasn't even possible at the time. Now it's barely possible. Can't wait to see what evolves. We need people who know the right questions to ask! Infinite processing speed will mean nothing without knowing how to use it. This is a good use.
Just found your channel, Pecos Hank sent me! Great work and collaboration, keep it coming!
your model of tornadogenesis helps explain those whispy looking "mini tornados" that I see in videos near the main funnel that always form and quickly dissipate. fantastic work!
I've been a tornado enthusiast since I was 5 when the movie twister came out. Since then, I've studied and studied about super cells since I was a kid. Im 30 now and I'm inventing an instrument the will tell us why some supercells produce tornadoes and why some don't. I watch Orf's computer models and help me understand more and more. I watch tornado data every day and study the US radar every day during tornado season.
How its goin?
Yeah, what's brewing?
This looks similar to Reed Timmer’s drone video of the Andover, KS tornado in April 2022.
You can really tell where the multiple vorticities show up within the larger tornado circulation. Never seen that before in such detail in a simulation. You're definitely on the right track.
Really enjoy these uploads and look forward to see where this journey goes
Anyone know why he hasn't uploaded anything in a year?
Layperson confession here: that is just gorgeous.
Apparently, based on how the Ward chamber model worked with multiple vortices it allowed for a center where there was a downward draft which could be a path to a seat for the funnel vortex to sit in. Wall cloud has negative-particles and the debris shaft has positive+particles thus 2 different fluid types that don't mix.
this would make Tim and Pete so happy!! RIP Samaras Family!!!
All models are wrong, but some are more useful than others. This work is incredible. Bravo!!!!
@Critter145 - You're not wrong...but with a caveat. The laws of physics aren't wrong, as long as they are not used outside the context in which they apply. So, they're "models" in the sense that they accurately describe the behavior of the systems they represent. But those are all *_simple systems._* That's why the equations of physics work. Models that attempt to describe the behavior of complex systems (human behavior, the weather, the climate, etc.) are doomed to fail.
The reason is in the very definition of a complex system, which is a system whose behavior cannot be modeled by a finite algorithm. The kind of models you're referring to - all of which are wrong - are consigned to failure from the get-go, because they're an attempt to treat complex systems as though they are simple systems. It can't work.
It's true that, depending on the system involved, some models can provide useful approximations. But with a few rare exceptions, believing that the approximation is reality and betting your future on it is generally inadvisable.
Having been directly hit by a tornado last year, I’m watching this and confirming everything I felt in the air just before it hit. This is unreal!
Sorry you had to go through that. What do you mean about "felt in the air just before"? I'd be interested in your experience.
Leigh Orf's Thunderstorm Research thank god it was a weak tornado. A brief explanation: I’m an intermediate level chaser with 5 years of experience, and found myself stuck in the worst possible situation. My truck was buried in mud, and I had to walk over a half mile away to get phone signal. As I got signal, the air changed. The wind direction changed 180° and the temperature dropped sharply very quickly, and my ears popped. I knew it was a pressure drop, and I was in the inflow notch. I saw a funnel in the sky over the tree line very close to me. I saw a cattle guard and wrapped my arms and legs around it because they’re usually built into the ground. As for tying this video to my experience, before the tornado hit me, I felt the twisting and winding air being pulled upward around me. It was an extremely weird sensation but it was almost like I could literally feel the individual suction vortices. Seconds later the weak tornado has gone right over me, rain and wind being swirled around me, and tree limbs breaking. It had passed over, and I still held on to the cattle guard in case I was going to be run over by the ghost train. It dissipated shortly after, and the firemen came and took me to their station. I was unharmed, just embarrassed I ended up in that situation!
Leigh Orf's Thunderstorm Research I apologize for the novel. But when I watch your videos, I instantly tie it back to that day. Seeing the visual representation of the air, and relating it to the physical sensation makes it ultra real and amazing. So thank you for your continued research
@@IGY6Brah Thanks for relaying your experience... fascinating... if you could actually feel the updraft on the ground, you were in a really bad spot for sure!! Glad you came out of it unscathed.
Leigh Orf's Thunderstorm Research it could have been worse than it was! The fireman told me they thought they were going to recover a dead body. Sent a chill down my back realizing how lucky I was. I still chase today though, and I actually have a question for you! If had an idea of getting a high res thermal camera to film a tornado, just to see if it would actually be able to capture the difference between the hot and cold inflow. Do you think that would be possible, or a waste of time?
Nothing is more interesting to me than someone who is genuinely passionate and interested in their work.
I love your "revolutionary" work. I have been following since dark skies 2017 seminar. I wonder when/if you can put surface friction into the ground boundary? It seems that the simulation had a fine enough resolution to put 10m grid spaced 10-100m apart for studying the effects of surface friction and the disruption of cyclostrophic pressure on the genesis of tornadoes.
Also, I wonder why that one vortex that formed inside of the parent tornado "wobbles" around. It's strange as it moves from the presumable radius of maximum velocity where the subsidiary vortices form to the center of the tornado where the main central downdraft is located. Additional, it's strange how it lasts for so long. I hope it's not too important when you have to compose the research paper of doom.
Maybe it's the suction vortices "growing out of the ground" during the late maintenance that causes the sudden left turn seen in large tornadoes.
Next year there will be a free phone app that does this.
Humble Brag: "I don't know what it means but it looks pretty cool..." This guy seems to be a computer scientist that minored in Sever Weather. Anyway, this is the coolest thing I have ever seen. Thankyou for sharing your work on yt!
Thanks for sharing this, massively inspiring also in other fields of physics.
You should try studying the TRI state tornado. It lasted for 2 and a half hours.
And was the fastest moving tornado and it went through well 3 states.
13:02 it’s crazy that the motion of the primary vortex is related to the wobbling path of the overall hook on storms, when they bunker left or right
Thank you for sharing this simulation on youtube, it was very interesting to watch! This makes me wonder if there have ever been attempts to find initial conditions that produce extremely strong or wide tornadoes. The physical limits of tornado intensity is something that I am very curious about but there doesn't seem to be much research on that. Have there ever been simulations that tried to produce such extreme tornadoes?
I think it would be cool if you did a video on the hardware and how you configure and run your cluster... for us nuts that run clusters too but for other things ;-)
He uses the Blue Waters and Frontera supercomputers for this work.
This is extremely exciting. Is there adequate data from the Stoughton storm and the, what, 16 others produced that day?
Horizontal vortices seen in real event videos possibly be these anticyclonic vortices you're simulation is showing? This is impressive work.
if I am ever successful enough some day, I will find Dr. Leigh and give him additional funding. Loads of it.
I’m sure it’s been said before but keep in mind the storm this simulation is based off was exceptional and perhaps prone to skewed data. This feat is definitely impressive but I hope developments will come once more is learned and more people get interested in this
Impressive work!
Cool to see props given to Hank!
Incredible. Question (for the freak chance you read these comments): Do you think the same multi-vorticy carousel of the maintenance cycle is equally present in violent "drillbit-style" tornados, just more tightly condensed? For example, visually, neither the Manitoba EF5 nor the recent Minnesota EF4 appeared to have any dance partners to their incredibly tight funnels for long, sustained periods of time.
Alex, it is likely that there is a lot of variation between tornadic supercells. Keep in mind however with a dry boundary layer you might not be able to see visibly these little vortices, but they could be there. In very humid boundary layers you tend to "see" vortices as condensation occurs in the center. So I would say that yes, there is abundant vorticity in most supercells, specifically along cold pool boundaries, but I make no claim that the "parade of vortices" and SVC are necessary for tornadoes. I suspect these things occur more frequently in the big monster storms.
@@LeighOrfsThunderstormResearch Thank you so much for taking the time to reply! That makes sense. El Reno is such an interesting storm to be studying because it feels like the entire meso just lowered itself to the ground. I can't wait to see how your research develops! Your work is truly awe-inspiring.
@@AdevebA Thanks. To be clear, the El Reno storm I'm simulating is the first one, 24 May 2011. I am however working on 31 May 2013... stay tuned...
@@LeighOrfsThunderstormResearch THAT explains why I thought it looked small! Haha
10/10. I'm not a scientist, just a dude watching this video tonight as two storms passed over, getting 3-5% better at knowing what I was looking at 😂. Hey thank you for sharing your work! 🤟 I also saw that vid of Greenfield. Wow! And clear similarities to your model. Cool! A question... I heard "solving turbulence" is something uniquely difficult, and it's written that it's "solved in physics but unsolved in math". Do you have thoughts about this? 👍 Thx again!
man this guy is very sharp it seems
AMAZING simulation data.
Not sure if this will be answered, but here's my question: Are there any characteristics of your models that are fundamentally inconsistent with real-world data?
Also, I just wanted to say you have done some amazing work, so thank you for sharing! I'm fresh out of a ME degree and this subject pokes at my very limited fluid mechanics brain, as well as my childhood fascination with tornados. When I was young and "Twister" came out and I was sure I was destined to be a storm chaser... lol, glad that didn't pan out.
We don't have friction with the ground right, but neither does anyone else. But other than that, no major weirdnessess in the simulations. Heck, we discovered stuff in the model that later turned out to be in the atmosphere! Perhaps the atmosphere needs to pay more attention to these simulations (kidding of course haha).
Am I understanding correctly when I say that the vortices are the force of air on the ground rushing to equalize pressure in the atmosphere underneath a super cell. Similar to the effect a hole in a spaceship would have between inner and outer space?
I like the part showing the central vortex with a ring of suction spot vorticies orbiting the main vortex
I don't even know how I got here, but this is cool as hell
We indeed live in a complicated universe.
This is rocket science never before air simulate air flow and direction of a tornado can you imagine the simulate the air pressure, doctor Fujita would of been happy to be at this lecture!!
Great work Dr. Orf
Grant Sanderson's 3Blue1Brown CZcams channel has some awesome videos explaining partial differential equations. Even if you are familiar with the subject, they are still works of art worth watching. But for those less familiar, they are very approachable and clear. You might find that they help you appreciate Leigh's work more, since his model(s) are iterated systems of partial differential equations.
Question: Does Theta prime subscript rho mean Angular Velocity with respect to density? Trying to figure that out. (First time re-watching after a ChemE Fluid Mechanics course)
(Also, is there a place to look at the mathematics behind your simulations/an explanation behind them? I'd like to be able to use the knowledge gained after this admittedly hellish course in Fluid Mechanics for something interesting)
@@garethfuller2700 Potential temperature = theta. Theta prime is the deviation from the background theta (i.e., the environmental sounding). Positive theta prime basically means positively buoyant, negative theta prime means negatively buoyant. The rho subscript refers to density potential temperature perturbation. This includes the effects of hydrometeors which make the air "heavier".
The document you want is this, which describes the CM1 model and has the equations: www2.mmm.ucar.edu/people/bryan/cm1/cm1_equations.pdf
you know right the cold pooL in the Fore Front DowndRaft and the RFD is often a warm inflow but it onlyhelp strengthening the tornado helping the cold core of the twister, that reall looks like a Tight coiling DNA like spiral, and the FFD, is now more known and accepted as the SVC Streamwise Vorticity current is on the cold pool, the cool that get get sucked up by low pressure pertubation within what really makes a tornado spin and create those deadly spiral that may obliterate a house EF5 style but leave the next one slightly damage but the wrapping of the RFD, that rises and coils easily and snuffly allong that cold stream as it will mix and war until 33k feet or so and the SVC streamwise vorticity current also known as the main inflow get, eiteir 60 - 30 or 2 degrees of the ghost train clouds, and forms a oclusion cloud at the base of the condensation field of a conado* or a stovepipe twister* above that or around that often obsuring is the wall clouds ...POINT IS that that tail and ghost train cloud is a result of real twisting tightly rotating SVC on the ground invisible BUT IT's there, did you check out lee orf work especially the video that compares it with real storm chasers like PECOS HANK had a look with doctor or professor (idk) lee orf and abng everything matches up with the supercell, i mean EVERYTHING and it's obvious now that the EF tornado's especially feed of the cold pool and the FFD keeps cooling that field and the RFD is just there as a consequence but is also important of tornado genesis as it created a presure gradiant within the meso cycone, it's now about the micro bursts youd find there, no more way allong the storm front several FFD's tehat get big, LIKE REALLY BIG and last 5minutes and they are popping all the time i think lee orf needs to focus on those if he wants to proof his work and as a very cautious scientist he's being afraid of screwing up the modes he neds funding so he can use better super computers like nasa's VEGA. So he can create a lager and more dynamic domain. because what do we see in EF5 storms? Lots and lots of rain and cold and on different side lightning warmth winds and in front of the dry line it get moisty warm hard to breathe as the pressure drops and BAM, i've talked with him last week about the tornado outbreak and told him all this and he said jeez man, i really need to take a look at the forward flaknk of a storm and i said nonono not just that study multicellular low end storms, lower end storm front AKA squall lines and or indeed multi celular non rotating storm fronts COMPARE THEM and i hope that in yur office or the computer room you prnt out a bunch of paper and SHOUT, ...
EEEEEUUUURAAAAAKAAAAAAAAH, and i i dare if that turned your work upside down or has gotten way ahead ready for the research of TORNADO WARNINGS as early as 24- 48 hour til..... a few hours... to finally start that work yyou get me a green card, and we meet up explain our work and fly to japan and freaking do Japanese karaOKE, the song...... Full version of eurka 7 Opening 4 'Sakura.' or opening 1 and two. and eat ramen, because we like noodles in every way imaginable go on vacation in FIJI or indonesia or something may northern australia, and meet another meteorologist that does the same work u do just without chips and computational fluid dynamics (what a shame, but she holds the key) and get a weeks off vacation and get me a lawyer as i am on the brink of suicide right now as an unfair case is popping up. Hehehehe, that's what i'm telling you lee orf, awaiting your reply BTW these two cents are worth allot more AS WE HAVE SEEN AN F3 TORNADO IN THE NETHLANDS IM NOT FOND OF THE eNHANCED SCALE IT WAS AN ef1 and it wasn't storm associated but definitely culombinimbus atmospheric 26k feet at most but sculpting and rotating due to insane wind shears and we have had allot of those also on land and on land coming waterspouts turning into tornado's. not rare but the rarity lied within the amount and duration strength overall, we have constant dry lines warm moist air filled with rain, cloud heating, and cold dry air, daytime ground heating, in a sea climate, not rare but the amount of low pressure and high pressure super localized coming in and over ,arch winters,, hail and warm temperatures for the time and very cold nights a typical battle for low and high pressure but oscillating in a very high frequency, creating the most insane atmosphere for including winds and Jetstream's, completely in shear. look up the northern netherlands Friesland, (stations (EHLW, EHGG, EHKD, check the northern islands, seriously those are separated by just sludge and a meter of water, creating sprout RIPE WARM POOL conditions and, what is the weather like the past month or 2 we have had lots of spouts. associated with rain, often in the evening. and morning even., can u research this i think along with supercellular microburst or plain down drafts coming with rain and winds (vertical downwards. just get sucked right back up in the US creating these monster F6's and EF5's or el reno.... that was allot of air that had to be balanced out in a short time aka storms bursts, micro/macro bursts straight line winds, the forming of the cold pool tell me how does the pressure perturbation get so strong anyway. we both know but is a consequence of air masses needing to swap position eventually though a narrow gap, and the quickest way to empty a bottle is with a whirlpool, when a spin in a supercell initatesthat or windshear of all sized ef0 to ef5,thats the quast way, Oh an popcorn flying up in the air in [movie: 'supercell'] 60m/s will do that and not just with popcorn right :) u are eating popcorn then and people seeing that scene look up above them hehehe i mean It as in ITALY, here are unique formation of water sprouts the netherlands where i live the north sea sees more tornado's that where ever the F-work in the fu**ing world F-0's water sprouts everyday there' at least on average 30 you can always see them, the water is cold so it wont condensate and is associated with just an experience of a prolonged unidirectional wind gusts, but that where you catch the most fish underneath one, just not on fully developed and dark cloudy condensation funnel f1 ef0's woth strengthangth and updrafts enough to lift a human body or even capsize fishing boats with wide and EF1 F2 ones.
Final thoughts, i see allot of you in me, and once were working on something we forget the greater picture in tunnel vision, NOW i am not accusing u of that, but u need a real hardware experts MULTIPLE with YOU in charge to learn your ways and help u, now i cant program at all a little xhtml and barely some python scripts but its not my thing, hardware configuration, c
I think all kinds of storms needs to be researched, and also multiple types of cells and storm fronts/squall lines. and MicroBurst Computational fluid dynamics just to confirm or debunk their effect or find things. We haven't thought about, and there's no trigger? maybe the strengthening of the cold pool and 'hot pool'
I do not claim to have a PHD in meteorology all though i have a bachelor's degree of it, and i used to be an airline pilot until i got sick.
IMMA MAIL YA Dr. Orf. hands down best researcher ever.
on top of that in a FL600+ overtopping cell with rain being dumped upon in FFD's, where does that extra Cold come from hey? that's right WIND ALOFT and the cold reached of SPACE, like how hurricane's are being sustained, i suggest compare em too.
Very well articulated model of an enormously complex process. I think it would be fascinating
and instructive to link the simulations to hodographs that would be likely associated with
various supercell intensities and morphology. I realize that's a tall order, but it poses
considerable useful forecasting potential. Same could be said for the thermodynamic profiles.
That's what we're working on, lots of simulations in different environments - just at say 75 meters, not 10 meters. It's a fun slog.
@@LeighOrfsThunderstormResearch Yeah, I can appreciate how the lower resolutions could still shed light on why certain shearing environments produce splitting storms, or supercells that are not efficient tornado producers compared to those that are. I'll anxiously wait for the simulations.
Mr. Orf, I'm currently at a crossroads In my life but I think I want to go back to school and do science because that's what I love , but I'm not sure where to specialize. I was studying pharmacy before, I kinda want to study plants but I'm really interested in weather and tornadoes, always have been, and I love this new generation of computer modeling storms! I would love to do stuff like analyze models, and uncover mysteries in tornadoes.
What would I want to study to go in this direction?
Take a look at a typical undergraduate meteorology degree curriculum. 3 semesters of calculus, linear algebra, differential equations, 2 semesters of physics, 1 semester of basic chemistry. And of course all of the meteorology classes that are sequenced to work with the prereqs. It's been a while since I taught undergrad but I'm pretty sure that about covers it. If you are into computers, excellent, the more programming you know the better. One thing I have painfully realized is I am no 'true' software engineer, I'm a scientist who does a lot of scientific programming. Good luck.
amazing
Kudos to Dorothy 👍
Leigh, I understand how air at ground level is drawn up into the super cell thunderstorn. Again, what is causing this air to turn into these many,spinning, individual, vortices at ground level? Your narration in the video stated that these vortices seemed to form out of the ground. Can this be explained in plain language?
(Repeating my answer given on another video):
You get a lot of those near the boundary between the cold pool and the ambient inflow/environment. It is typically a region of large amounts of wind shear. Shear lines can easily "wrap up" into individual vortices. But there are other things going on - the cold pool itself is a region of extreme turbulence, swarming with 3D vorticity, which can then be consolidated and amplified. It's really how the tornadoes form in these simulations - by consolidating and amplifying "vorticity of opportunity".
I have an idea, but I assume with such massive data and computational requirements this would need heavy funding; nonetheless, here is my idea: what if you 1) ran the empirical model that you currently use for some specific supercells thunderstorm + tornadoes and save all of the 3d rendering outputs; 2) make sure you have the associated real life footage of the same storms. 3)Then: train a neural network based on the real life footage (i.e. use generative AI model); 4) train another neural network based on the 3d renderings from your model output; 5) train another AI model, lets call it a correction model: this model will essentially use the weights from the other two models, and compare them to ground truth (real life footage) and to each other, and make a correction in the data; 6) now you can run your emprical model, and feed it into the AI correction model - in doing this, you can incorporate real-life Computer Vision data + non CV data collected + 3D rendering generated from the non CV data, and get a 3D rendering that more closely matches the real-life 3D output. Maybe that will give some more insight and better visualizations? Maybe some of what I said can be adapted in another way? Anyways, that is my general idea. Btw, I am a CIS senior at the University of Michigan-Dearborn, and I love CS, and I love Tornadoes since I was 7yo!
I think you are the person to see what this might be like!
Really cool
To 5m resolution and beyond!
what model did you use? and is compiling it required? if it is then is there a precompiled build there somewhere? im pretty much a first timer in programming so compiling stuff gets complicated really quick
www2.mmm.ucar.edu/people/bryan/cm1/
Jesus, my brain is overheating! Great video!
I hope you took good care of my Gavril D series, not a scratch, not a dent. I'll be picking it up on friday.
@@lociflow6154 Well... are you fine with my old Covet? Your Gravril D-Series looks like an accordion now.
@@markischermaximilian Open the door smithy! I can hear you in there, hiding the remains of my car. I'm not leaving!
@@lociflow6154 I'm afraid not! You can get the rearview mirror, that's the only thing without a dent or a scratch.
@@markischermaximilian well, after 2 hours of waiting on your porch I'm starting to give up, but please tell me you still got the key? I got Gabester's autograph all over it... And I'll take that mirror as well, i can at least put the mirror on that shity covet of yours.
What is the pressure doing (relative to the genesis phase) near the the end of the maintenance phase when the sub vortices begin dancing around the inner vortex?
The pressure of the tornado during the phase you commented on would have many individual minimums corresponding to each of the individual vortices. As the parent vortex expands during the maintenance phase, there is a general trend where the overall pressure of the main vortex is probably decreasing and is not as strong as the earlier phases. I may be wrong and this may be for only wind velocity maximas. Additionally, the one vortex that forms inside the parent tornado that dances around for long periods of time is something I am not sure about but looks like torture to me as it has to be investigated for future research papers.
Crazy Erich The Scientist
Yes but I’m wondering about the overall pressure in the mesocyclone above in correlation to the sub vortices. I understand each would have there own pressure which may vary between them. I did a little experiment and I see sub vortices fire up in it but the don’t last and they are in the beginning phase. From what I can see faster moving air comes into the vortex and gets whipped around into a sub vortex it momentarily.
@@FishersofTheRemnant You might need to clarify a little more but I found that the type of tornado produced by the mesocyclone and its parent storm doesn't really determine if multiple vortices are produced are not. What I found is the low level updraft has a lot to do with the structure of the tornado as it modifies the most important part of the tornado that determines its structure, the ground boundary layer. The inflow into the tornado must remain "shallow" (exceptions though) and the low level updraft be strong enough to stretch the vortex, but not too much as it turns the tornado into a turbulent single celled vortex (two celled depending on where you are). You may know a lot more than I do since I am still learning (started a year ago).
Sorry for unstructured post. I lost all my brain power from frustration due to physical modeling.
Crazy Erich The Scientist
I actually did a little silly lab experiment by creating an area of low pressure above the ground. On the left side only it produces a vortex that forms and hold together momentarily and then falls apart, repeats over and over again. In these cycles it produces sub vortices, a horizontal vortex that rides up the funnel, an octopus arm and even a satellite vortex. Since the ground surface layer of air is visible with dry ice fog you can see what it’s doing when these features form.
@@FishersofTheRemnant Can you record this please? I am very interested. I stopped physically simulating tornadoes after this.
imgur.com/a/Jbjsvda
The first image is my first prototype, the rest are improved but still have "frustrating" problems.
Wow, think of how quantum computers are going to ramp up these sims in the coming decades, and think of the data volume and throughput headaches...
good job! +1like +1follow from France
Did you ever cover the rear inflow jet in your research or did i miss out on it? Would be interested in what it does in the simulations
"Rear inflow jets" are not really identified within supercells, but within MCS/bow echoes.
@@LeighOrfsThunderstormResearch I think the commenter ment the inflow feature caused by the surface interaction of the tornado. I am not entirely sure though.
5:20 looks like the Andover Tornado
Voxel simulations, aww yeah
How much data was accumulated in total uncompressed?
We average something like 10:1 compression ratios, so 2 PB or so? And I could easily "bump up those numbers" but the point was to compress it as much as we could to get the most mileage out of what disk space we have. Also consider compression speeds things up (it's not just about saving disk space).
@@LeighOrfsThunderstormResearch woah, that's a lot of data. Did you have to create your own visualization software for the animations?
@@xeonthemechdragon Thankfully, no. There are three robust visualization programs I use regularly: VAPOR (NCAR), VisIt (LLNL), and Paraview (Kitware). I create files from my raw model data that these codes can read - thankfully there are standards so I only have to generate one type of file (NetCDF). It's all open-source (free) and continuously developed. I am very, very grateful for all those viz coding folks out there who make my data look good :^D
Take the sky imagining that being a body of water, you got cold air wants to fall when you have enough hot air rising in a certain area that cold air has to go somewhere and that’s what creates a downdraft And just like a body of water draining threw a pipe it will start twisting.
A tornado is nothing but hot air rising around a column of cold air falling around each other creating a twisting affect.
Dust devils are not very strong because it doesn’t have moisture (water droplets) to intensify the moving of air down around it.
Air does not move air well but water droplets intensifies the whole situation.
Water it’s self displace oxygen go into a cold room take hot water at least 180° and just let it run down from the shower and watch the small water bubbles Form in a room make it difficult to breathe that is a low air pressor Replaced with water droplets that when cold water Droplets can fall at much greater speed then terminal velocity ( The speed at something will fall with a resistance of air slowing it down ) There’s no air there to interfere with The cold water droplets falling
Allowing hot air to rise violently trying to replace air the cold water droplets left as a vacuum in the space above with hot air wanting to rise just intensifies it.
Sorry, tough read, but what I could understand seemed logical. Sounds like your head kind of where mine is, maybe the wind starts the storm rotation but the oxygen content/temps/pressure/rain actually kick off the tornado development from there.
G L I have a tendency to text without using my periods maybe this’ll make it better to understand
G L Take the sky imagining that being a body of water, you got cold air wants to fall when you have enough hot air rising in a certain area that cold air has to go somewhere and that’s what creates a downdraft And just like a body of water draining threw a pipe it will start twisting.
A tornado is nothing but hot air rising around a column of cold air falling around each other creating a twisting affect.
Dust devils are not very strong because it doesn’t have moisture (water droplets) to intensify the moving of air down around it.
Air does not move air well but water droplets intensifies the whole situation
Water it’s self displace oxygen go into a cold room take hot water at least 180° and just let it run down from the shower and watch the small water bubbles Form in a room make it difficult to breathe that is a low air pressor Replaced with water droplets that when cold water Droplets can fall at much greater speed then terminal velocity ( The speed at something will fall with a resistance of air slowing it down ) There’s no air there to interfere with The cold water droplets falling
Allowing hot air to rise violently trying to replace air the cold water droplets left as a vacuum in the space above with hot air wanting to rise just intensifies it.
@@Robert468us I dont believe it's that simple. I believe it actually wants cold air. I feel like if it dumped too much of one or the other its development and structural integrity (cell as a whole not just the nader) would begin to weaken. Maybe that is why they go through cycles; they lose too much of one needed element and have refuel/rebuild. I truly believe these things FEED on environmental factors like we as humans do. We could intake an excess of one nutrient and a lack of another and have issues. But if we intake everything we should, we develop very well and are finely tuned beings. I'm not saying a storm is breathing and thinking, but I do believe once its kicked off, it will get what it needs from the atmosphere around it. Some have RFDs and some dont. Maybe that's a way of purging unneeded energy and unspent "fuel?"
An RFD could be a purge valve or vacuum leak! Haha
any new videos, what's going on?
Hangin' out in my basement staring at computer screens LOL! I hope to post new material soon...
@@LeighOrfsThunderstormResearch I understand. Not all your work results in pretty pictures :) I will be patient
@@pratwurschtgulasch6662 If you go to my web page I have a link to a recent podcast interview (Chaser Chat podcast). No cool videos though!
All my life I have been fascinated by tornadoes I did not have fancy labs But I did study them closely here what happen hot and cold make water or moisture moist air rising moves a lot of air just like a fire hydrant hose moves air when it’s put on a spray The water bubbles make more air move and than just air it self moving air displacing cold air from the inside at the super cell When you have a balance of hot air that can only go so high creating water droplets displacing cold air wonting to fall with cold water droplets when that cold air is dropping that’s where the twisting starts coming into the system Just like water draining from a reservoir start turning developing your tornadoes To stop or prevent a tornado the warm moist air feeding the super sell from the side in flow band Tail Needs to be replaced with cold air that prevents the hot air from displacing cold air and making cold air drop Creating tornadoes Hot and cold air does not mix together well cold air wants to drop hot air wants to rise when the to meet each other they don’t have time to equalize what better way to start the transfer is spinning just like water draining through a pipe from a reservoir always start spinning when there is a down force being made if there’s someway in advance of the storm the hot air could be replaced with cold air the tornado would go away that stops the flow of hot air rising and cold air falling What creates the tornado is cold air falling from high above -30 to 40° wanting to fall just as fast as hot air is rising meeting air 80 or 90 or 100° rising Neither do away with the cold air dropping or hot air rising and you will not have a tornado
I don’t have a lab but I do have a local pond with a pipe that creates small tornadoes I put an object down towards the bottom of a tornado is impossible to stop but you can stop the down drafts at the top of the tornado it takes very little trouble to stop a tornado in the water just simply using draining water pipe I did experiments and put something in its way at the top to stop the twisting effect Put the same object at the bottom of the twisting water tornado it does very little to affect it same thing here some super cell does not have enough warm energy air rising to create a tornado when some have enough warm air rising creating moisture displacing cold air does create a tornados at the very top of a tornado developing having some sort of a gun that could shoot fire directly at the top of the tornado where it is Forming would disrupt the airflow down and would simply stop the tornado Just like a small object can stop the downdraft and stop the tornado in a water simulated
This is some fascinating and amazing work! I am mesmerized. I've been negative toward modeling, specifically as I'm climate models because they have been so bad and..it seems like the garbage in factor is often politically and ideology driven. But, I can't see any reason to apply that to your work as you are trying to understand tornados and save lives.
I could watch this simulations all day!
Suggestion... to subsidized your work, have you considered using these models for phone and or even Pc screen savers? May not seem like its worth the trouble but...find some young coder at school, make a mobile app and sell for a couple bucks...could make some bank! Would love this to be on my home screen or on my laptop. I would pay $5 for that. 😋😁. Something to think about. Really are beautiful.
Yeah Tornado Is Insible
Maybe it’s time to consider the electric model of a tornado to see if there is a better way to predict and forecast them.
“Now let’s consider the most dramatic weather event of all, the tornado, and how these massively destructive whirlwinds are also formed by a plasma corona in a thunderstorm.
For air to become plasma and carry current, the air has to be partially ionized. A plasma state can be defined by “plasma density” - the number of free electrons per unit volume, and the “degree of ionization” - the proportion of atoms ionized by loss, or gain of an electron.
A gas with as little as 1% of the particles ionized is a plasma, responding to magnetic fields and displaying high electrical conductivity. A partially ionized plasma is often referred to as a cold plasma and highly ionized plasma is referred to as hot. Discharge from a corona is predominately a cold, dark current, invisible to the eye.
Cloud-to-ground arcs come from high charge density regions of the corona, primarily surrounding the central updraft where current from the updraft generates ions. Ground charge builds below this region in response, and the electric field strengthens, magnifying and focusing electron avalanche the way a lens focuses light, into a continuous plasma channel. When the channel connects with ground and discharges a hot current, it wraps tightly in its own magnetic field, in what is called a Z-pinch.”
www.thunderbolts.info/wp/2017/06/13/tornado-the-electric-model/
Greg Hummel Bravo!
That's the outside the box thinking we need to bust us all out of trying to just confirm prior beliefs, and bring us back to the true scientific method. PROVING sonething before we call it fact instead up saying, "well, it SEEMS like the most logical POSSIBILITY." Words like possible and seem only belong in the early phases of the scientific method when we hypothesize and begin down the path.
I have an issue with that theory as I think it's not that important. In the two tornadoes I have been in (one being the canton Tx ef4 or ef4 , can't remember; the one that hit the dodge dealership was the one I was in), I felt absolutely no electric buzzy feeling. All I felt was the pain from my ears popping and the fear that my car is going to become an airborne torpedo. It may be something to look into though. Still open to ideas.
Look I’m not a scientist but I have a weird intuition and tornados intrigue me...... I’ve been doing tons of research and while I’ve gone through all your videos, you’ve failed to mention one very important tornado...... it was the jarrell tornado in Texas back in the 90’s...... it was perhaps one of the oddest phenomenons the world has ever seen. It didn’t leave a path of destruction (so to say) because it was so powerful it left everything to ashes. (Now I have a confession to make it may have not been the jarrell tornado some of my saved research YT has taken down but I do know it was a Texas tornado and it happened in the late 90’s) what was strange is that the conditions this day was great, nobody thought a tornado could form. (Man I wish YT didn’t take down this video so I could have linked it down below) the only thing was that it was very hot and humid...... I do believe it was the jarrell tornado please look into this and do some research......
Here are some very unanswered questions....... 1. why is the dry line moving East? Given our worlds past, (I don’t really believe in global warming to the extent the media propagates this) one would assume tornadoes would have been prevalent in other locations (maybe millions of years ago).
2. Why do people believe the funnel forms from the sky? My intuition tells me the ground has something to do with the creation
3. Why do some form backwards? And what drives their location? They have no rules lol.... they can take a hard left turn, go backwards, stay in place, etc.
4. Why can they be weak one moment then create momentum at other times?
5. Why do tornadoes form in the dry season during winter time?
My biggest question. I live in the northern pan handle of WV. And we have never ever had a tornado! And yet everywhere around us, even 10 minutes away, have had some and yet wheeling WV remains tornado free. Why? Makes no sense.
Edit: I think I found the tornado! It was jarrel.
czcams.com/video/8hrhxlAXkNo/video.html
This is the link to the tornado but I’m still trying to find the weather conditions of this day.
Making sure you’ve seen this... czcams.com/video/iOOQaE0WR2I/video.html
Indeed I have, and it's tremendous!!
This is cool but it doesn't get much across to the layman.. im sure its created for your peers but you talk so fast and use lots of jargon that most people don't understand. I think it would be helpful if you had a slower video explaining some things to get regular people interested in pursuing this field. Just a thought
Yeah, these are just "straight" scientific talks. There are other videos aimed at a broader audience here: orf.media/education
@@LeighOrfsThunderstormResearch oh wow, awesome thank you man. Keep up the good work!
@@LeighOrfsThunderstormResearch Thank you! Just shared that on Twitter.