@GraceCWeather Very cool! I'd love to know some more details. What type of ML algorithm does it use, and what are the datasets its trained on?
English
Daniel Wefer
23 posts
Daniel Wefer
@WeferWX
NIU '26. Storms are cool.
DeKalb, IL Katılım Mayıs 2021
289 Takip Edilen51 Takipçiler
The webpage hosting my ML tornado forecasts is now public! check it out!
cohenwx.pages.dev
English
@CameronJNixon Really cool seeing some of the seasonality of different regions on these soundings.
Dixie alley is all early season HSLC setups, and the upper plains are all summertime modest shear cape bombs
English
Anyway here's a map of all EF2+ tornado soundings for select locations across the U.S.
English
@oldscarf1stweek Here’s the website
nimbus.niu.edu/heatmap2/index…
English
Officially in May Range on the CFS chicken cutlets.
Tornado season so far is cancelled but could return any day.
English
@ConvChronicles In my experience, lid strength, storm mode, and anchoring bias are nearly always to blame for forecast busts.
With how many CAMs we have, it's extremely easy to hyperfocus on solutions that agree with you after the initial forecast is made.
English
Quick breakdown of a few of the meteorological factors that may have allowed today's Moderate Risk severe weather outbreak across the East Coast to underperform.
English
@DCAreaWx IMO the lack of a capping inversion/EML wasn't considered nearly enough prior to this event. With as strong of synoptics as were present, a substantial warm nose would've been needed to prevent overconvection in the warm sector.
English
...along with the potential for lingering convection from a severe threat the day prior, created a legitmate failmode of prematurely killing the already marginal instability. This is something which I need to consider more strongly in the future for improved messaging. (2/2)
English
Hindsight is always 20/20, but today's underwhelming severe weather is a good reminder to me to consider the thermodynamics of an event, even in the face of high-end dynamics.
Strong synoptic ascent driven by an intense upstream trough in an uncapped environment,... (1/2)
English
@a_guidarelli I got stuck south of carbondale because of bad road options💔💔💔. Saw nothing but a wall cloud
English
Marion, Illinois ef4 190. Audible roar from 3 miles away. The most violent thing I have ever seen in my life
Vincent Lilly@VincentWX_1
What is the strongest tornado you have ever seen?
English
@TeamGodsFinger Herbie python package makes it easy.
herbie.readthedocs.io/en/stable/
English
Does anyone know how to get old forecast model runs? Working on a march 6th, 10th video. #wxtwitter
English
@MatthewCappucci From the south it was a visible stovepipe for around a minute before it went fully rain wrapped.
English
Fellow chasers — how long was Kankakee a stovepipe? I was only about a half mile to a mile behind it and had a massive rotating cylinder that I at first thought was a ground-scraping meso, but evidently it was the tornado?
English
@VORTEXJeff This was one of my main reasons to stay with it. Lack of better options and warm + heavily backed winds at the Kankakee ASOS were what kept me east.
English
Leaving the storm would have meant driving home since I wouldn't have reached the western storms until sunset and I *DO NOT* chase after dark. Thus, I likely would have followed the storm to I-57, my quickest way home, and obviously stayed with it afterward. 2/2
English
I couldn't chase Tuesday, but I've been asked several times if I would have hypothetically stayed with the Kankakee storm. I wasn't actively decision making as the storm was ongoing, but my answer is that I likely would have stayed with it. 1/2
English
@WxGabe Agreed. One of my biggest regrets of my life to date is falling behind to collect hail. I was like 7 minutes too late. 🥲
English
Not my best capture, but it was dark. Anyway, here’s the Lake Village, Indiana tornado… though I was just south of it, I had to keep asking myself “that’s the tornado, right?” because of how enormous it was.
I was behind it in Kankakee as it wedged out and it was huge.
MyRadar Weather@MyRadarWX
This is our photo of the Lake Village, Indiana tornado — apparently the same that began in Kankakee, Illinois. It was on the ground for 1 hour 21 minutes, killed two people and tracked at least 36.6 miles.
English
@oldscarf1stweek If there had been a single rotating storm within 30 miles I would’ve 100% bailed. Glad I didn’t
English
This is what Kankakee looked liked 40 minutes before it wedged out.
Surface observations/parameter space will take precedent over the visual cues. What a tricky day for experienced chasers.
Ian Shiery@eagles_fan_05
@GKellerwx some nice structure just west of Reading from that dying supercell!
English
@Sawyer_WX @gensiniwx conduction on the contact layer. Sun bakes the surface and the air touching the surface warms more rapidly than the air above it.
English
@gensiniwx Absurd amounts of cooling just above the surface. What do you think was the cause of this?
English
2030z sounding from yesterday (3/9) near Wenona, IL.
English
@CRyanMauk A big fear of mine is a chaser pulling someone out of wreckage, not realizing they had a spinal issue that needed to be stabilized.
English
I don’t think it’s a storm chasers’ responsibility to stop and sift through wreckage to “save lives“. I think people expecting that they would do such a thing is just as ridiculous as people saying they’re chasing to “save lives”. I think it largely comes from misplaced grief. Which is understandable. But it is shooting the messenger.
English
@CameronJNixon @Get_EMLd @SimonStormRider @VORTEXJeff @JustonStrmRider @AaronRigsbyOSC @JordanHallWX Something I noticed while chasing it was that during the high base/pseudo-elevated stage, it was ingesting a lot of rain-cooled air from its own FFD. I figured the LLJ would help it take a right turn and ingest cleaner air.
English
@Get_EMLd @SimonStormRider @VORTEXJeff @JustonStrmRider @AaronRigsbyOSC @JordanHallWX I have to say what makes me believe Simon's decoupled boundary layer hypothesis is that those HCRs are just a classic, classic feature of tornadic environments, and you can see things change abruptly from Cu's to HCRs just west/east of Kankakee
English
I don't think the Kankakee, Illinois supercell became "elevated" north of the cold Lake Enhanced Boundary between Pontiac and Kankakee: here's why (THREAD)
@JustonStrmRider @AaronRigsbyOSC @JordanHallWX @CameronJNixon #ILwx
English
@ContentWxGuy The natural land is pretty much already gone in IL/IN. All I’ve seen solar farms replace are corn ethanol farms, which are MUCH worse for the environment.
English
Good.
This natural land destroying nonsense must end until a single panel can do what these eyesore fields can.
Maybe nature is sending a message by destroying that trash heap... into... a trash heap.
Jaden Pappenheim@PappenheimWx
ENTIRE SOLAR FARM DESTROYED in Wheatfield, Indiana after a strong tornado struck last night. Unfortunately, multiple house sustained significant damage here as well.
English
I’m sorry I probably tweeted and retweeted 1000 times today
GIF
English
@skydrama I was next to you at the Wally’s! Didn’t end up seeing the Pontiac tor, just saw the wall cloud. Kept following the storm despite its elevated appearance out of a lack of better options. Caught up to it SE of Kankakee and got a nice stovepipe at aroma park
English
I too, observed the Pontiac, IL tornado after getting on the storm at initiation, only to leave it as it wandered over to the cold side of the boundary and back, to resume producing tornadoes from Kankakee into IN.
Will recap tomorrow. Some lightning from round two after dark:
English
@Emokwx It was those SE winds out of kankakee that kept me on it. Looked awful for a while there
English
Large tornado near Aroma Park IL around 18:27. Has since become a rain wrapped wedge @NWSChicago
English