Grace Cohen 🌪️

This would be a really good wind profile had there been instability

@GBAMowerRescue This was pre-calibration. See my quote rt for the post-calibration version.

@GraceCWeather How was the og super outbreak so low????

More data will be needed to help get the probabilities where they need to be, but things are looking promising!

@simbasad2 oh shoot i forgot that one

@GraceCWeather you also need to add a disclaimer that it is 2026 bc what if the public thinks its 1974 🧠🧠🧠

Following the recent comments I guess I need to start posting these like this

@TheChrisBarrows @AriWeather @aguacerowx Its great, we’ve had these vertical wind profiles already here in japan, but now having it for the US is wonderful.

@AriWeather @GraceCWeather @aguacerowx The cross section functionality won me over, finally a modeling application that you can see a deeper look into the atmosphere.

I have used @aguacerowx pro for one month and here are my thoughts. 1. can actually use it for here in JP w/o being hard to see. 2. Tons of customizability 3. Good model selection I rarely spend money on subscriptions but this one might be the one to change me.

@weatherlord_wx ML, a lot of data, and a good PC

@GraceCWeather how do you build a model like this?

Actually, calibration might have just solved the issue.

@wxKobold Who knows, maybe they are time travelers from April 2 1974

@GraceCWeather Good grief you weren't lying about the comments, haha.

@FredCun93117348 good thing they didnt see the calibrated one LMAO

@GraceCWeather yea I thought the 10% was too widespread for recent memory so I started going "hmmm". Looked at the margins for extra details and 1974, April 3rd and it clicked in my head "yea that makes sense" lol

@WxGabe thank you!

It is! This is a phase diagram. The various phases are labeled 1-8, and correspond to lines of longitude in the equatorial space. The further out you go from the origin, the stronger the MJO. The dots show the various ensemble predictions for strength and position of the MJO lift max. Colors correspond to days from initialization.

Global models agree that severe weather is unlikely through the end of March. But as others have noted, the beginning of April is looking quite interesting. The Euro is forecasting a move into the “tornado hot zone” (Phases 1 & 2) of the Madden-Julian Oscillation. We shall see!

@FredCun93117348 It is twitter though, people are going to misinterpret everything unless it is painted out in big bold font lol

@GraceCWeather Damn. I mean, all they had to do was look up...

@ChrisGensch Chris, please check the date at the top of the image.

@GraceCWeather We in tornado alley do not see this as promising.

@wakeupnowpeeps This is not a current forecast, it is a test of my machine learning forecast. Please check the date at the top of the image before you comment.

@GraceCWeather "Promising" for what? Destruction of property & resources, lives torn apart, lives lost....come on, at least remove your exclamation mark......

@itz_aj_crxsh This is a test using the 1974 super outbreak data of my machine learning forecast that I am developing. It is not a current forecast.

@GraceCWeather is this real?

@Brady_Wx A lot of ERA5 and binary classification RF

@GraceCWeather Cool! I am interested in hearing about the architecture & data used

@mack65723 Please check the date. This is promising because it is a Machine Learning forecast I am creating.

@GraceCWeather I wish chasers would start utilizing terms better, because I guarantee the people in those areas don't think of it as "promising"

@Cdunlapweather Yes once I finish training and adding CIG’s

@GraceCWeather Will this become public?

Did one last high end test before I go to sleep.

Mulling over ROC-AUC values….I don’t know if it is good enough yet. It’s been sitting right around 0.95 but I really want to push for 0.98 or so…i need to sleep its 2am

@ScottDimmich Thank you! Once I feel they are ready to be used I plan on creating a webpage for them. I want to make it a resource for people to use.

@GraceCWeather Nice plots. Publicly available?

@CameronJNixon @backinblack_wx @wxkylegillett All credit to Chris on this one. I mainly handle the stuff once it becomes 3D.

@backinblack_wx @GraceCWeather @wxkylegillett Excellent analysis y'all. You definitely found something Trey and I missed, and he's working to include this in the analysis. The main question I have with this kind of thing though is how to communicate it. I personally have a problem calling this a "nudger" (even if it nudged)

The Kankakee, IL tornado radar analysis is now uploaded to the tornadogenesis page. A merger-drc seemed to occur! @CameronJNixon @wxkylegillett tornadogenesis.org/radar-analysis…