Modernization of radar imagery prior to modern dual pol

[UncleJuJu98]

So I've been mulling and sitting on this idea for about a year or two. Discussed it with @ColdFront and @WeathermanLeprechaun. I thought of this cool idea or a project of modernizing dual pol radar imagery prior to its latest modernization in the early 2010s.

And I have used AI to try and recreate the oak Grove f5. If anybody wants to try it or create some template to modernize.

 

[UncleJuJu98]

Theres a lot more smarter people on this forum than me that can probably bring this vision more to life than I can lol.


Alabama palm Sunday is one that I would love to see a modernization of.

Palm Sunday Tornado Outbreak 3/27/1994 Weather Data

 

[WeathermanLeprechaun]

Honestly while it's a interesting idea, AI is quite literally just using a basic supercell and kinda imitating it off that. You're not gonna get much any different results as such with it. It will most likely generate the same looking supercell, velocity, CC and ZDR.

 

[UncleJuJu98]

Im going to try and plug in this velocity couplet from the f4 on palm Sunday here in Alabama. Later when the refresh allows me.

 

[ThatGuy;P]

This video includes both simulated reflectivity and velocities of fictional tornadoes, perhaps something similar can be done, but modeled off of exisitng radar data of previous events.

 

[WeathermanLeprechaun]

This video includes both simulated reflectivity and velocities of fictional tornadoes, perhaps something similar can be done, but modeled off of exisitng radar data of previous events.

This was done based off WRF. it's possible but it takes hours and hours to actually run these environments

 

[UncleJuJu98]

This is chatgpt best try at the ragland f4 from the 1994 palm Sunday outbreak in Alabama. It's getting closer but no cigar.

 

[UncleJuJu98]

Tg

This video includes both simulated reflectivity and velocities of fictional tornadoes, perhaps something similar can be done, but modeled off of exisitng radar data of previous events.

That's rather interesting honestly

 

[UncleJuJu98]

I dont think a perfect recreation of any velocity couplet prior to modern radar will be possible but I think it could get close enough. For a general idea of how it would look on modern day.

 

[UncleJuJu98]

I think this maybe the best recreation so far. This is the December 2000 Tuscaloosa F4.

 

[UncleJuJu98]

This is a go at the enterprise Alabama ef4 in 2007. I'm curious to if a more premium version of chat gpt would do better in the recreation regards. The created image has too violent of wind velocities, but then again it could be fairly accurate. Because the hook echo and tornado were violent.