So sure, these probabilistic weather forecasts are getting better at whatever they have been trained on to do. But in contrast to a physics-based simulation, I suspect they will not be good at predicting what they have not been trained on. And with climate change, there will be a lot more previously unheard of weather events.
https://www.nature.com/articles/s41586-024-08252-9
Hanno Rein /
npub1rq…r7gnu
2024-12-05 14:12:48
Author Public Key
npub1rqq5rv8tr2yu3cssea82js0mwrms5k3jm9h49d2k6tuhew8w9khsjr7gnuSeen on
wss://relay.nostr.bandPublished at
2024-12-05 14:12:48Event JSON
{
"id": "fe95358660d7d1536ffc663d1f9893be2d1b03af84bce1a9eef35c6effbd9f44",
"pubkey": "180141b0eb1a89c8e210cf4ea941fb70f70a5a32d96f52b556d2f97cb8ee2daf",
"created_at": 1733407968,
"kind": 1,
"tags": [
[
"proxy",
"https://mastodon.social/users/hannorein/statuses/113600624619983983",
"activitypub"
]
],
"content": "So sure, these probabilistic weather forecasts are getting better at whatever they have been trained on to do. But in contrast to a physics-based simulation, I suspect they will not be good at predicting what they have not been trained on. And with climate change, there will be a lot more previously unheard of weather events.\n\nhttps://www.nature.com/articles/s41586-024-08252-9",
"sig": "61f4a3568dd3a6823ab4a588af46274d362d6924fb57bfc6445b0326949c9501d42572fd7a71035c1a9647d72f0149d1cf2d848c99cab0b4177da3ed3f23a483"
}