Tim Fernholtz
speaker
84 appearances
2 recordings
2 series
first heard Feb 2026
last heard 13 Feb
Tim Fernholtz’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 2 in all, peaking in Feb 2026 with 2.
Appearances
Marketplace All-in-One · Inside the "biggest deregulatory action in U.S. history" · 13 Feb 2026
podcast
So it's an ongoing conversation between all sort of elements of the weather system.
It's lovely to be here.
Thanks for having me.
So I won't say traditionally because actually it's been an expanding field for the last few decades.
But in contemporary times, we basically have a government-run system that collects data from all kinds of sensors all over the world, weather balloons, buoys.
Every commercial plane in the United States coming down is sharing its measurements with us.
the National Weather Service and the National Weather Service has a huge supercomputer and it plugs all of that information into it and it goes through these scientific models that meteorologists have spent generations building and spits out a forecast and it's a very complex and expensive process and it's been refined and refined and now we have pretty good weather forecasts out say 10 days.
So now we have deep learning software models where we can feed data from the weather agencies into them and ask them to do forecasts.
And it turns out they're actually able to predict the weather better than the traditional way we have done it.
And then in particular, these models are very good at doing things like tracking hurricanes, tracking cold fronts that are trickier for the traditional models to do.
And so scientists are trying to figure out why exactly they can do this, because you have to sort of understand why it works in order to implement it in a useful way.
Well, you hear two approaches to that.
One is if it works, it works and we should use it.
But what is interesting is that if we understand how it works, we can use it to learn more about the atmosphere itself.
These technologies are very new.
Only in 2022 did we first start to see these models coming out.
And they are an opportunity to learn.
And like a lot of artificial intelligence models, they create a very complex way of managing data that's sort of hard to peer into at first.
And just like with the LLMs and other forms of this technology, there's a lot to learn about what actually is happening.
And then can we take that and use that to learn more about science?
Showing 1–20 of 84 · page 1 of 5
Next →