12-11-24 Interview - Weather Wednesday with FOX31's Dave Fraser - podcast episode cover

12-11-24 Interview - Weather Wednesday with FOX31's Dave Fraser

Dec 11, 202410 min
--:--
--:--
Download Metacast podcast app
Listen to this episode in Metacast mobile app
Don't just listen to podcasts. Learn from them with transcripts, summaries, and chapters for every episode. Skim, search, and bookmark insights. Learn more

Episode description

GET YOUR WEATHER QUESTIONS READY Because Fox 31's Dave Fraser pops on at 12:30 to answer them all. I've got one about the new weather forecasting AI that is probably going to be game changer. Read more about that here.

Transcript

Speaker 1

Hello, Dave.

Speaker 2

It's a nice day to day after yesterday's poopy weather.

Speaker 3

Monday and Tuesday were tough. That that storm really hit the roads hard. But yeah, today it's nice. I'm enjoying the sunshine, very very nice.

Speaker 2

What do we have coming up here as we head to the actual You know, we've got fourteen days until Christmas. I know we're not predicting the weather out that far, but is there anything you like out in the out on the Pacific or the Pacific Northwest that is going to give us a chance for snowy Christmas even though we don't normally get that.

Speaker 3

No, And I knew we would get to that question. So I looked at our long range computer models while I was waiting, and I can't actually see out ten to fourteen days. Yesterday when I was looking, there was one of the long range computer models was really robust on a storm on the twenty fourth, and I was like, oh, all right, we'll keep an eye on this. Well today, when I looked at that same long range outlook, it's not there anymore. So for right now, like you said,

we're fourteen days out. The seven to ten day is dry. We might get a rain or snow shallow come Monday. Mountains will get another inch to three inches of snow late Thursday Friday, But the overall pattern, there's just nothing out there that gets me excited about getting some snow in here at time for Christmas.

Speaker 2

How are we doing with the snow in the mountains, because I know they've been getting a lot more snow than we've been getting on the Front range, but I mean, how are we doing at a snowpack goes We're fantastic.

Speaker 3

We're over one hundred percent. Most of the basins are at more than one hundred percent. Wow. The deepest snow basins we have is in the southern mountains and southeast Colorado is doing really well. Where they're below average a little bit like eighty five to ninety percent is the northwest corner of the state. They could use a little more.

But the Front range is doing well, and the total state, I think the last time might look last Thursday, we're at about one hundred and thirteen percent of average, which is good this early in the season. Remember, we really front load the snow water usage come the spring, so to be ahead this early in the seas is fantastic.

Speaker 2

That's what I was about to say, you know, we have so many new people because I remember when I first got here and people were talking about snowpack levels, and I'm like, why do we care? Why is that important? Especially if you're not a skier. But it is all about our water. So that is why we talk about that specifically. Now, Dave, I got a question for you.

So a listener said it a super cool story, like super cool story about gen cast, which is a new AI model that they say predicts weather, delivering faster, more accurate forecasts up.

Speaker 1

To fifteen days ahead.

Speaker 2

Plus it's more accurate on determining the path of hurricanes, which is huge. I mean, that's huge. Is this the kind of stuff I mean in your industry? How much are you talking about adopting these AI tools? And how long do you think it will be before something like this becomes a part of Fox thirty one's weather forecasting.

Speaker 3

Well, it already is, and really, if you think about it, forecasting as a whole is more and more computer models were being folded into the process going back decades. That is really AI at its beginning, because we were using computer models, mathematical algorithms to kind of predict what the

atmosphere would look like in the future. So kind of, if you think about it, all of the computer models that we've been using for decades are in fact AI, and currently all of the broadcasters in town and across the country, for the most part, use the same set of the same company. It's called the Weather Company and it's owned by IBM, and they have their own AI computer model do we actually show on the air every night.

It's beyond some of the other ones. I'm sure if you're a weather geep you've heard us use terms like the euro model, the cfs her, the NAM. We look at all of these models, which are futuristic, they are AI. They're trying to tell you what the future is going to look like, and we kind of we know some

of them have biases. Some of them or two pass too slow, too cold, too warm, and so we know what those biases are, and that's where our understanding of the atmosphere and making a forecast comes in to be able to weed out those biases. So we have one it's called the graph, and that's actually what you see me for our use on the air when I'm doing future casts and showing you where the snow is what time it arrives, how much it might get. It's all

of that kind of blended the gen cast. It's interesting gen cast is looking more at bigger events so that emergency managers and states can brace and kind of hopefully reduce the risks on the costs of destructive weather. However, here's the funny thing. Gencast is run off of the European model. Right. Without the European model, gen casts can't stand alone learning. It's learning, but it's not there yet.

Speaker 2

I just find all of this super fascinating, and it's like, you know, I was talking about, I would love for this to be you know, adopted in if we if we just get more weather forecasts, and especially for hurricanes.

Speaker 1

Because to your point that you just made, if you could.

Speaker 2

Direct all of your resources to exactly the right space, right so you don't have people all up and down the coast of Florida buying plywood and supplies, and if you had, if you could clarify that with some you know, instead of the cone of uncertainty that you get, now, that could be game changing.

Speaker 3

That's one hundred percent, one hundred percent. And I even look at it more so in the day to day forecasting, so you're looking at something there that's a larger event, if you will, something where let's just use your home state up Florida for instance, where where the governor and the emergency managers and all the task force are sitting down and they're looking at these updated, high resolution AI generated images to say, Okay, this is where we need

to put our resources and be prepared because it looks like this is the bullseye. I would love to see it go even further and be able to be used on a day to day more localized level or forecasting. Say here along the front range, I mean that storm the other night. We always mandy get two or three snowstorms a season that either overperform or underperform. That storm we had Monday night at Tuesday overperformed in the amount of snow it's delivered, especially on the southwest side where

places like ken Pale got six inches of snow. If we can get AI to kind of be more pinpointed in that direction, that may help us a little bit in getting those more finite details out. You're never going to get a perfect snow forecast, because I've said before it's a painting. Right you stand back and you look at it, you can see, oh, that shade of blue is this much snow? This deeper blue, is this much snow.

But when you zoom in, it's a mosaic, and there's going to be little squares and little communities that are always going to be a little higher a little lower. But if we can get that extra help, two brains are better than one.

Speaker 2

Yep, speaking of two brains, A lot of people on the text line are asking if you have any thoughts about Mike Nelson retiring tomorrow.

Speaker 1

I mean, you're part of the same fraternity.

Speaker 3

I do, Mike and I and I'm glad you brought that up. His last day is tomorrow. His last show will be his late evening show, the ten o'clock over on Denver seven. They had a nice going away celebration for Mike last Wednesday. I attended the log with a lot of the other broadcast meteorologists and of course all the anchors and the behind the scenes folks that have worked with them. A very nice presentation. Governor gave him a proclamation for twelve twelve. He picked twelve twelve four

twenty four. I'll tell you the story. He picked twelve twelve twenty four, because if you have that up, it's forty eight. That's how long Mike has been a broadcaster on LEA. Thirty eight of those yeah, thirty eight of those years have been here in Denver, split between Channel nine and Channel seven. And I'll just tell you something. Mike has been a leader in our industry. He was out there selling the first color graphics computers on TV. Prior to that, it was all the magnets and the

dry boards, and Mike was a big beginning. Yeah, you should see go watch, Go watch seven tomorrow night because they'll have some great video. I've seen it over the years. And Mike was the first one to call me when I rolled into town almost twenty five years ago. Within the first week I was here, he picked up the phone,

introduced himself. He says, listen, I think you do a fantastic job, and I'm happy to have you in town so that we as a collective group, even though we're competitors and on different stations, can get the message out about what the forecast is going to be. So I I've known Mike for a long time. His wife Cindy will see each other after he retires. But best wishes to mic and it's been a long, stellar career. He's well deserved his time off.

Speaker 2

I got one last weather question for you, and that is from Noco Dan for Dave Frasier. Our lenticular spaceship clouds more common along the Front Range recently. I don't remember even seeing them until the last five years or so.

Speaker 3

Yeah, it just depends on the weather pattern there. We do better with lenticular clouds, and you need too ingreedy. You need a westerly wind coming up and over the mountains so that the air is kind of curling like the curvature of a lens, and you need moisture. Sometimes we can have wind events, but it's dry and you're

not going to see the lenticular. So what you need is the right ingredients, the perfect soufflet of the right speed, the right curvature of those that wind coming from west to east over the mountains, and then a nice moisture flow in there to kind of condense the cloud to get the lenticulous. We haven't had a lot of those this year. You know. One of the things that we've been talking about is, I don't know if you've noticed it, the leaves hung on the trees a longer this year.

Speaker 1

They sure did.

Speaker 3

They think, Yeah, I'm no arborous, but we think it's because we just haven't had a lot of wind so far this fall and winter. So I think that's been part of the component, and that's probably why we haven't seen a lot of articulars so far this year either.

Speaker 1

I that's Dave Frasier, Fox thirty one. You can always see him.

Speaker 2

That is an incredibly stunningly accurate forecast over there.

Speaker 1

We'll talk to you next week. My friend

Transcript source: Provided by creator in RSS feed: download file
For the best experience, listen in Metacast app for iOS or Android