This is the best Monsoon tracker in Arizona

[00:00:02]Hi everyone, I'm meteorologist Royal [music] Norman >> and I'm meteorologist Pete Manion.

[00:00:06]>> This is Sonor and Sky Live, a service of First Alert Weather and Arizona's family. We have kind of a really crazy good show to talk about because we're going to be talking with the modelers from the University of Arizona that model the monsoon and they're legendary.

[00:00:21]They I mean they've been doing for sure they're legendary. Everybody follows their stuff in the state of Arizona and I think this is going to be exciting.

[00:00:29]First of all, just to find out how they put all their products together and just uh how much pressure there is when that forecast is out there. Shall I introduce them?

[00:00:38]>> I'm kind of excited about it. Yeah, go for it.

[00:00:39]>> Uh Patrick Bun and uh Pat Hullbrook are joining us uh today from the University of Arizona. Uh Patrick, I'll just ask you first for your your formal title there with the modeling group.

[00:00:51]>> I'm a research scientist with the powerful casting group. And Pat, your title, please.

[00:00:56]>> I just I'm a meteorologist working here.

[00:00:58]>> Oh, that's fantastic. I love that. Well, okay. So, um, Patrick, tell us a little bit about the history of this. I've I've told you I've been watching this since day one and looking at the models that you all produce. It's been, has it been 20 years yet or is it just shy of >> Yeah, about about 20 years since Mike Luod started the UA Wolf Weather Forecast System. uh and then he was running it and developing it for a number of years and then we've got our first contracts with the regional utility companies. So APS and TP were the first uh to provide some funding and then SRP recently in 2023 they also joined in. So >> I had to jump in a second too. Idaho Power also was providing power funding for it for a while.

[00:01:44]>> Yeah. And so the bottom line is what is your goal um during the monsoon? You you put out a discussion which all of us read. You put out maps which are really automatically rendered but what is your goal every day? What's the workflow like?

[00:02:02]>> So uh we start with the model. So one of the really unique things about it is it's a really high resolution um weather forecast model. Um so it's part of the um weather research and forecasting model that was initially developed by the National Center for Atmospheric Research and then we run a custom configuration of it uh that's specifically designed for um the Arizona weather and climate. Uh and the one of the unique things about it is it runs at a 1.8 km grid spacing. So that's really really small grid spacing that's uh specifically designed to better forecast these small scale thunderstorms that we have particularly in the monsoon season.

[00:02:38]And so we run that weather forecast uh system on a high performance computer that's here at the University of Arizona and that's as you said it's it's automatically running. So every three hours it cranks out a new forecast and new plots that go up onto our website and then every morning we start the kind of weather forecast discussion process.

[00:02:56]So >> and basically it's a top down approach.

[00:02:59]We look at the upper levels. We look at what happened previous day because a lot of days you know like today uh it was pretty up day yesterday. So today should be a down day and but but independent of that um the it's kind of a top down approach. We start up at the higher levels of the atmosphere about 300 millib bars and work our way down just analyzing maps down to that then look at our uh just to get a feel what what's happening in Arizona. And then once we uh once we get that feel we kind of put the forecast together from using a model and um and then trying to forecast out for three days basically. And we're and we're trying to help out the customers because we want to uh look for interesting thing things that may be happening uh that may affect them. Uh high wind gusts, hail, uh that type of things, obviously flooding. So uh that's that's one thing we're trying to do is help the customer out.

[00:03:52]>> And you put together some amazing very detailed discussions looking at the model data, kind of describing the setup for today and then also the forecast.

[00:04:02]How long does that take to put together?

[00:04:04]Because if I had to put that together, I think it would take me hours. It looks like a lot of work.

[00:04:09]>> Yeah. Uh, you know, actually I I I'm looking at it the night before basically because I got to see what happened and uh it basically uh I start about 5 in the morning done by about uh 10:30 that type of thing. So it's it's takes that long to do it. So how many meteorologists are involved in the whole process there from start to finish and then you know people who you know help on the help on other things that sort of bring this all together into the you know the presentation that you make on a daily basis.

[00:04:39]>> So in terms of meteorologists it's me and Pat that's just just us two but we have we have a team of software developers behind that are responsible for helping run the weather forecast model.

[00:04:49]>> Time out on the field. Time out on the field. No time out on the field. Do you believe there's two >> That's amazing.

[00:04:56]>> That's That's amazing to me.

[00:04:58]>> That is lean and mean.

[00:04:59]>> I would I I thought you would have like 10 or 12, you know, people around sort of, you know, putting things in there, but it's you guys that are doing it, huh?

[00:05:09]>> Yeah.

[00:05:09]>> Yeah.

[00:05:09]>> Well, I've been doing monsoon forecast since I got here in 95 at the National Weather Service. Now, I'm going in and out of that. I was up in doing cloud seating forecast up in Idaho. But uh but monsoon is, you know, basically I got a pretty good understanding of how it works.

[00:05:24]>> That's an incredible task that they're doing here. I I got to ask you though, back in if you since you referenced 95.

[00:05:30]Here I go, Pete. I'm going on one of my >> This is my favorite part of the show.

[00:05:33]>> I'm going on one of my sojourns to the past. We true or false. We did not have near the ability to forecast the monsoon in the 90s that we have now.

[00:05:46]Uh, no. I would say uh it was it was it was something we had to learn because we didn't have a whole lot of data. Yes. Uh it was that type of stuff, but uh we had some uh good meteorologists uh doing uh good research down here. They did uh two research projects. One of them called the Southwest um >> Southwest Arizona Monsoon Project back in the 90s. Uh that was done by Bob Maddox. And then uh in 2000 in the late 90s I think we did a a thing called swamp. So >> those two those two research projects gave us pretty good idea of how the monsoon works.

[00:06:25]>> But Pat, let me just ask you this. I'm sorry. Is that is that a lot of the knowledge that at least I possessed in those days was passed to me from other meteorologists who'd been here for a while. It wasn't like I could pick up a computer and start tapping out precise computer models for this that or the other or we had cell phones that would show us where the storms were going to be. I guess that's what I was asking. It was a whole different way of forecasting, wasn't it?

[00:06:51]>> Yeah. Yeah. So, and and like I say, the the models we had back then were were primitive. Uh so, it was it was they weren't weren't very good models. And then remember, especially in Arizona, I always I always tell people where our weather mostly comes from is is Mexico and the Pacific Ocean. And we don't have a lot of data in either place. So >> uh and before satellite pictures, it was it was even worse. So it's it's it's it it's really advanced over those 30 or 40 years. Yeah.

[00:07:22]>> You know, I was wondering, are there certain atmospheric setups during the monsoon where you know, hey, the U of A model is going to nail this thing because this is a situation where the model tends to do really, really well.

[00:07:37]Uh I would say not necessarily like exact setups but there's definitely uh things we look for um certain monsoon flow patterns particularly at the higher levels and and the mid-level 500 mibars looking for good flow in those directions in that in those levels of the atmosphere and then looking for the right cool temperatures or just the right kind of temperatures at 500 millibars. So anywhere from -5 C to - 10 has to be cool there just to allow thunderstorms if they do reach that point to continue rising and create these cumulone nimbus clouds that we're we're so used to in the monsoon season.

[00:08:15]Um and then the surface level we're one of the unique things we do with the model is to change soil moisture. So the the fluxes that are are happening from the the surface up into the near surface atmosphere, they're really important to keep the the surface of the atmosphere humid enough to in the model to start uh convection properly. So I would say one of the really good things about UA Warf is is the um convective initiation. So where storms start. Um again, it's kind of because we have that high resolution um in the grid spacing, it really well represents the terrain features we have.

[00:08:50]So we have a lot of mountains, complex terrain uh in Arizona and and UA War really does a good job of that compared to the the publicly available weather forecast models that are run at a a greater grid spacing.

[00:09:03]>> So one thing I like go ahead.

[00:09:05]>> Yeah. Is uh uh upper level divergence >> uh because if if you don't have divergence when the storm grows it it does not it does not evacuate that stuff at the top. So it basically falls back in on itself. So, I like seeing upper level divergence. I think that's areas where um you know they can you can the big thunderstorms grow. Not not the everyday ones, but the big ones.

[00:09:31]>> Yeah. And the the amount of data they have is amazing.

[00:09:34]>> Absolutely. I I was just thinking about my uh time in grad school of being very nervous during a test and and the the question would be find the area of upper level divergence on the map here which is like you said very important.

[00:09:46]>> Yeah. Yeah.

[00:09:47]>> And it's analyzed for us now.

[00:09:49]So you uh I would say so far this monsoon you guys are on a roll. And now I can say it. It's just these two guys.

[00:09:57]>> Oh yeah. Let's say it. [laughter] >> Yeah. Own it.

[00:10:00]>> Um you're doing you're doing okay. I mean you've called a couple of big storms in Tucson. You've called some big storms up here in Phoenix. A lot of mountain storms as well. I hate to make it city centric but because your discussions, by the way, folks ought to check these discussions out. Um they're detailed but they're not overwhelming.

[00:10:18]And there's a lot of uh graphics that people can actually learn along with reading. So I I would tell people to go if you want to learn about the Arizona monsoon, go read these discussions on a day-to-day basis. So without being city centric, you guys uh first of all, how does it feel that you're you're kind of doing okay so far this monsoon?

[00:10:39]>> It's been challenging. So [laughter] >> actually actually the one I like the best forecast I like last was from last year when we forecast at Phoenix Haboo really well.

[00:10:50]>> So but >> August 25th.

[00:10:52]>> Yeah. August 25th.

[00:10:53]>> But uh it's it's challenging this year because there's more weather. Remember last year the monsoon was kind of a dud.

[00:10:59]Yeah.

[00:10:59]>> So it's much more weather. Much more interesting.

[00:11:02]>> But it must feel kind of good that like like I said you've called well you called that first huge storm in Tucson which was a monster. Um, and it I don't know. I know you don't feel you're Hey, knock on wood, right? You don't want to brag. I I get it. Well, I'll knock I'll say it for you. It was a good call. And let me ask you this. Is that because you are you guys constantly tweaking your models or is that something you look at say in the off season where you say, "Hey, you know, we didn't do so good here. Maybe we could work on that."

[00:11:30]How's that work?

[00:11:32]>> Yeah. If we if we make any changes, we try not to to do it during monsoon. may need to keep consistency in the model because one of the things Pat really does is is look at trends through time.

[00:11:41]So if the model yesterday was underestimating then you might anticipate that it's going to do something similar uh today. Um but using intuition and having the experience that he does is really helpful. So I appreciate Pat.

[00:11:54]>> Cool.

[00:11:55]>> Yeah.

[00:11:57]>> I would remember something from 1999.

[00:12:01]>> That looks just like that. No, I don't do that. I can't remember back then that. [laughter] So, >> all right, guys. We're going to take a break. We're going to talk about a little bit about I used to jokingly call him U of A Mike uh the founder of uh and talk about a little bit of his history, but also how uh things will go in the future. We're going to do that right after this [music] break on Sky Live.

[00:12:29]>> Welcome back to Sonoran Sky Live, a service of First Alert. I'm meteorologist Royal Norman.

[00:12:33]>> And I'm meteorologist Pete Manion.

[00:12:34]>> And this was happening in the Prescat area yesterday.

[00:12:39]>> Yeah, they had a whole lot of rain in a short period of time. Prescuit Valley got hit with some major league flooding caused well a whole bunch of problems to some apartment complexes up there. Those giant containers that you see in the middle there actually floated down a couple of blocks. Uh it rained so hard.

[00:12:58]So, they had one of their uh big ones.

[00:13:01]And that's why the monsoon is so interesting because uh even on a day like yesterday, we did have the flood the flood watch out, but even a day like yesterday that started quiet can turn on us very quickly. We're talking with Patrick Mun and uh Pat Hullbrook from the University of Arizona. They're the they're the major modelers down there.

[00:13:19]They get that UVA model going, which uh quite frankly is probably the best model at this point for handling the monsoon.

[00:13:27]Uh there are a couple of models you guys that um you know the folks up here at the weather service will use on a more shorter term only because yours isn't rocketing you know every hour or so. Um but it really is good to be able to lean on your models. Tell us about UFA Mike Leold who got this started years and years ago. He really was dedicated to this wasn't he?

[00:13:48]>> Yeah I mean before it was me and Pat it was just Mike all on his own. So he would run the model and do all the weather forecast discussions all a oneman van job. So yeah, really uh owe a debt of gratitude to Mike Luod uh started the model in the early 2000s um in the hydrarology and atmospheric science department um and then yeah continued to run it for a decade gaining funding for it and then eventually he retired in 2024 after handing it over to me and my group now.

[00:14:17]>> How do you start a model? I have no idea. Do you take question do you take a model and then start tinkering around with it? a model that you like or do you start from scratch?

[00:14:26]>> Yeah, so it's like a open-source piece of software uh from the National Center of Atmospheric Research. It's called the Wolf model, the weather research and forecasting model. Um so it's kind of free to download. Anybody technically could go and download it and and run it even on their laptop providing it's a small domain. Um but yeah, he he had a dual uh bachelor's in in meteorology and computer science. So that really enabled him to uh set up the weather forecasting part of the model but also the the computing side. So being able to run it on a high performance computer takes a lot of specialized knowledge uh in terms of parallel computing as such. So he was really uh really instrumental in doing that for us.

[00:15:08]Do you keep track of the skill accuracy either through an automated form or just kind of like eyeballing it? Were you looking at what the forecast was and what actually happened? [clears throat] >> So in terms of uh precipitation, we do post some observations uh to kind of do a a real-time evaluation of most recently made forecasts. So, we keep the last 5 days worth of forecasts and the last 5 days worth of observations and that's all available on our on our website um weather.arizona.edu.

[00:15:39]Um but then in terms of the long-term trends in terms of like statistical performance uh that's much harder to quantify um a lot more of a a research um job that needs to be done for that.

[00:15:51]Um but we do we do evaluate um temperature and dueoint temperature and um precipitation where we can. Uh we have a power forecasting website which is also what some of the utilities fund because we we use output from the weather forecast model to uh create renewable power forecasts for um solar panels and wind turbines too.

[00:16:11]>> Uh it is the best model in the west for the monsoon. Pat, I want to ask you this uh question. When I used to talk with Mike a lot and I used to call him all the time. You guys are lucky I don't bug you as much as I bug Mike. Right. Okay.

[00:16:23]Um, and one of the things he used to tell me, I would say, "How confident are you on? How confident? How confident?"

[00:16:29]And he would say, he would tell me his confidence level with a certain forecast, but and then I would say, "How far out?" He'd say, "Royal, I'm I'm never confident more than 24 hours out."

[00:16:40]Pat, I want to ask you how what's your confidence level like where the models are today?

[00:16:45]Well, uh, when I used to work at the weather service office here, I would issue a forecast at 3:30 in the afternoon, and I'd walk out the door and I go, I wonder what's going to happen today. Uh, [laughter] uh, not so much now. Uh, the models the models are really good. Uh, the the problem is only problem you really have is the randomness of thunderstorms, the outflows that happen, those type of things. So, that one side of town will get really heavy rain, the other side won't. So, it's uh but but really the models are very accurate.

[00:17:17]>> And is there something specific that your model does that's kind of specific to the desert southwest that the European model, which you know is a common weather model used by meteorologists across the country does not do because I do notice when I first came here a couple years ago, Royal told me, Pete, be careful with the European model here during the monsoon. So what specifically is the UFA model kind of picking up on that the European model isn't really looking at?

[00:17:43]>> So uh we start the model off using some data from the high resolution rapid refresh model and that has a very sophisticated data assimilation scheme.

[00:17:53]So it ingests um a lot of observations um so brain gauge observations, dupoint and temperature observations from the surface in situ observations and then satellite information and radar information >> airplane too.

[00:18:07]>> Yeah. And then it tries to uh ingest those into the model to try and kick the model off as close to the current conditions as possible because as you make the forecasts as you integrate through in time your errors are only going to grow. So you really want those conditions at like forecast hour zero uh to be as close to the current conditions as possible. So I think us benefit benefiting off that uh high resolution rapid refresh um initialization data uh is really beneficial to our forecasts.

[00:18:37]>> Yeah, it's a good it's a great place to start. That's for sure. Pat, you dodged my question. So, how far off in the future are you confident of your for like for instance, I'm sure you're pretty confident of the rest of today and tonight and you should be, but but because of those variabilities you talked about because one storm could go this way or that way or one outflow boundary could miss that or the other.

[00:18:58]Uh where does the confidence start to taper off?

[00:19:01]>> Uh day two and day three eventually. Well, you know, basically what you're doing once you get past A3 is you're heading back to climatology. There's a 30% chance of rain, you know, that type of thing. Yeah.

[00:19:14]>> So, so basically what we're doing with the warf model is we could uh, you know, identify what's going to happen today and tomorrow really well. After that, you start getting into just the variability of weather. It also be nice to have all the upper air data in state that we're missing. You know, Tucson doesn't do them anymore >> and and Flag only does it once a day.

[00:19:35]It's it's good that our partner at SRP does it up at Phoenix, right? And uh the Marines over in Yuma do it in the morning. So that's that's very very good information we need for our models.

[00:19:47]>> We need the upper air information. That is for sure. Guys, we're going to take one more break. When we come back, I'm going to prepare you. We're going to talk about is Super El Nino going to impact your models? We'll do that after we return.

[00:20:06]Hi everyone. I'm meteorologist Royal Norman >> and I'm Pete Manjone.

[00:20:09]>> Yeah. And we're talking with Patrick Mun and uh Pat Hullbrook at the UVA. They're the model masters. We've had a different name for you every segment, guys. That's just the way it works here. And I I'm We flip the coin. Pete's going to run that Super El Nino question, right?

[00:20:22]>> That's right. Well, Roy, you introduced the topic, so I'll pick up that and run with it. Yeah. Um, you know, there is a lot of buzz over the current El Nino and the super El Nino and we know it's a complex process. You know, my understanding is that it tends to sort of affect late autumn rainfall and precipitation, winter precipitation more so than it does summer rainfall. Um, what do you say to the idea that, you know, because I think I saw this somewhere, some people are already saying, look, Super Nino is causing the monsoon to be active already, but it's not really broken down like that, right?

[00:20:56]Look at that. He's like, "No, >> exactly."

[00:20:59]>> Sorry.

[00:21:01]I looked at that in a study a while back uh early 2004 2005. There's there's no calling Mons both the El Nino and the PDO don't really affect our Monsson precipitation.

[00:21:16]>> That's interesting. You're just saying flat out it doesn't. There's no relationship.

[00:21:20]>> No, no, no.

[00:21:22]The thing it might do is is impact the East Pacific hurricanes. So if there's more hurricane activity, that can that can affect more tropical moisture into the region. And so that's why you might get that um autumn uptick in in rainfall as Pete mentioned. So I think that's one thing that we're looking for particularly late August and September if there's any hurricanes moving from the east east Pacific region into the model uh into the model domain at least.

[00:21:50]Yeah.

[00:21:50]>> Yeah. And and a lot of times that's what happens. We get a tropical system at the end or just after the quote unquote monsoon and that's where we get a ton of moisture from. So um what what are we going to tell all these people that keep writing headlines about hey the super El Nino just crushed Prescat Valley last night. What do we do about that?

[00:22:13]>> I don't know.

[00:22:14]>> Yeah. [laughter] I don't know. I I just basically say it has nothing to do with Super Elino. That'll be this winter because that's when it that's when it does its effects and that'll make us wetter than normal.

[00:22:25]>> Okay. So, I do have actually have a question though. Have you That's legit u although it might be a rookie question.

[00:22:32]I I never deny asking that. Do you over the years have your models been influenced by what we'll call climate change. It's getting hotter. We seem like we're having um maybe less actual rain or thunderstorms, but they could be more severe. Are your models impacted by that?

[00:22:51]>> So I would say the they they would be in terms of like just the way the the models ingest data from from other sources. So they're ingesting those current conditions. So those current conditions will be affected by natural variability and climate change, but also there's a kind of climate change signal.

[00:23:07]Um and yeah, I I would say that the one of the climate change signals for the monsoon is for it to be a kind of more extreme. So, >> yeah.

[00:23:17]>> Uh, less frequent maybe storms, but when they do come, they'll be more intense.

[00:23:22]>> All right, Pat, you got to tell us what you're uh chlingling about there.

[00:23:26]>> I I I hate saying this, but I'm a I'm a climate change and I or at least, you know, the global warming part of it.

[00:23:32]Yeah. So, it's it's it's just that's how the earth uh is made.

[00:23:38]>> But that's okay. That's okay. That's we have that's why we have this show is to talk to everybody and and hear everything. So, that's not a big deal at all. I would like to ask this question, though. Are we seeing are we going to continue to see a pretty decent monsoon in I know you guys aren't climatologists and long range forecasters, but what does it feel like to both of you? I want you to both answer that question.

[00:24:03]>> You first.

[00:24:04]>> Well, [laughter] for the for the for this rest of this monsoon.

[00:24:08]>> Yes.

[00:24:09]>> Yeah. uh you know it with El Nino going on it should end early. So we should by the 30th or of August uh that should be the end of it. So you got 40 days to get rain.

[00:24:21]>> Oh >> so uh uh in Phoenix uh you know uh you haven't got much rain, have you? Except on the east side of town.

[00:24:29]>> Yeah. East side. Yeah.

[00:24:31]>> Yeah. So uh hopefully we'll get one you know one or two good storms uh in Phoenix before the monsoon ends.

[00:24:38]Patrick, what do you think about the rest of the monsoon? Same.

[00:24:41]>> So, yeah, it's tough. Um, I would say it looks like we're going to have a dry weekend and then there's one maybe one more week for July to kind of bring us back to maybe a normal level of precipitation. And then August, it's going to eb and flow, right? So you were saying earlier actually Pat that we're dependent on a Gulf surge in August to kind of bridge this gap between >> the kind of monsoon season ending and then hurricane season ramping up and getting some of the tropical moisture there. My bets are going to be late August, September, we're going to see a lot of hurricanes bringing in moisture and they're not necessarily going to be monsoon dynamics, but we're still going to have a lot of rainfall in that those months and that hopefully will take us back to normal in terms of preip, but we'll see. That is very interesting. So, the monsoon is well, you know what? I love it. It's a mystery to all of us, even to two of the people who understand the monsoon the most in the state.

[00:25:36]Patrick and Pat, it's been great talking with you on Snor and Sky Live. Keep up keep up the good work. You really have the best model in the west when it comes to the monsoon. A lot of people depend on you. And folks, just like I I urge you to check out the website. It's uh weather give me the website again, you guys.

[00:25:53]>> Weather.arizona.edu.

[00:25:55]>> Yeah, it's that easy.

[00:25:55]weather.arizona.edu.

[00:25:57]Check it out, read it, look at their forecast. It's amazing. And guys, keep up the good work and thanks for joining us today.

[00:26:04]>> Thank you very much. Thanks for having us. Thanks, guys. Have a good day.