Subscribe on iTunes | Android | Stitcher | SoundCloud | Spotify | RSS feed
Show Notes & Resources Mentioned:
Contact Information:
For questions or comments, contact Joe via email at weather@wsu.edu.
What is a podcast?
For those of you who are newer to the medium, a podcast is like a pre-recorded radio show. In the same way that you turn on a talk radio show, you have to turn on a podcast. The major difference is that while our cars are equipped to find radio frequencies, they are not built to accommodate direct access to podcasts. On your smartphone or computer with internet access (since the files tend to be on the larger side), you can discover podcast shows of any kind, in any field, on any topic.
Listed above are some of the most used podcast hosts. iTunes and the iTunes Podcast app are preinstalled on your iPhone and are the simplest tools to use. You simply search for “WSU Wheat Beat Podcast” in the search bar, hit “subscribe” and the download arrow, and listen whenever it’s convenient for you.
If you use an Android or use another type of smartphone, you will need to find a different podcasting app because those devices don’t come with a preinstalled app like Apple. If you don’t know which podcast app you’d like, simply hit the “Android” link above and it will show you to several Android podcast apps for you to choose from.
After you download an episode, you can listen without using data any time of day. Our goal is to post a new podcast every other Monday. Your podcast app should automatically load our new episodes and download them for you (on WiFi), hands-free if you choose that in the app settings.
If you have further questions about what a podcast is, which app is best for you or need more assistance with getting started with podcasts, don’t hesitate to contact us.
Episode Transcription:
Drew Lyon: Hello. Welcome to the WSU Wheat Beat podcast. I’m your host, Drew Lyon, and I want to thank you for joining me as we explore the world of small grains production and research at Washington State University. In each episode, I speak with researchers from WSU and the USDA-ARS to provide you with insights into the latest research on wheat and barley production. If you enjoy the WSU Wheat Beat podcast do us a favor and subscribe on iTunes or your favorite podcasting app and leave us a review while you’re there so others can find the show too.
[ Music ]
Drew Lyon: My guest today is Dr. Joe Zagrodnik. Joe is a Postdoctoral Research Associate at WSU AgWeatherNet. He joined WSU in May 2019 after receiving his Ph.D. in Atmospheric Sciences from the University of Washington in Seattle. Joe’s research has spanned a number of meteorological topics including tropical cyclones, mountain waves, cloud microphysics, mountain hydrology, high-resolution modeling, and Pacific Northwest climatology and extreme events. At AgWeatherNet he is studying the effect of irrigated fields and orchards and weather observations and is working to develop site-specific weather forecasts for agriculture. Hello, Joe.
Dr. Joe Zagrodnik: Hi, Drew.
Drew Lyon: So most of use our Smartphones to get a weather forecast. There’s these forecasting apps on our phones, can you give us a brief background about where these forecasts come from?
Dr. Joe Zagrodnik: Absolutely, so there’s a lot that goes on under the hood to make what looks like a relatively simple forecast on your phone, and all weather forecasts originate from what we call these physical base weather models. And what they are is a giant three-dimensional grid of the atmosphere that’s created using weather observations. So ground weather stations, but also weather balloons and satellites, and so they create that grid and then they step forward to the actual equations of motion, so that’s Newton’s Laws, that are stepped forward in time in order to basically move the atmosphere into a future state. And, as you might imagine, that causes our weather forecasts to be wrong in many ways because errors grow exponentially in time, so that’s the butterfly effect, basically, and they just keep growing and growing the further out you go. And also this grid can only be so detailed because literally computing power is the limitation, so many processes are too small or detailed, such as for instance the model does not see individual clouds so those have to be parameterized. So what atmospheric scientists do is they have to post-process these models, either so the weather service kind of does this manually based on the experience of meteorologists with decades of knowledge on a local area or increasingly post-processing algorithms are used which calibrate a forecast to an individual location and that’s often what you’re seeing is these private companies like the Weather Channel applying a post-processing algorithm to the forecast in order to try and get it for a specific location. But from a researcher’s side, these are kind of a black box because they’re proprietary and we don’t always know what’s happening and also if there is no observation to the location, so for instance if you’re trying to forecast for a field and it doesn’t have a weather station it’s ultimately still a little bit of a guess, so there’s bound to be inaccuracy for a number of reasons.
Drew Lyon: Okay, well, that kind of explains why I can get on three different forecasting tools and get three different forecasts?
Dr. Joe Zagrodnik: That’s correct, yes, each one has its own algorithm and who really knows where it’s coming from. A lot of times, you kind of have to play with them and see which one works the best for your location.
Drew Lyon: Okay, well, we know that weather forecasts still lack some accuracy, they can sometimes be quite accurate and other times be pretty far off, especially it seems here in eastern Washington because of all the topographical differences we have across the area. What are some of the challenges associated with forecasting in this part of the world?
Dr. Joe Zagrodnik: Well, you already mentioned the big one, which is the complex topography. So again that’s when you look at these actual physical weather models one grid box is maybe seven or eight miles in length, so everything within like a seven or eight mile square is going to be represented by a single number. So that’s fine if you’re in Kansas where it’s completely flat and kind of uniform, but when you bring in topography you bring in microclimates, inversion formation is a big one, at night the cold air tends to pool in the low lying areas, that creates complicated belly flows of wind. Then there’s irrigation, which further changes the land surface, the effects of evapotranspiration, and of course, then you have our precipitation around here which as we see in the spring tends to be extremely site-specific and isolated where one area has really heavy rain showers and it’s nothing and a couple miles away [ laughter ]. So trying to get all that into a single model grid box is quite a problem. But from my perspective, from a meteorologist it’s really fascinating, it makes for a lot of interesting research topics, but if you’re just trying to get a correct forecast for your field then it can be pretty frustrating.
Drew Lyon: Yes, you have your work cut out for you as a forecaster in this part of the world. So what are you working on in AgWeatherNet to improve these forecasts in this highly diverse area we live in?
Dr. Joe Zagrodnik: Yes, so we’re seeing a lot of really interesting things. So as I mentioned temperature is one of the things that is especially troublesome, and it turns out this is one of the easier things to fix or at least improve. And we’re using machine learning tools to do this, and the key thing that you need for the machine learning tools is validation data. So a weather station, which is exactly what AgWeatherNet has quite a few of, so if you can validate and correct the forecasts using a weather station you can get more accurate temperatures. And this just makes sense for AgWeatherNet because we’re trying to emphasize site-specific forecasts anyway, rather than using these grid-based forecasts that just don’t make a lot of sense in this part of the world. And kind of one of the interesting things about our forecast model is that we developed it when I was a Ph.D. student at the University of Washington and we used our model to help our team win a weather forecasting competition against other meteorology schools. So it was initially kind of started for fun, but then we started to see how this could be used in the real world. So that’s the main one and at least the first one that will likely be deployed to everyone, but we also have other projects actually trying to improve the physical model data, so using higher and higher resolution models, which as computers get more and more powerful is increasingly possible. And then also using what’s called an “ensemble approach” to weather forecasting, which basically means that you run the model lots of different times in kind of slightly varied the way you run the model each time, so then you get 20 or 30 forecasts or even 50 forecasts for one location and that not only is the average of that better than using just one weather model. So, for instance, if you got the forecast from like 30 weather, if you did like every weather app that you could find and just averaged all 30 of those forecasts it would probably be better than just picking one that may or may not be accurate, so that’s an approach that we’re trying to apply to the forecast as well.
Drew Lyon: Interesting, so temperature, I know as somebody who makes recommendations on herbicide applications, things like wind speed and inversions are all pretty important, are these things you’ll be working on going forward as well?
Dr. Joe Zagrodnik: Yes, yes, wind speed is another one that goes along with, which again is really dependent on having observations in that area and also something that’s a higher resolution model, you see a lot better than these global models. And in a lot of cases that data is already out there and it’s just a matter of collecting it and delivering it. And for inversions we are doing a number of things, we’re deploying, right now we’re installing meteorological towers in several locations in Washington State, we can actually measure the inversions. And we have some proposed research out to do more interesting things, such as monitoring inversions on balloons attached to tethers or attaching weather stations or weather observations to a drone. So I think in the next couple of years we’re going to learn a whole lot more about inversions and how they vary and how they form, so it’s really exciting.
Drew Lyon: Yes, that’s very interesting, that’d be a great step forward if you get to that point. So what is the accuracy of AgWeatherNet’s forecasts and why do you think the AgWeatherNet forecasts might be better than some existing sources out there?
Dr. Joe Zagrodnik: So they’re actually pretty good. For a next-day forecast for tomorrow we’re pretty good at getting within three degrees Fahrenheit or so overwhelmingly and the majority of the time, and that includes the trickier forecasts, such as the low temperature and clear night. And we don’t really, where these machine learning approaches really are good is when you go out to three, four, five days, it doesn’t really get that much worse than that. So we’re pretty confident that we can go up to seven days and the accuracy at day seven is probably more like four or five degrees, but it’s still not terrible. And the reason is that it’s the site-specific nature of them, so we’ve tested at 21 locations where there’s a weather station and we’re expanding that to all the AgWeatherNet sites. And when you run the forecasts with that validation data you can be way better than these gridded forecasts, which include companies like Dark Sky, which a lot of people like. So we’ve compared our skills of the forecast to Dark Sky and we’re consistently able to beat them.
Drew Lyon: Excellent, so that’s something people might want to access to get AgWeatherNet forecasts. How do growers go about getting access to those forecasts and when are they going to become available?
Dr. Joe Zagrodnik: So AgWeatherNet is about to roll-out an app called AWN Farm, and this is currently in beta testing so it should be this summer. And everyone will be able to download that first release. Initially, it’s going to use National Weather Service forecasts, but there will be a transition to the forecasts hopefully by the end of this year in the fall or winter. But the key thing is those still will be based on the forecasts at the nearest AgWeatherNet station, so I strongly encourage anyone who would like their own site-specific forecasts to contact AgWeatherNet about installing all-in-one weather stations. And I know that a number have just been installed in wheat fields to further look at falling number issues. So they’re relatively inexpensive, they’re only about $2,500 or so, and we need, my tests show we need about two years of data to start giving you accurate forecasts. So basically the sooner the better in terms of having observations available for us to make forecasts from.
Drew Lyon: Okay, so how does a person contact AgWeatherNet?
Dr. Joe Zagrodnik: weather@wsu.edu is the easiest way so, or if you forget that you can go to our website and hit the contacts form, weather.wsu.edu and we basically are building this program to help with that, so it’s part of our long-term strategy and a number of growers are already coming on to our network. So if you install the weather station you get access to everything that AgWeatherNet provides on the site, so I highly recommend it.
Drew Lyon: Okay, well, I can only think of one or two things that farmers are more interested in than weather, so I think this is something that they will be interested in, and I appreciate you taking some time to talk to us about what AgWeatherNet has to offer.
Dr. Joe Zagrodnik: Excellent, yes, thanks, Drew, and I love talking to farmers about weather too, so happy to, if anyone wants to reach out to AgWeatherNet I’m happy to chat. So I learn as much from farmers as they can learn from me, so the more we have a two-way platform the better.
Drew Lyon: All right, well, excellent. Thank you very much for your time, Joe.
Dr. Joe Zagrodnik: All right, thanks, Drew. It was great to chat.
[ Music ]
Drew Lyon: Thanks for joining us and listening to the WSU Wheat Beat podcast. If you like what you hear don’t forget to subscribe and leave a review on iTunes or your favorite podcasting app. If you have questions or topics, you’d like to hear on future episodes please email me at drew.lyon — that’s lyon@wsu.edu –(drew.lyon@wsu.edu). You can find us online at smallgrains.wsu.edu and on Facebook and Twitter @WSUSmallGrains. The WSU Wheat Beat podcast is a production of CAHNRS Communications and the College of Agricultural, Human and Natural Resource Sciences at Washington State University. I’m Drew Lyon, we’ll see you next time.