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 most 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 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.
[ Music ]
Drew Lyon: Hello and 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. We have weekly discussions with researchers from WSU and the USDA-ARS to provide you with insights into the latest research on wheat and barley production.
[ Music ]
Drew Lyon: My guest today is Dave Huggins. Dave is a USDA-ARS soil scientist with the Northwest Sustainable Agroecosystems Research Unit in Pullman, Washington. Dr. Huggins conducts research in sustainable farming systems, including the use of crop modeling to predict yield, as well as other soil-related topics, such as management impacts and solar organic matter, soil acidification, soil fertility, and nutrient use efficiency. Hello, Dave.
Dave Huggins: Good morning, Drew.
Drew Lyon: So crop modeling, it’s something I toyed around with a little bit when I was a dryland cropping systems person in Nebraska, kind of a useful tool. You’ve been doing some modeling work here lately. What are some of the questions you’re trying to address with crop modeling?
Dave Huggins: Yeah. Good question, Drew. And to back up a little bit and put things into context, you know, our capacity to collect data seems to be ever growing, whether that’s weather data, or if it’s soils kinds of data, or economic data. So we find ourselves in a position of we’re kind of drowning in data but starving for knowledge. So how do we start to put all that data into a meaningful package? So we’re kind of translating data and forming it into information and then knowledge. And hopefully, you know, that knowledge starts to inform our decision making. So crop modeling is one way to start to gather all that data together and put it into a useful context from the standpoint of our knowledge, hopefully translating it into kinds of useful, a useful tool for making better decisions. So, you know, in terms of our activities today, we’re mostly been working with Dr. Claudio Stockle. He’s with WSU Bio[logical] Systems Engineering.
Drew Lyon: Bio Systems Engineering, yes.
Dave Huggins: You know, Dr. Stockle has worked with what we call a process-oriented model called CropSyst. And this particular model is able to start to grapple with the complexities of how plants grow, and how they get resources and use them from the sun, or from water, and from the soil, et cetera. And so we’ve been using that model to try to look at specific questions. And one of the questions we have is, you know, the use of, the potential use for flex cropping and making flex cropping decisions. So many farmers in the area, they’ll make a decision from the standpoint of whether they want to fallow this year or if they want to grow a crop. And every year tends to be a little bit different. And we see these kinds of decisions playing out in the field in terms of kind of an intensification in wetter years more crops being grown and then in drier years more fallow being out there. So our question was can we use a crop model to basically try to predict what the yields may be for this upcoming year? And to do that for some of the flex cropping options that you may have from the standpoint of a spring crop. So one of the projects that was part of a Ph.D. dissertation by Harsimran Kaur was to basically go through and model spring canola, spring wheat, and spring peas across the dryland region, so of Oregon, Idaho, and Washington, and to model that so using the process model but also using the weather data that we’ve had during the course of the fall, and the early winter, and winter, and the accumulation of precipitation that occurs, so the current year’s data. And we found that, of course, water drives a lot of our yields so we’re thinking, and we are in a Mediterranean-like climate so much of our precipitation actually occurs during the winter that kind of determines a large portion of that yield potential. So we’re looking at the current year’s weather data, looking to see how our soils start to recharge in terms of water and then running the model based on the current year’s data. But then for the weather that we don’t know in terms of the rest of the season right up through harvest, we basically use historic weather, a 30-year normal average, well, we’ll use 30 years of data to basically simulate use the model every year those 30 years to kind of to append to the model that we’re doing with the current year to give a sense of uncertainty around the results. And so from that standpoint what we did was we modeled, again, these are simulated yields for this coming year. On December first, and then January first, on February first, and on March first, and April first to see then how well we’re able to simulate an actual year that’s already occurred. And so from that standpoint we start to see how well the model is performing from the standpoint of predicting a yield, a given yield for that year. And a little bit of what we’re finding there is that, that as we get more information, as we get the weather accumulating, our results in terms of our capacity to get closer to those simulated results for that year tend to improve as we go from December, to January, to February, and then after February, at least in comparison to March and April, we’ve kind of maxed out and we’re seeing that we’re almost predicting about 70% of the variability of the yield by the time we get to February for that spring crop that we’re looking at. And so we’re thinking then, and then we can create maps of these crops across the region in terms of what’s the, kind of the mean yield in any given location. And we can combine that with kind of the uncertainty. That’s what we call the coefficient of variation that kind of looks at, well, how scattered is that particular yield in any given location as well. And hopefully this starts to provide some information that growers can look at and say, “Hey, like for instance this year, we’re at above average precipitation in much of the region. Does it start to make sense to look at more of a flex crop scenario and replace some of our fallow typically with a spring crop?”
Drew Lyon: I could also see something like that working well for making a decision on some nitrogen application if it looks like your yield come February or March is looking really good, maybe you can go out and supplement with a little because as yield tends to go up, protein levels tend to go down, and you might make, if you’re growing a hard red wheat, you might want to add nitrogen or make some — Your ability to make decisions in the crop year might be improved if you have a better idea what your crop’s going to do.
Dave Huggins: Yeah. I think you bring up a good point. And, you know, in terms of trying to use crop models like this that one of the objectives is to try to kind of get a crystal ball out into the future, and to do that before some important decision points that you’re coming up to from the standpoint of either whether to grow a crop or not, but other inputs as well, like you mentioned in terms of nitrogen. Certainly that feeds back into what we might be thinking our yields will be for that particular year. So from that standpoint, right. And so, and this is a little bit of the power of models if they’re actually performing well which we could talk about. But if they’re performing well, then, my gosh, they give you some insights into what the future might be. And you look into the future and some of our capacity to predict some of the more midrange weather and what might be occurring is getting better and better. And so that type of information can feed back into the models to give us even better kinds of predictions in terms of outcomes for that particular year.
Drew Lyon: Okay. So let’s talk about that. How well are these models doing at predicting? I would assume you’re verifying some of the model output in the field. Are you doing a pretty good job on that?
Dave Huggins: Oh, that’s a great question and it really gets back to how much confidence we have in any particular model. And I’ll mention that the model that we’re using currently is the CropSyst model. And this is a model that was developed at Washington State University so it’s kind of an in-house model. And much of the testing of this model has involved a lot of field experiments. So the process is, okay, we go and we grow various crops. We monitor lots of variables from the standpoint of how much moisture in the field, how is the crop responding to that, and how is it growing throughout the season? And this model then takes that type of information and puts it all together from the standpoint of predicting what yield will be. So it all really though depends on how well it’s able to do that. And then, you know, of course, that represents, that field data represents data from that particular year and, of course, every year and every place is different. So then you start to think, “Well, how well will it perform in other locations?” And so that’s where you really, if we’re going to push this whole modeling effort forward, we’re going to have to basically set up a network of field types of situations where we’re growing different crops. And we’re also measuring things like weather and what are the initial conditions from the standpoint of soil, water, et cetera, in order to feed into the model and drive it. And then check the model to see how it’s doing. So this is a little bit of the process that we went through from the standpoint of creating this regional map, was to basically use a variety trial data for the last five years and look at those, how well the model was able to actually predict the yields that we’ve found in those variety trials that were scattered about the three states in terms of canola, and spring wheat, and peas. So we did kind of okay but I’m not, you know, I can’t say that I’m real happy with that and certainly we can make improvements moving forward, particularly if we wanted to get a little bit more aggressive from the standpoint of using these models. I think we’re going to have to set up a network that — and I think, you know, in my way of thinking we already have a network of variety trials that are out there. I think a logical step might be to start to make sure that we’re collecting the data that drive the models, like the weather data, the initial conditions. And then we can start to check and run the model and see how well we’re doing at any particular location. And that helps to see, okay, are there improvements that we can make to the model in terms of making it better? And certainly I think that’s the case. But it also starts to tell you, well, how well is that model actually performing? So that’s the kind of process that you would go through.
Drew Lyon: Okay. So we’ve been talking about using a model for flex crop decision making. What other uses do you see for crop modeling in this part of the world?
Dave Huggins: Yeah. So basically, you know, if you look at CropSyst anyway, it’s able to handle different crops. And so, you know, in terms of assessing how alternative crops might do, we’ve modeled flax, and safflower, and corn, and other kinds of crops that we — and then create maps again of yields of those particular kinds of crop options. And so assessing alternative crops is one thing to do, particularly if we don’t have a lot of field knowledge about that particular crop in terms of how it might perform, a model might be a good first step at trying to assess how it may do in various parts of the region. So I think that’s an obvious type of application. But also, you know, even the crops that we grow now, it would be very useful to know what the yield potential might be for winter wheat and other kinds of crops that we grow, chickpeas, et cetera. And that’s another application that this model can be used for in terms of basically mapping regionally the output or the yield from these particular crops.
Drew Lyon: You know, my experience with crop models, they’re pretty good at predicting potential yield and so if you can figure out what the potential yield is and oftentimes we don’t reach that, then you can kind of look at why aren’t we reaching those potential yields and try to get at what are the factors that are limiting us from obtaining those yields.
Dave Huggins: Yeah. That’s another good point, Drew, because, you know, our models are okay. This CropSyst model is primarily driven by water, and temperature, and the like, and so from that standpoint it’s not including issues that you may have with disease, or with weeds, or with fertility issues, these kinds of things. So from that standpoint you might, the crop model kind of tells you what a potential yield might be. But, of course, what actually happens on any given landscape is much more complicated than what this model is going to do. So from that standpoint that’s simply where we are from the standpoint of this whole modeling effort. Moving forward, I can see though that we start to bring in some of those other factors but we’re not there yet from the standpoint of integrating some of those other factors in terms of how they impact yield.
Drew Lyon: I think crop models can be a great tool. We have a ways to go to make them as good as we’d like to make them but I think it’s a really interesting area of research. And I look forward to hearing more about that and what you’re doing with models in the future. Thanks, Dave.
Dave Huggins: Thank you, Drew.
Drew Lyon: Thanks for listening to the WSU Wheat Beat podcast. If you have questions for us, that you’d like to hear addressed on future episodes, please email me at firstname.lastname@example.org. You can find us online at smallgrains.wsu.edu. You can also find us on social media on Facebook and Twitter @WSUSmallGrains. Subscribe to this show through iTunes or your favorite podcasting app. The WSU Wheat Beat podcast is a production of CAHNRS Communications in the College of Agricultural Human and Natural Resource Sciences at Washington State University. I’m Drew Lyon; we’ll see you next week.