 Hi, I'm Mark from the Systems Integration Facility. In this video, I'd like to show you how to use WMT, the Systems Web Modeling Tool. If you're unfamiliar with WMT, it's a web-based application that allows users to build and run coupled surface dynamics models on a supercomputer. WMT can be accessed at this URL, systems.colorado.edu.wmt. Note that we use HTTP secure to protect your login. Alright, let's take a look at WMT. WMT always starts here as a sign-in screen. To use WMT, you'll need to create a login using an email and a password of your choice. I'll sign in now. Note that WMT remembers your email address, and if this is your first time using WMT, you'll be asked to repeat your password. Once you've signed in, you'll see the main screen of WMT. The main screen is divided into two panels, the model panel and the parameters panel. The model panel is where a stand-alone or a coupled model can be created. The panel provides a set of buttons that can be used to open, save, and run models, as well as a tree structure, which can be used to design a model. Though initially blank, the parameters panel is where the parameters and model components displayed in the model panel can be viewed and edited. Let's use WMT to build a coupled model. I'll start by selecting the driver for the model in the model panel. This box is what we call the root of the model tree. From the available components, I'll select CEM, the coastline evolution model. CEM is now the driver for this model. Note that the parameters panel now contains information, the parameters for this model. We'll come back to them in a moment. If I select CEM, you can see that I can show parameters, which I'm already doing. I can also get information about this component model. I get an overview. If you'd like, you can also click the link to go to the Model Help page on the system's website. I could also delete the model. That's not bad. I can always add it back. Let's visit some of the parameters of CEM. I'll keep the run time, so that's 10 years. Let's make a smaller domain size. And output files every 100 days. There. Now you can see in the model tree that CEM exposes two usage ports. These ports can be filled by another model that provides these outputs. Let's choose first to fill the Waves port. There's currently only one model in WMT that fills this port. It's called Waves. If I select it, I can show its parameters. Now shown in the parameters panel. I could also get information on it or delete it. Next, let's fill the River port. Currently, Evulsion is the only component that satisfies this, so I'll select it. Note that Evulsion fills CEM's River port. But it also exposes two ports of its own. Discharge, and this is Elevation. So Evulsion provides the River port for CEM, which in turn provides the Elevation port for Evulsion. This represents a two-way model coupling. In WMT, this behavior is represented with an alias component. You can see a little symbol here. So the CEM component at the root of the tree and the CEM component that provides elevation for Evulsion are in fact the same instance. If you'd like to learn more about model coupling, please visit the system's website. So let's look at Evulsion's parameters and modify them to match those of CEM. I want to change the number of rows, the number of columns, grid point of the inflow, and rows and columns. This will be at the midpoint, and then again the output. Let's finish by filling the Discharge component for Evulsion. Here I could choose Hydrotrend or River. I'll choose the River service component. Again, if I wanted to get information on it, show its parameters. At this point, the design of the model is complete. Let's next label and save the model. Prior to saving a model, it's useful to label or tag the model with descriptive keywords. These keywords can be used to search for the model at a later time. The labeling system in WMT is designed to mimic the system used in Gmail. To add labels, select More, Manage Labels. Note that a label has been created with my username and there's also a label selected for the driver of the model. By way of example, I'll select the Test label, which is Public and therefore available to all users. I'll also create my own Private label and call it Tutorial. Last, I'll choose to make this model Public by selecting the Public label. This way, anyone who uses WMT can access this model. Now let's save the model by selecting the Save button. You can give the model a name. I'll choose to call this WMT Video Example. When I save it, note that the model is saved to the WMT database server, not to your local machine. This means that you can access save models from any web-accessible computer. For example, I've created this model here at my office. I could go home and open the same model from my home computer or from my iPad, for example. Now to show how labels work with saved models, I'll delete this driver component. Now I'll try to open the model I just created. You can see that there's already quite a few models that are saved in the WMT database. So let's filter the results that are returned by looking for a label, like maybe the Test label, for example. So you can see that by selecting that label, I've got a filtered set of results, which is much easier to see. In fact, note that labels are conjunctive. If I'd filtered on both tests and say tutorial, I would have got a very restricted list of models. So let's open that. You can see it's my model. Alright, so now let's run the model and get some results. Before I run this model, a few points. First, a model needs to be saved before it can be run. Second, you must be online in order to submit a job. However, you don't have to stay online and wait for the results. And third, you'll need sign-in credentials for the supercomputer on which you're running your simulation. If, for example, you don't have a sign-in for the system supercomputer, please see the system's website for information on obtaining a sign-in. Okay, to run this model, select the Run Model button. You'll be prompted for your sign-in to the system supercomputer beach. I'll sign in. Note that your sign-in for the supercomputer may be different from your sign-in from WMT. Okay, so I'll sign in. I'll hit Run. If your job is submitted successfully, you'll see a success dialogue like this. You can select the View Run Status button to see how your model run is doing. Alright, so this part of WMT is still under development. I'm going to need to refresh in order to see how it's doing. Okay, so recall it's going to run for 3,650 days. Alright, so I could keep hitting refresh to see what happens. But I think that's going to take a little bit of time. It takes about 10 minutes maybe for this particular model to run on beach. So what I'll do is take a look at the example that I created earlier. So you can see that you can download the results as a TAR-GZ bundle. And this will be available on your computer. So in summary, with WMT a user can select a component model from a list to run in standalone mode. You can also build a coupled model for multiple components organized as nodes of a tree structure. You can view and edit the parameters for these component models. You can save the models that you create to a server where they can be accessed on any computer connected to the web. You can share models that you created with others in the community. You can also run a model by connecting to a remote high-performance computing cluster where the system software stack is installed. Alright, thanks for watching. Don't forget the URL for WMT as well as the URLs for the WMT help and WMT tutorial. Alright, buh-bye.