This video illustrates the use of the Distance Speed and Feature Following Zoom tools.

In this lesson we are interested in the motion vector of storms, and in applying linear fits to them. We’re going to spend a little time here understanding the behavior of the Distance Speed tool since it is fundamentally the same as how storm motion estimates in WarnGen are generated, something you’ll want to be very comfortable with.

Before we get into the data, there are just a few quick setup steps I need you to follow along with me on. As with other lessons, we’ll have already loaded the WES-2 Bridge test case in D2D and set the clock to 2340Z. We want a few configurations to our maps and display, which we will simply pull up with a bundle. 

So, we can hit CTRL-O on your keyboard to open the GUI for loading procedures (or select procedures from the file menu). Then, within the OUN group, we’ll go down to W2B_Exercises and double-click that. From within the Procedure window, we’ll double-click the first and only item. In short, this procedure pre-sets useful maps, as well as display characteristics like magnification, density, and the frame count, which will be useful for the products we’ll soon load.

Now, to help you and I be certain that we’re looking at the same features in this lesson, I’m going to enable the lat-lon readout at the cursor tip. This is done by right-clicking and holding the right mouse button in the main pane, and selecting the lat-lon readout option from the popup menu if it’s not already checked. Now we can see the latitude and longitude of wherever we point the cursor.And I’m actually going to go up to the magnification, and reduce it to 1.5.

So, now we’re ready to look at our data! Let’s load some KCRI data by going up to the KCRI menu, which is just to the left of “radar,” and selecting the first item on the list, “0.5 Z+SRM8.” Toggle to reflectivity by hitting the period key once.

And then we’ll again use a utility we demonstrated in the distance bearing video for adjusting brightness, to help the tool that we’re about to load stand out better against this radar data. On your keypad, type CTRL-I… this opens the “Imaging” window. Grab the slider in the lower half labeled “brightness” and drag it down to somewhere around 50%... that’s close enough.  Confirm that the brightness of your radar data in the background did reduce… and then we’ll close the Imaging window. 
 
You’ll also notice, now that data is loaded, that the cursor readout has reflectivity and velocity values, in addition to the lat/lon at the bottom. Let’s use the scroll wheel on the mouse to zoom in a little bit.  We want to keep Weatherford roughly at the left, and Oklahoma City and Norman visible on the right. We notice that the radar data is set to 23:38Z. 

And we’ll take a minute now to loop through all the frames of data we’ve loaded, using our left and right arrow keys. I’ll just hold the right arrow key down for a while. Keep in mind it might take a few seconds for all of the frames of data that we’ve requested to continue loading.

What we see is, in the beginning of our time period, no storms have popped up yet.  But by around 21:30Z some convection initiates along a southwest to northeast line near Weatherford.  And the storm which has strengthened by about 21:45Z goes on to experience a merger with a storm to its southwest. We can also see that the storms’ motion deviates to the right, from its initially northeast motion. So, although a straight line wouldn’t necessarily be the best way to fully describe the storm’s path over its lifespan, we can still unlock useful features of AWIPS by applying a best guess linear fit to the storm. So, let’s try that!

If you’re not already at your most recent frame, let’s go there. You can either use your arrow keys to advance to the end, or, to go directly to the last frame, hold the CTRL on your keypad and then hit the right arrow key. We’ll just verify that you should be at 23:38Z in the product legend.

Now let’s load the distance speed tool. We’ll go up to the tools menu and select “Distance Speed.” We’ll see a pop-up window, and I’ll be sure to set Mode to “Point” and the legend to “speed” if they aren’t already set that way, and then we’ll close this window. We should also now see “Distance Speed” appearing in the product legend, and it should say “Editable.” If it’s not, then we can middle-click on the legend item to toggle its editability.

The distance speed tool will have placed a small target symbol labeled “Drag me to feature.” With my left mouse button, I’ll click and hold on it to drag it, and position it over the core of the storm between El Reno and Yukon. We’ll repeat this step later so accuracy isn’t crucial at this point. As soon as you do, AWIPS will use the last known storm motion estimate to position a notched line through our point… and it might be very wrong, but don’t worry, we’ll correct this next.

Now, how do we get a good guess for the linear motion of a storm? Well, here’s an important tip: don’t use just one frame of motion.  Some storms can appear to change path erratically from scan to scan while in reality maintaining a more consistent motion only visible over several scans. So, a best practice is to span at least three frames, or even more, to find the storm again. This helps to reliably identify persistent motion as opposed to small variations from one frame to the next, which can dramatically change your resulting motion estimate. 

So let’s practice.  Let’s use the left arrow button to go back more than three frames here … I’ll actually choose 10.  One… two… three… four… five… six… seven… eight… nine… ten. You’ll see the target symbol retreated one notch back along the line for each frame that we moved. But also, pay attention to this… you’ll see the position where we first positioned our target symbol appear as a small “x”… it’s a little hard to see, but it’s a small “x” on the path. Any change we make with the target symbol at our new time will pivot around the line… will pivot the line around that “x.” 

Let’s try this… with the left mouse button, click and drag the symbol to any new position. Observe how the x remained fixed and acted as our pivot. If we stay on the same frame, we can continue to adjust this line using the same pivot. So, having done it again, we see that the original “x” still acted as our pivot point. But, once you change frames… so let’s hit the arrow a couple of times… your most recent change becomes the new pivot.  We can see the “x” is no longer where we originally placed the marker, but is where we made our last adjustment.

Now that we’re more familiar with how the distance speed tool works, let’s actually focus on a linear motion estimate of the storm over its entire lifespan. Let’s go to the first frame. We can do that by holding the CTRL key and hitting the left arrow, which takes us to the first frame.  You’ll see 21:03 in the product legend. And then we’ll advance forward with the right arrow key. We want to go to our 21:44 scan, which is about 19 taps of the right-arrow key for me… and you’ll see, we’re on 21:44.

We can again see our storm in its early development just southeast of Weatherford.  I want to use this as the starting point for our linear motion estimate. So, I’ll find the current symbol, and I’ll click and drag it with the left mouse button and position it over the middle of the storm. The lat-lon readout where I’ll placed my symbol is approximately 35.47N 98.63W. I’ll wait a moment to let you position the symbol on your screen at roughly the same location.

At this point our motion estimate is using this new marker at 21:44Z and the most recent adjustment prior to that (you can see that marked with the x partway down the line). We want to use the last available frame for our second point, so we’ll go to the end by again hitting CTRL and the right arrow on our keyboard.  Now you see we’re at 23:38Z again. When we’re at the last frame, let’s click and drag the marker with the left mouse button and place it east of El Reno, about halfway to Yukon at a lat-lon position of roughly 35.49N and 97.82W. Again I’ll pause here while you put your marker at the same location.

 
Now, let’s review our motion fit by zooming out just a little with the scroll wheel, and then we’ll use the right arrow key to cycle through our frames.   What we wanna do is check that the symbol appropriately follows the storm. You should be developing enough comfort with this tool to adjust the line as necessary, but I’m fairly happy with this fit and if you followed along yours should look good as well, so we will not do any further adjustment.

Now, the moving symbol will have a label for the speed and bearing approximated by this linear fit. In our case, it is approximately 20 knots at a bearing of 87 degrees, or roughly east. 

Now, this might seem a little anticlimactic after all that work fitting the storm’s motion, but remember that this must become second nature to you when fitting storm motion for warning generation. What’s more, AWIPS can key in on your defined motion vector and pan your view to automatically follow the storm. This is called the feature following zoom tool, and it is very useful for getting a storm-relative view of how things are evolving.

So, again, having already defined our motion vector with our distance speed tool, let’s now go up to the Tools menu, and select “Feature Following Zoom.” Now when we cycle through time… and I’ll just go up to the loop button in the toolbar and press that… we should see that AWIPS is automatically panning our view to follow the storm. The effect is a little bit more obvious the more you zoom in… so I’ll use the middle mouse wheel to zoom in… and this essentially allows storm detail to be studied in a moving frame which keeps the storm in view. So, for example, we can see the effects of a smaller to the southwest merging with our main storm, and causing some spin up.

Now, as we wrap up this exercise, some caveats to keep in mind: the feature following zoom is only as good as the motion vector that you defied. If you find your frame is not panning with the storm as well as you’d like, you need to redo the motion vector from the beginning. Keep in mind, also, that these tools can only handle linear motion, so for curving motion you’re out of luck unless you manually update the trajectories when needed. Moreover, only one distance speed and feature following zoom will work in a single editor. If you have multiple storms with different trajectories, sorry but you’re going to have to do distance speed and feature following zoom separately for each, either by shifting emphasis in one editor window, or by loading a new editor and setting things up for the additional storm.

Finally, an almost identical functionality for setting a storm motion vector exists in WarnGen, and they will conflict with each other. In any single editor window, the latest storm motion vector, created either from distance speed or WarnGen, is the one which will govern the feature following zoom. So be aware of that when using this tool. 

We’re going to stop the loop at this point. When you’re done with feature following zoom, the only way to disable it is to unload it from the product legend at the bottom right. The distance speed tool, meanwhile, can either be unloaded completely or, rather than doing that, just click on it once in the product legend to toggle its visibility.