[MUSIC] Hello again and welcome back. In this lecture we're going to put some of what we've been talking about about projections into practice in ArcGIS. Specifically, I'll show you the project tool and the define projection tool. I'll also take a look at different properties of projections and how they change when we change the projections of the underlying data. To start it out, I just have a rectangular feature here thats upper corner is in Seattle and its lower corner is in Austin. So remember that, because as we work on the data, those points will be pretty important. Initially, I created this feature in the Web Mercator projection, the same projection that your data would be in if you created it in the same projection as a base map. That's the default projection of a base map. Remember that Mercator projections are good at preserving direction, but are bad at preserving scale, or area and distances. This is an inherent property of the projection. Well, let's say that instead of using the Web Mercator projection, I want to reproject this data to a new projection. To do that I can use the project tool which is in the data management toolbox under projections and transformations. Note that there is a separate project roster tool. Rosters have their own different algorithm for projections in part. And this project tool only works on feature data. So, If I double-click it to open it up, the tool pops up for me and gives me familiar illustrations of what's going on. And I can select my input feature class for what I'm projecting, and it detects the input coordinate system. And I'll put the rectangle_Teale_Albers, which is the name of the production that we're going to use. And I'll select my output coordinate system here. And I'm going to select one that's really common for California, in part because it's a bad projection to use for this, and I want to show you what happens when we select the wrong type of projection. So I'll use a shortcut to find it. And that shortcut is known as the EPSG code. And it's a code that each projection has that's a unique identifier so that you can look it up that way, and reference it to other people based upon that unique code. Now, not all projections have this geographic transformation option. But, the Web Mercator projection is based on the WGS 1984 datum and the California Teale Albers projection is based on the NAD 1983 datum. In order to project between them, remember that I need to not just translate the coordinates, but I also need to translate the datum. And so it automatically selected a geographic transformation for me. There are potentially other transformations that could apply here, but it selected a generic one that does this well across the entire United States or entirety of North America for me. Now as usual, to run it, I just click OK. And when it finishes running I get it back in my table of contents. And initially, I don't see any outward change here in my map document. Now remember that ArcGIS, in order to display the data to us, or in order to analyze the data, has to bring the data into a common projection. So what's happening here is, while the underlining data has been re-projected, it's again projecting it on the fly. It's re-projecting it on the fly for us, so it can display it on the map. So it's showing it in the same projection as the rest of this data, which is also in Web Mercator. If I want to see visually what the transformation has done, I'll need to add a new data frame to my map. So I'll go to Insert > Data Frame, and now I'll drag this across to the data frame. And right way we can see the difference. Where before we were visualizing a rectangle that fit on the map with straight up and down sides and completely horizontal top and bottom, now we have this sort of distorted view. And if I add a base map again, I'll add the same base map. We'll see exactly what's going on here. And now let's make this somewhat transparent so that we can see a little better what's going on. And as we talked about, projections are optimized for specific locations. So, in California, there's very little distortion. But as we get further away, into Texas and Boston area over here, where the rectangle ends, the data is getting more and more distorted out further there. One of the reasons for this is the optimization as I talked about before, but the other is that the projection I chose is an Albers projections which is an equal area projection. It preserves areas well but it doesn't preserve distances or direction. And while the Mercator projection doesn't preserve distances, it does preserve direction. So we're optimizing for different properties. And if I want to see that in action, I can measure from corner to corner here, so from Seattle to Austin. In this not properly chosen projection, if I measure from Seattle to Austin, in this not correct for this used projection, I get about 2,800 kilometers. Now if I switch to this other projection in the other data frame, if I measure it here. Again, not a distance projection, but a different set of properties being preserved, I'm still going to get a very different answer, 3,700. So 2,800 kilometers versus 3,700. That's a big difference. And that's all just due to reprojecting the data. Now, under the hood, remember we're not actually destroying our data. It's referencing the same locations on the map, but, it affects the use case for that data what projection we're storing it in at the moment. So keep that in mind. We haven't really destroyed our data. We can re-project it again, but the use cases for our data are based upon the projection it is in, and the projection is being displayed in on our map. Okay, lastly I want to talk about the define projection tool over here. A lot of times people get this confused with the projection tool, and I want to make it clear. The project tool is like a re-project tool or a transformed projection tool, where it changes the underlined data in our feature class. The define projection tool takes a feature class that doesn't already have the projection information specified, where the data is already projected. So the data is already stored in a projected coordinate system, but ArcGIS doesn't know what it is. And the define projection tool specifies what projection that data is in, so that ArcGIS can understand it. And the consequences for using it incorrectly are huge, because if you tell ArcGIS some data is in a different projection than it's actually in, it's going to display in the completely wrong part of the world. So, let's try that now with our rectangle here. And I'll do that by telling it that the Web Mercator projection is actually in the Teale Albers projection, and we'll see where it goes. So if I take the Web Mercator projection as the input data set, and it's going to warn me that it already has a projection, because you really don't want to do this unless you know it's already in a wrong projection, or if it comes in without projection information. And I'm going to specify the projection again. So I'll type in again that spatial reference ID, and search. And I can pull it off of the existing layer or just off the search. Now I'll click OK to define the projection. Now I can still see the Teale Albers version of the projection, but if I turn it off, all of a sudden the other one has disappeared. Where'd it go? So, to find out, let's right-click and Zoom To Layer. And look, it's over Indonesia now. It's in the completely wrong part of the planet. I didn't actually even know where it was going to end up, I thought it was going to end up somewhere else. And so, it just goes to show that when you use the define projection tool improperly, the results are a little unpredictable. Basically what's happening is, it's now interpreting the coordinates from the Web Mercator as if they were in the Teale Albers coordinate system. And they use very different sets of numbers, so when you apply the numbers from one coordinate system into the other, it's going to show up in some weird locations. So, don't use the define projection tool unless you have a dataset that is of an unknown projection, and you are trying to tell ArcGIS the projection you know that data to be in. When your dataset is in an unknown projection, ArcGIS will warn you when you try to add it to your map document. And it'll usually give you a little warning icon in the table of contents, because it can't adequately draw it without projection information. Okay, that's it for this lecture on projections. I just wanted to give you a quick introduction to the tools we use in ArcGIS to manage projection information on our feature data. Namely the project tool that we can use to transform our data between two different projections, as well as the define projection tool that we can use to tell ArcGIS what data or what projection a dataset is in, when we know what projection it's in, but the dataset doesn't have that information stored. I also showed you a little bit of the consequences of storing data in different projections for our analysis work. I hope that helps bring some of these concepts home just a little bit more so that you can understand them more deeply. There's a lot more we could talk about about projections and transformations. You could run a whole class on it, but in this case that's all the time we're going to spend on it. So, if you have more questions, bring them to the discussion forum or look up more information on the web. It's a humongously deep topic that can be really difficult and really rewarding at the same time. Okay, I hope that helps. See you next time.
