When we're digitizing things like areas or any object really, we have to think about what scale are we at when we're creating these objects, when we're digitizing in relation to what scale we want to actually show those things on in our completed map. So, you'll see here. I haven't talked about scale much so far, but you can see that the scale here is actually at 1 to 5,287. Really, for now, you can just think of it as zooming in or zooming out in the scale. Let's just turn off our trees for now. If I wanted to trace the outline of this park, I have to create a new feature class for polygons. So, I'm just going to call this Park. As a polygon, and you give it the same coordinate system as my trees. I'll just quickly put this together. I'm not going to actually put any attributes in here and I'm going to ignore this warning. So, now, I can start editing and select my park. Okay. So, I've clicked on the Create Features tab here at the bottom, that may not be obvious. So, I select Park here. So, now, when I go over to the map, I can trace out the edges of the park. So, in order to do that remember that with vector primitives, a polygon is created using points that are connected by lines to create a polygon. So, I have to click on a location where I think the edge of the park is and then I click a second time to create another point that are connected by a line. Really, what I'm doing is I'm clicking every time I think there's a change in direction in the shape of my object that I'm trying to digitize, to try and capture the essence of that shape as I go along. Part of this has to do with making decisions about, well, where is the edge of the park? Is it the edge of the grass? Is it the sidewalk? How many points do I want to create? As you go around, you'll see that it's finishing that there and then I can just right-click and say Finish Sketch. So, now, I have a polygon that represents Queen's Park or part of the Queen's Park area. If I save my edits, Stop Editing and just for now, so you can see how this works, I'm going to remove the fills, say No Color, change the outline to something that's easy to see and maybe increase the thickness a bit. So, that's now a vector polygon that represents the outline of that part. You might look at that and say, "Wow, that's pretty good." I actually did a fairly quick job there. I wasn't paying a lot of attention. If I was only going to show this vector polygon at this scale then you might think, "Well that's a pretty good job, you didn't do too badly there." But if I zoom in, maybe I want to make a map that's at a much larger map scale, in other words, more zoomed in. When I do that, you might start to notice, well, actually, I didn't do such a great job tracing the edges of the park. You'll see that some places, this is a good example over here, is that I missed part of the park there. So, really, it should have come out that far. Here, I went a little too far in, so that should be out there a little bit. So, I didn't really trace it that closely. So, if I say created that dataset, shared it, let people download it online, and then you went and created your map at a zoomed in more with what we call a larger map scale, you might say, "Wait a minute, that's a lousy dataset. Look at how badly he traced that." You could hurt my feelings by doing that but nevermind that's okay, I can live with it, I've got a thick skin. But more to the point is, that it's not necessarily that it's a bad data set, it's that it was created at a certain scale for a certain purpose, and you're trying to use it at a scale at which it's not appropriate. That's really what it comes down to. That's my rationalization anyway, so you don't hurt my feelings. Just kidding. So, if I zoom out here, to let's say that scale, then it doesn't look so bad. Or what if we zoom out to a scale of 1 to 24,000? We'll just let that load. Suddenly you don't notice those little inaccuracies in the way that I traced it. So, if you're working at a scale where you're zoomed out more than the way it was traced, then you will notice those problems and that's an appropriate scale in which to use that data. So, the rule of thumb is that if you're creating data, digitizing it, I recommend that you zoom in farther than you need for your finished product. So, that you can trace those details and get it as accurate as you think is useful for you, so that when you zoom out to the scale that you actually want to present that data at, no one's going to notice those those little errors or problems that you had. So, anywhere that I zoom out from here, those accuracies will not be evident and everything will be fine. Just to illustrate this with some preexisting data. So, what I was doing before was creating my own data and digitizing it. But when you download data or you get data off of Esri's website or wherever you get it from, somebody else had to digitize that data and that's what I'm looking at here. This is a dataset from Esri, Canada, that shows the provinces and territories for Canada. At this map scale, when it's zoomed out quite a ways, you might think, "Well that actually looks pretty good, there's no problem there." That's exactly right. At this map scale, it's appropriate to use this dataset. As we zoom in, we start to see that, okay, that's still looks pretty good. If we continue to zoom in though, we start to see that not so great here, we actually have some problems. Is that if you compare this outline, you'll see that when this was created, they didn't do a very detailed job. I'm not going to say it's a bad job because it's just what it was designed for or not designed for. Like so, for example, this bay is missing up here, the shoreline doesn't match very well over here. So, again, this is an example where the data was created at a certain scale, it's meant to be used at a certain scale and if you try to misuse it, you can't really blame the data provider. It's important that when you get a dataset, you find out from the metadata, which is the information about the dataset, what scale was used to create it so that you can decide whether that scale is appropriate for your use. So, here's the metadata for this data set. This is the description that was given. You'll notice that it actually says, "Suitable for use at a scale of one to two million or smaller." Which means zoomed out beyond that. So, what I've done here is I've zoomed in to much larger map scales, so that they're telling me they've been warning me to say, well, wait a minute, that's not what it's meant for, so don't use it at that scale. So smaller map scales have less detail, larger scale maps have more detail. It's the opposite of what you might think but that's how it works. Another example I can show you here is that this is a map of Canada with some major water bodies on it. At this map scale, it's about probably 1 to 36 million, I think I made this at. It's great but if you take that same data set and zoom into a much larger map scale and look at one of the lakes, well, that's actually a really lousy version of Simcoe again. It just wasn't designed to be used at that scale. So, I've got a different dataset that was created at a larger map scale that has more detail, if I want to use it for a larger scale map.
