While mapping the real world, we have a choice between, representing it with the Vector data model, or the Raster data model. Here, I just want to give you a quick comparison between the two, and show you how you can actually use both of those inside the software. With the Vector and Raster data model, a very basic guideline just to get started with is that, discrete objects are easily represented with the Vector data model. Continuous phenomena are more easily represented with the Raster data model. But this is really just a guideline to get us started. Here, we have an image that's been taken by a weather satellite, and what we're looking at here is, a tropical storm, this was Isidore. The question is, how would you represent that? Either as a continuous phenomenon, or as a discreet object. So, in one way, it's a continuous phenomenon because really, it's just a collection of barometric pressure, or it could be the amount of moisture, or temperature, all these things are related of course, in terms of whether a storm exists or not. I mean, we could easily say this is a discrete object. We can draw a boundary around it. Of course, it's a bit arbitrary. Do I include all of this out here as part of the storm, or not, or is it just above a certain air pressure? I'm sure there's a definition for that. Do I just base it on the eye, and that becomes my discrete object, or do I just treat it as a continuous phenomenon, and not try to create it as a discrete object at all? So, in other words, what I'm trying to tell you is that, you have this option, depending on what it is, is you can either represent this thing as Vector or as Raster. With elevation, you could represent it as Raster as I've done here. This is a digital elevation model or DEM. We have squares that are representing the elevation values, but you could represent elevation as Vector lines. So, these are contour lines, and each line represents a line of continuous elevation. So, for example, if this line here represented 100 meters above sea level, then anywhere along that line, that contour line, would be 100 meters above sea level, and maybe the next one is 150 meters above sea level. So, that would be the next contour up. So, sure, we've lost some information here. We're not getting the richness of data that we would with the Raster elevation model, but it's taking much less data to store, and it doesn't dominate our map so much, as if we wanted to include other things on our map, and these contour lines might be a more subtle way of being able to depict that three-dimensional elevation continuous phenomena. So, here I'm showing both just so you can see that you can represent it either as Vector or as Raster. You wouldn't normally have both of them on here, that might seem a bit redundant, but I just wanted you to be able to see the comparison between the two. We can capture discrete objects as Vector points lines and polygons, but we can also convert them to Raster. Here we have a point, we have a line, we have an area, and these would be in the Vector version points, lines, and polygons, and they can be converted into Raster. So, we can capture those and convert them into Raster values. One thing to remember with Raster data though, is that, unlike the Vector version where I can click on a particular polygon and select that, and see what the attribute is for that, or whatever I want to do with it. With Raster, they're just cells with values. You can't actually select a polygon or a discrete feature because they're not there. They don't exist in this version of the data. All we have are cells with values. So, it's sure, there are ways where I could just isolate values that have, cells that have a value of one, and I could say, those represents buildings or whatever it happens to be, or isolate cells with a value of two and those would represent water. But that's a very different way of looking at things. With Raster is that, really they're just cells with values. I don't overstate that, but it almost requires a different way of thinking when you switch between Vector and Raster because you're representing these things in such different ways. So, I've color-coded those in the first version, but here this is really all you're seeing is that, they're just numbers that are representing something, and it's up to you to determine what those things really are. So, you have to know that zero means land, one means building, two means water, and so on. So, you can color code those to be able to see what that might look like, or to help you imagine what it is. But really, you want to try and treat these as numbers that you're seeing in a dataset, you're able to see into the data, you want to be able to see what's going on behind the data, be able to see the values like in the Matrix maybe. I know it's kind of an old movie but it's one of my favorites tonight. I love that kind of idea, that you're seeing into the data. Okay, I think I've made my point. Here, we have a Vector dataset with some points, lines, and areas. So, the points are schools, lines are roads, the areas are zoning. So, we have commercial zones, industrial, and so on. So, that's just to give you a sense of the scale here. What I'm going to do is, convert this into a Raster version of the same data. So, we have points here as, these are actual real points. This is a screenshot or an image that I captured from ArcMap, and these are the Raster equivalence of those school points at a particular spatial resolution. So, that's a comparison between the Vector version and the Raster version of those schools. We can do the same thing for lines. This is a Raster version of the lines captured at a spatial resolution of 25 meters. I did do this on purpose so that you could see, that we're getting this very kind of staircase version of the roads. Of course, this is not very realistic, but that's part of the trade-off that you get in capturing that variability or trying to capture it with Raster is that, if the roads are roughly say, 20 to 25 meters across, or maybe even less, then you're going to get this kind of staircase look to your dataset. So, we could increase the spatial resolution to capture more of it, but I wanted you to just be able to see it easily. So, there's the Vector version of the roads, and the Raster version. Here's the same thing for polygons, and you get the same idea of this kind of staircase boundary between the polygons. So, that's the original Vector polygons and the Raster version. Here's our Vector polygons zoning, and if I select one of these things, you'll see that in the Vector version, we have a table associated with it. I've selected that particular polygon, I can do something with it, isolated it, do whatever I want. So, we have one record per polygon. With the Raster version though, we just have cells with values. So, now I just have no individual polygons, just cells with the same values. All I know is that, we have a value of three, that there are 383 cells that have that value, and that that cell value represents employment industrial for this land use data. So, here's our comparison between the two. So, then the nice thing is, is that at this point in time with a software, it's quite easy to be able to convert data from Raster to Vector. Either way, you can load Raster and Vector data into the same map, you can actually, with a lot of these tools, be able to create a Raster analysis from say Vector points. It was not always the case. In the old days, the dinosaur days of GIS, there was actually Raster GIS software and Vector GIS software. They were completely separate. They didn't work together. They were completely different data models. But of course, over time, they were able to make the software more sophisticated, so that you could work with both of them at the same time, and work with them interchangeably, which is a really nice thing to be able to do with the software. So, I hope that gives you a good sense of the two data models, and the comparison between the two, the Vector and the Raster, and that they can be used for discrete objects, or continuous phenomena, depending on what it is you're trying to do, or depending on who created the data that you plan to use.
