Raster Data Theory (2/6): Raster Fundamentals

hello everybody and welcome to this video lecture where we're going to get started learning some foundational information about the structure of the raster data model and how it contrasts with the vector data model so if you remember a while back ago we said there are two major ways that we can store information or data about geographic features on the planet we have two major data models for this and the first one was the vector data model we've looked at that extensively at this point worked with it a lot and I hope that if somebody says vector data model vector GIS data you should immediately think about points lines polygons or points lines and areas thinking about that geometry and then thinking about an attribute table and then thinking about that connection between an attribute table and some geometric representation and that's really what made the vector data model work that's a simple idea that at its core but there is a whole lot of things that you can do with that and it gives you a tremendous amount of power that's the core of the vector data model and we're going to contrast that now with the raster data model the raster data model is about maps as numbers you frequently have this phrase that you hear when people are talking about raster's which is maps as numbers and so basically what raster's allow us to do is to be able to take a map and then convert it into numbers rather directly and then we can do all kinds of mathematical manipulation on those numbers in order to solve problems answer questions and conduct whatever analysis that we want to do if you are mathematically inclined if you like math you're probably really going to like working with raster's because if you can think about how you're going to conduct your analysis if you can think about what you want to do in whatever it is that you're analyzing whatever analytical process that you want to accomplish if you can think about how to do it mathematically then you're going to be very you might be able to vary easily and directly apply that mathematics to the rafters in order to compute what you want to compute so in some ways the raster data model is easier for computation than the vector data model we have all those specialized tools in the core GIS toolkit that we were talking about which are sort of our foundational tools we have less foundational tools actually in the raster data model because the tools that we have are much more encompassing because they are just about mainly doing raw mathematics on these rafters and then producing some output that has meaning so we'll look at what kind of the equivalent of our raster of tools that in future videos but also raster's are at their core a grid of cells okay so when you think vector points lines polygons raster data model grids of cells maps as numbers so what we're going to do what do we do with all those numbers how do we generate those numbers or what do we do with them well we're going to have this grid of cells that we're going to overlay on the area that we are interested in studying and that we are interested in representing some geographic feature or phenomenon inside and then we're going to put numbers inside each one of those cells in order to represent the feature the way that we would like to have it represented so grids of cells with numbers in them are the very core of what it means to represent something in the raster data model so in the raster data model you do not have geometry and you do not have an attribute table we're not talking about geometry and links to attribute tables here we're just talking about grids of cells with numbers in them ok let me give you two foundational rules for raster's and these are pretty simple rules raster's are conceptually simple the vector data model was conceptually simple as well but the raster data model is is very conceptually simple and we have these grids of cells and then we're going to follow these two very simple rules the first is that every cell can contain one and only one value okay we're going to be able to put our number inside each one of these cells but it's critically important that every cell only gets one you can't put more than one value per cell so every cell can contain one and only one value or number that means we're going to have to be very deliberate about what number we put into each cell we want to make sure it's a good one so to speak because we only get one per cell and we'll look at different techniques for putting numbers into these cells later but every cell can get one and only one value or number and also each cell in a raster has to be exactly the same size so rasters do not look like what I have over here on the right every cell in a raster must be the exact same size so let me X that out now of course you can have different rafters with different size cells different size cells across different rosters but every time you have one single raster they are all going to be the same size cells so we take this grid we make them all every cell the exact same size and we put only one number in it well basically we are off to the races as far as the raster data model goes because we have our sort of corefoundation then let me also introduce to you this idea of spatial resolution spatial resolution is another very core component of the raster data model and something we need to understand because it's always with us so basically what spatial resolution does is communicate to us the amount of area that each one of those cells on the planet represents so each cell in each particular raster represents some area on the planet right we're going to geo position this grid of cells over a planet so it's got to look over the planet so we've got location the location for it and then we're going to look at every one of those cells and we're gonna say okay this cell represents this certain area so it's very important when you're working with raster's to communicate what the spatial resolution of the raster is are you talking about each cell representing some giant amount land area or a much smaller area does FP land could be ocean but what is that in this case for instance we're looking at let's say a 30 meter raster so this is what you'd call a 30 meter raster spatial resolution of 30 meters what does that mean that means that every single one of those cells remember they're all the same size is representing an area that is 30 meters by 30 meters on each side so in this case that means that every single one of the cells in this raster would represent a location with an area of 900 meters square so I would have to represent everything that has to do with the feature phenomenon that I'm representing in this particular raster with one value over a 30 by 30 meter area if I have a 30 meter raster and raster's do come in varying spatial resolutions and that's why you have to be able to look that up or know where to find that information when you download a raster or otherwise acquire one or when you make one it's important to include that information the metadata when you exchange it with somebody else I do want to note that all of the rafters that I'm showing you here in this video lecture have been simplified I'm just going to be showing you a relatively small number of cells larger so that we can see them and we can understand what's going on as far as their manipulation goes but raster's often contain thousands of cells tens of thousands of cells or more and that's typically the way that we work with raster's when you're in software and you're downloading them but for our purpose of understanding the theory behind them I'm simplifying to much simpler raster's that we can we can visually see them think about when we were looking at the weather map that had all of the temperature across the United States we just saw a gradation of colors over there we weren't looking at individual cells across that map and that's because there were so many cells and they are so small so we don't typically when you're working in the software work with raster's with just what is this nine different here we're working with rasters with thousands and thousands of cells and that's why using computer systems to manipulate them computers are very good at math of course that's how they help us with projections for one thing or they also help us manipulate our raster's because we'll need to do mathematical operations across thousands and thousands of cells potentially or more and a computer will help us out with that and it's very happy to take care of that for us and we don't have to go in and look at the raster's cell by cell like this anyway just keep that in mind so if we have this simplified raster right here one value per cell then what kind of information would you like to store in that raster well we can use it to store all kinds of information do you want to store information about land cover or elevation or rainfall or the depths of the sea or ozone concentrations or population density or whatever we can represent all kinds of geographic phenomena using the raster data model so you choose what you're interested in representing and then what value you assign to each cell is sort of up to you now you're probably going to determine some methodology you want the computer to execute in order to assign values to all of these thousands and thousands of cells but the value you assign and how you do that is up to you and I will also point out that the meaning of the values cannot be determined simply by examining the raster you know I can fill up this raster that I see right up here with all kinds of numbers but what do those numbers represent are the numbers supposed to represent rainfall total or total elevation or does it represent the both imagery of a lake or is it population density there's no way that you can just determine what it is that is being represented or by these values just by studying the raster itself and that's why it's also important to make sure that you have metadata for this reason because the metadata has to accompany the raster in order to let somebody else know what it is that that is supposed to be representing and also how the raster was created because that's important as well we'll talk about that later let's just take a look for the purposes of our examples right now at land cover let's do let's look at the land cover say we want us to our land cover and a raster for our examples let's look at that in by also introducing this term here binary raster binary rafters are going to be very helpful to you in a lot of different circumstances they're very useful a binary raster here like I only have here is one in which every cell is only allowed to be coded 0 or 1 1 I suggest the technically binary in a broad sense any one of two values but you know traditionally we use zeros and ones and then also from a mathematical standpoint there are reasons why we would want to have 0 and 1 here they will talk about as we go into how to do raster processing and solving problems and answering questions with raster's 0 becomes very helpful in that so here we're looking at a raster that is storing land-cover information and we have one where water is and we have 0 where water is not so raster's a binary raster's of this sort are often used to record the presence or absence of some phenomena so here we're talking about water we're just saying 1 ok that means there is water there and then 0 no water and this gives us an idea about the distribution of water in the area of interest and of course again if I had a much larger raster here and had 10,000 cells or a hundred thousand cells and was covering a large chunk of the earth maybe at a spatial resolution of 30 meters think about the spatial resolution of this if I had a special resolution of 30 meters then every 30 meters I would go by and assign a value is there water or is there no water that's a binary raster let's look at a series of these so here is a series of binary raster's for land cover information maybe these rafters represent are located in the same area they're representing the same thing but here I'm interested in or land-cover information than just where's the water I've got water in this area I've got forests in this area and I've got grassland in this area and so I have three different binary rosters one that's showing water one that's showing forests and one that's showing grassland and then I can use these in my analysis and do whatever kind of mathematical manipulation I'm going to want to do with them I don't have to represent the data in this way though what if I wanted to move to a non binary way of representing it so here is a single combined land cover raster this is no problem I can I can rasters in general can store more values than just 0 or 1 if I want to move away from a binary format so this is not a binary raster and in this case I'm saying one means water two means grassland and three means forest so here I just have a qualitative kind of data scheme you know one that's just a label right I'm using that value 1 to represent water of course two doesn't mean that's twice as much as one in this a particular raster because 2 is just a label for grassland 3 is just a label for forest you can assign whatever colors you would like to your raster so we were looking at a color stretch earlier with temperature across the United States but I can go in and I can assign hey I'd like all of the cells with one to be represented or to be shown in blue and then I've got these two different color greens for grassland and forest and the computer will symbolize a raster this is the way that you symbolized raster's we've already talked about how do you symbolize vector data sense well this is the symbology that we can do with a raster data set so as you might expect just as the symbolization of vector data sets do not matter for analytical purposes likewise color of raster data sets the symbology of your raster data sets does not matter for any analytical purpose the analysis is done on the numbers on the values that are stored in each cell we don't often look at the numbers in each cell because like I said these raster's that I'm showing you up here are much simplified and enlarged so that we can see what's going on but we'd often don't just look you zoom into a raster and you don't see a value there you just see whatever color whatever symbology has been assigned and really you don't often see the it cells themselves either so if I were to symbolize this and this blue and green color scheme I might just look at this and this looks like what you would see if you zoomed into a much larger raster on your computer and we're taking a look at the the individual cells here but for our purposes right here I'll probably look at them a lot like this with the numbers up so you can see exactly what's going on in them okay well I guess I will leave that right here for this video and I will see you again in the next lesson [Music]

Channel: GeoMindz.com

Views: 1,427

Rating: 5 out of 5

Keywords: geography, GIS, GISci, Geographic Information Systems, Geographic Information Science, geospatial analysis, map, mapping, cartography, problem-solving, geoprocessing, raster, raster processing

Id: 7uu_fkHaGqA

Length: 16min 45sec (1005 seconds)

Published: Wed Dec 13 2017