[MUSIC] Hello everyone, and welcome back. In this lecture, we're going to continue our lesson on coordinate systems and I'm going to cover datums, which are foundational to coordinate systems and projections in GIS. You'll also learn how we model the earth's surface, and different datums we use in practice, that underlay the coordinate systems that you will use. First, for those of you who get distracted by poor grammar, it is pronounced datums, not data in this case. Besides being confusing, if we called it data, datum is a name so we pluralize it as datums. So first, what is a datum? It's the known locations to which we reference our geographic information. Remember that coordinates are meaningless without knowing where they are relative to. So this can take a few forms. It can be a point or a set of points which is common in surveying, or it can be a continuous surface which is common in desktop GIS in general. Many times, those surfaces are developed from those known points that you get from surveying. Some of you might be thinking, why do we need some sort of surface model with known points? If the Earth is a sphere, can't we just use a sphere? I have big news for you. The Earth, despite what you've probably been told, isn't a sphere. At least not a perfect one. It's slightly wider than it is tall, something we call flattening. Much of this is from the rotation of the earth. Angular momentum forces mass outward toward the equator, and our sphere shrinks and becomes more akin to an ellipsoid, which is to an ellipse what a sphere is to a circle. Spheroids and ellipsoids are foundational to datums. They form the basis for our models of the earth's surface that we use to build coordinate systems. With these, we can build ourselves a datum that incorporates local changes in elevation, or the lumpiness of the earth, to make our coordinate systems even more accurate. And just like we'll learn about for projections in general, datums are tuned for specific uses. Some of them are good for the entire planet like WGS84 or the World Geodetic System while others are good for certain regions like NAD83 or North American Datum. This tuning comes because we don't just have some perfectly accurate 3D model of the Earth that we can use for everything. Instead you can think of a Datum as being built of mathematical curves that attempt to approximate the surface of the Earth. This limits the specificity of the model. But increases the ability to transform between different versions of these models. By localizing a datum to region, the curves can be tuned to that area at the expense of the others that aren't designed to be represented by that model. For example, the North American Datum is tuned for accurate representation of data in Canada, Mexico, and the United States. But would poorly represent data in the Indian Ocean on the other side of the earth. It can represent data there, but not as well as it can represent data for North America. In contrast, WGS84, which I mentioned earlier, is designed for use world-wide, and makes a different set of compromises to represent the whole world well, but no one area perfectly. So in a quick review. The Earth is a lumpy and imperfect sphere. And if we want to represent it by using an ellipsoidal model. Our model will be closer to the true surface in some spots than in others. We also have vertical datums. So far we've talked about horizontal datums. Which goes to our x and y on the surface of the Earth. But what if we want that z for 3D information or elevation? If horizontal is hard, in my opinion vertical is harder because what defines the zero point? The sea level changes. We often talk about mean sea level. But if sea level changes every day, how do we define that network of known locations that build our vertical datum? And air in the Earth's surface models creates additional error in the vertical datum. So now that we know what a datum is, let's talk about some different datums that you may use. There are many, many that cover the entire world in different regions, but here are a few that are common. I'm not as familiar with some of them since I've done most of my GIS work in North America. And so, I'll start there. To begin with, there's NAD27, which is the North America Datum from 1927. It's mostly historical and covers older data sets but you do get lots of data sets that come from that era or the era that that was valid, and that you'll need to import, and convert, and translate to modern GIS data sets. There's also NAD83, the North American Datum from 1983, which is a revised version of NAD27. And that's a current modern datum that we often use. And then in the vertical datum sphere, there's NAVD88 and NAVD1988, which the North American vertical datum that you can use to reference your 3D information too. Now, worldwide there's WGS84, the World Geodetic System from 1984 that I mentioned before, and that covers the entire world pretty well. Where NAD mostly focuses on North America. And WGS84 is used in GPS. So whenever you get GPS coordinates those come out in WGS84, and if you want to use them in conjunction with other data from NAD83, or something like that, you will need to convert it. Now, in other parts of the world, again, which I'm not as familiar with, there is ED50, for Europe, which is the European Datum of 1950. And then in South America, SAD69, which covers basically all of South America, and is optimized for that region just as ED50 is optimized for Europe. I mentioned before that if we want to analyze data that's in one datum with data in another datum, we need to convert between datums on at least one of those data sets. And this is a bigger task than converting between different coordinate systems that reference back to the same datum. Because we need a transformation to convert data relying on one datum to use in another datum. A transformation is a mathematical algorithm that can convert directly between datums or to a common coordinate format that can be used to determine the correct coordinates. And in order to do that transformation, we need to back out any projection we might have to the geographic coordinate system, and then transform the datum from one to another. So, why am I telling you all this? Why should you care about the datum if you're going to use the coordinate systems built on top of it? And the answer to that is that the difference in datums matters, not just in theory but in practice. The difference between two versions of the North American Datum can be as much a hundred meters within the contagious United States. That is the same coordinate in both datums can be as far as 100 meters apart on the ground. That's huge. And can make a big difference in all sorts of applications. In even more practical terms, getting a pair of decimal degree coordinates, say 94.81 degrees and 33.48 degrees, isn't as absolute as it seems. To put those coordinates into use, you'll need to know the datum and the geographic coordinate system built on it just like you need to know your map projection and coordinate system for all datum. So to wrap this all up. In this lesson, we learned about how we model the Earth's surface in order to establish coordinate systems based on it. We use ellipsoidal and spheroidal models as the foundations for datums which are tuned for specific regions of the planet, or the entire planet. And then common datums in use vary by region so you should find out which one is most common and accurate where you work. Okay, see you next time where we'll start building geographic coordinate systems on top of our datums.
