[MUSIC] Hello again, and welcome back. In this lesson, I'm going to cover raster mosaicing, which is the process by which, we can take many different rasters and turn them into one raster. Think about how rasters are very data with many rows and many columns of data points, we can get very large files very quickly. What helps us solve this, because many people and many distribution centers like data download sites like to divide up rasters into regions, so you download only the areas you need and it makes data management easier. Well, once you download all these different sub parts, these different regions, you need to put them back together again. And raster mosaicing is how we do that. So first, I have these folders here, where I unzipped six different data downloads that are digital elevation model files from the national elevation dataset, or NED. For part of New York State. And inside each of these, I have a bunch of different files. I have a shapefile containing some boundary information. I have an Imagine image file. And, then, I have thumbnails, and some metadata, and Other Read.me information. They come with a lot of different pieces but what I really need from the spatial side is this image file, and if you have forgot that imagine images were raster format, even just the size tells me here's the spatial data. So lets drags this into Art Map and Art Map will load it. And let's take a look at it. So, we have this practically square digital elevation model. For some area. We could add a base map and see, but let's not right now. We can also add our other layers. If we bring back up windows, I can drag these other files in here. Go back, drag the rest of these in. These are all independent files here. Occasionally, they come in slightly different folder structures but the matching image is what I really want. The dot OVR file here, as a side note, is the pyramids file that gets built by which helps it display rasters at different resolutions more quickly. It's sort of like an index that is uses to display rasters as I zoom in and out. So, then if I add the rest of these, still. And then I zoom out. I can see that it seems like I have six rasters here. And I can tell, even though they're basically seamless from perspective of data coverage, there are scenes from the perspective that they're each being scaled relative to their own minimum and maximum values. So I get this bright white here, even though the values at the edge here are probably similar to the values here. And that's because, they all have a different low and high values that they're getting symbolized by. Now, I can also, right now, turn one on and off. So, I can hide them and view them differently. But what if I want to analyze data that crosses some of these boundaries? I don't want to split put it up and analyze it against multiple rasters repeatedly. That's error prone and needless effort. Well that's Raster Mosaicing will help put these rasters together into one that I can use for my analysis. To do that sort of counter intuitively, I need to first create a mosaic dataset so an art catalogue or the catalogue window. I'm going to create a new geo database just so you can follow along with me, but you can use an existing geo database as well. So I'll right click. I'll go to new file geo database, and I'll just call this my digital elevation model database. And in any geo database, even one that already has data, I can right click on it and go to new, Mosaic data set. If I click on it, I get a new geo processing tool popping up saying, create a mosaic data set. And I can give it a name and I'll call it, Ithaca, New York because that's the area that's approximately covered by this And then I need to specify a Coordinate System and this is the Coordinate System for all of the items in mosaic dataset. So let's check out what that is. I'm going to right click on one of these and go to Properties and go to Source and see that these are in the Geographic NA 83 Coordinate System, so The North American 1983 Coordinate System, and I can find that when I go searching for it. And I know, in this case, that this one is not 1983, but let's try searching for the exact string that it gave me in the data properties, and that way you can find other ones as well, if you have different rasters. So, it's gcs underscore north underscore american underscore 1983. So, let's try that. Gcs underscore north underscore American underscore 1983. I search and it gives me four. And it's probably, this one. But if I really wanted to confirm, or I could do this another way, I could see that under the Layers option, it's telling me that this particular coordinate system is in use by one or more layers in my document. So we'll use this instead. This. And under product definition, we can tell RGIS a little bit about what kind of data we're storing. A lot of this has to do with aerial or satellite imagery. And so, we're just going to use none. And that indicates that there's no band ordering, because we just have one band in our rasters. And for now, we'll ignore the rest of the features in here. But, if you start creating more advanced raster Mosaic datasets, you will want to use these. And we'll click OK, to create our new mosaic dataset and it runs in the corner like all our other tools. And now inside our geodatabase, I have an empty raster mosaic dataset which is displayed as this sort of cascade of rasters here. I also get In my table of contents a new group layered of the raster mosaic data set which has the images themselves, the footprints, and the boundaries. But again there's nothing in it because, if I turned everything else off, by control clicking on it to turn everything off then turning back on, there's still nothing. And in fact, zoom layered I go all the way out to the whole world in the scale up here. I'm not seeing anything so let's zoom back to these layers, and let's add some some to our mosaic. So, there are two ways we can go about this. First, we can go to arc toolbox, and under the data management tools in Raster. And then, the mosaic dataset tool set, there are a bunch of tools for managing mosaic datasets here. And there's one called outer [INAUDIBLE] mosaic dataset. For now though, let's keep up the trend we're using in the catalog window. And I'm going to right click on it, and go to [INAUDIBLE]. And, surprise surprise, it pops up the [INAUDIBLE] mosaic data set tool. And one thing that's cool, is it gives me the option to provide a work space instead of individual files. Work space being a folder, or a geo database, or something like that. So, if I've already stored all of my [INAUDIBLE] in one spot, I can just provide that. And it will load all rasters from that location. Or I can choose to provide files or datasets. Where a dataset is much more like the rastors we're used to. So since we are using rasters in our table of contents, here I'll select data set and then lets drag all these over on to the source area here it's given me a warning it says data conflict between input and output. For now, let's ignore the warning and see what happens. We'll want to be aware that we had a warning, but it may be just fine. I know that I selected the coordinate system from at least one of these layers, and I believe all of the layers. because it only gave me a single coordinate system from the layers in the map document. So I think, I should be fine since the Mosaic data set and these layers should have the same coordinate system, but let's see. And I can have it update the cell size ranges, update the boundary and update overviews. I'm going to select those, because overviews provide us with a lot more Quick information about our Mosaic datasets. If I don't do that, I may not get to see the Mosaic dataset mark map very well. And then, I can specify things related to generating pyramids and a bunch of advanced options that we're going to ignore for now. And I'll click OK to add those to the Mosaic dataset. And it's running. And it can take a little while, but when it's done, notice that it pops back up in our map document, even though I haven't turned these layers back on. Our mosaic data set now has that information and I can turn off the images and still see the footprints and then the boundary if I turn that on will show up as well. So that's the overall mosaic boundary, and then I have the individual raster footprints that I can see. And then, I can see the images if I turn that on as well. And also, as long as I calculate statistics when I generate the mosaic, tells me my new high and low and notice that we now are symbolizing on one continuous basis between all these rasters. They are behaving as one raster now, because they effectively are. I can use this mosaic data set in most places that I can use a raster. Raster in ArcGIS. I can also export this mosaic data set now to a new raster using geo processing tools. If we had time, I would show you that, but for now just know that there are geo processing tools in here and the tool that let's us do that is down under Raster Dataset. It's the copy Raster tool that we'd used to Copy Rasters from any other format to any other format and we can select as it says a mosaic dataset that we want to copy and select a different Output Raster format. Another thing I want to point out that we're not going to cover in this particular lecture is that mosaic datasets aren't just useful for merging distinct pieces like this that already don't overlap. One of the main reasons you might mosaic data is, if you have overlapping imagery or overlapping sensor data. And mosaicing, helps you resolve conflicts in between your overlaps. You set rules of which data to use in the overlap, and you get a seemless raster out of potentially overlapping rasters still. That's a big use case for mosaicing. So, lets take just a quick look at that conceptually from ArcGIS help. So, what we just did was more akin to this top part here. Taking a bunch of different pieces, in fact, the same number just by chance, and turning them into a single raster data set. What I was just talking about with overlapping is taking images like this, where we have some overlap between these images. There's green parts here as well and this little feature here is over here and turn them into a single seamless raster. That's another use case from Mosaicking. We'll probably touch on that a little more in the next class, when we work with imagery more. But just know that that's a possibility. Keep that in the back of your mind for that day in the future, where you need to use it. For now, just remember that mosaicing helps us merge rasters into a single data set. It's a big topic on it's own, and one I encourage you to explore on your own if you need it. But at the very list with what we've shown you here, you can get started with it. Okay, that's it for this lecture. In this lecture, I showed you how to take multiple rasters and turn them into a single raster by the process of mosaicing and creating mosaic data sets in arcGIS. I hope you find that to be an interesting and useful tool. See you next time.
