 Hello, in this video we are going to geo-reference a scanned map. This video is part of the second edition of the book QGIS for hydrological applications from LocatePress. When you go to the data folder of the course data, you can open the JPEG in external viewer by clicking right on it, and you can use any picture viewer for that because the JPEG is not a GIS format yet, and if we zoom in on the map we can see that there's a coordinate grid with connected coordinates in Easting and Northing, and that we can use the nodes of the grid to sample so-called ground control points to geo-reference this map. You can also see here on the map that there's a description of what projections are used, and it says here that the projection and 1000 meter grid is zone 18 of Universal Transfer's Mercator, that's UTM. It also says that there's a 1927 North American datum used. The other information is all kinds of transformations or other coordinate systems printed on this map. Here we can look up these keywords at the spatialreference.org website. We use the keywords UTM 18N and NAD 27, and there we find that the result EPSG 26718 corresponds best to our search. We need to write it down or copy it to be used later in QGIS. Let's go back to QGIS and from the raster menu we choose the geo-referencer. You can maximize this new window because we will work in this window for the next minutes, and we choose this first icon to open our scanned map, the JPEG image, and here you see the full map, and the first thing we need to do now is to change the transformation settings. First the transformation type. These are the calculations, the equations used for the transformation with the ground control points that we give. If you don't know what to choose, start with a linear one, which is a simple rotation and scaling. For the resampling method we choose cubic because we want a smooth backdrop. If you want to do calculations, choose the nearest neighbor. We also choose the output coordinate reference system, which is the EPSG code that we looked up with spatialreference.org, and that's the projection that belongs to the grid, and we can see that indeed this projection is used over our study area here, which is Mount Marcy, New York. For the rest we keep the defaults. It will, by default, calculate a map with the file name underscore modified dot tiff, and we check the box to load in QGIS when calculation is done. Then we click OK. So now we can add our ground control points. So I zoom into a node and I read the coordinates that are printed on the side of the map. You need to zoom in very well because the more accurate we do this, the more accurate the results will be. You can also delete and move the points with these icons. So I add a point. I can still use the pen and zoom buttons or the scroll and the mouse, and I simply type the coordinates of the node that I have added here, and it plays this red dot. In the table below you can find the file coordinates in the source xy, the destination xy that is in the projected coordinate system, and some error statistics which are not yet available. We will see later when they come available. I place the second point on a node in the upper right corner, and here I also type the coordinates that are read from the side. Note that here the zeros are omitted. So it's in kilometers, but of course we continue in meters, and then we click OK. Let's go to a third point in the lower right. It's placed here, also here read it from the side, and you see now this red line. That's an indication of the error. We have three points and then it can calculate the error. The more points we add, the more accurate this calculation is. It will cover up for errors or irregularities in the map, so normally we would choose more than four points. It's really a minimum, so here I add the fourth point, and again I add the coordinates, and now we have four points, and the red line is quite big, and it exists at all our points. So if you have one that is off, you can delete or move a point, replace it. Here in this case the points are correctly placed, but still the error is quite large. I can change it here to a first-order polynomial to see if it goes better. That's not only a rotation or scaling, but a fit of a polynomial, and we see now that the error is within the pixel. We can also see that here in the mean error, which is 0.7. That's good enough, so I'll perform the georeferencing, and it asks for a transformation here. If you see the screen and you don't know what to do, you always choose the first, the default one, but for some cases you might need a different transformation. These are different equations used to make the transformation of projections. Here we see the end result of the georeferencing, and we see that our map has been scaled and projected in EPSG 26718, which you also see in the lower right, and that is because it's the first map that we loaded this project. It will take that projection as the projection of the whole project, which is called the on-the-fly reprojection. Every other layer that we add will have that same projection.