 In this video, I'm going to demonstrate how to calculate NDVI from a Sentinel-2 image and how to reclassify the result into vegetation classes. We can download Sentinel data from different websites here. I use the USGS Earth Explorer website. And you start by indicating the area. You can move the map to the study area and you can use map and then it will make a little polygon for that area and that will be the search area. In this case, I'm not interested in that area. You could also put coordinates or you can click to get a polygon. But what I'm going to do is to upload a zipped shapefile with a boundary polygon. And here I have the boundary. And then I can choose a date range. Let's say from 2020 April to 2021, first of April, I go to data sets. And here I can go to the Sentinel and choose Sentinel-2. And then I can go to results. And there you can check the footprint, you can check the metadata, or you can download after logging in. I'm logged in. And to have the full image, it's in the JPEG 2000 format. You choose this one. If you just want a picture, you choose the GeoTIV. In our case, we would use that one after choosing the correct one, which has the least clouds, etc. After downloading and unzipping the file, I find these folders and the data can be found under granule. And then the name of the tile and define it here under IMG underscore data. And I'm interested in all these raster bands, so I'm going to select them and drag them to the map canvas. And here we see them. And we can see that they are already projected to UTM zone 31 north. And that they have pixel size, in this case, for this band of 60 meters. And this band has a 10 meter resolution. So I'm going to make a map stack out of this to create a multiband raster. And I don't need a band 8a, so I'll just remove it. The quickest way to make a multiband raster of these individual bands is to make a virtual layer. So go to raster, miscellaneous, build virtual raster. Here I can select the input bands, select them all, click OK. And you can choose here which resolution you want for the output. And let's take the highest resolution, so the 10 meter, and make sure that it puts each layer into a separate band. If you have this unchecked, it will mosaic tiles, which are distributed in space. But in our case, they're of course stacked on top of each other. Therefore we check this box. And we keep the other things as default. And then I save it to a file. And then I run it. And this is very quick, as you can see. So I remove all the previous layers. And there we have our multiband raster. So we can use the layer styling panel, and it recognizes a multiband color. And here I can choose, for example, to have near infrared band 8 visualized in red. The green band in band 4, the red, and the blue band 3. And here we see that all the red parcels are vegetated. We can distinguish the forests, et cetera. You can play with these different band settings to see different features in the landscape. And we are going to use this now to calculate the NDVI. So what I do, I go to Raster, Raster Calculator. And here we see all the bands. And NDVI is for Sentinel is the near infrared is band 8. So this add symbol separates the name of the layer with the band name. So this is band number 8 minus the near infrared minus the red. So that is band 4 divided by the near infrared plus red. Let's save it. Let's call this one NDVI.tiv and we run it. There it is with values between minus 1 and 1. Let's style the layer, the NDVI, single band pseudo color, and let's choose greens. And here we see that the more green it is, the more vegetation you can play with these settings to get more contrast. If we want to know which areas are vegetated and which are not vegetated, we can simply separate them using the Raster Calculator. Because NDVI values that are larger than 0 are considered vegetation. So we create a Boolean layer and we call this vegetated.tiv and then I run it. There it is. And I use palette.uniquevalues and I say classify and then we can make zeros blue and the vegetation green. And here we see that it's not very informative, but we can get the lakes out of this. So let's make some more classes and we do that by going to the processing toolbox and use reclassify by table and we take the NDVI and we can make a table where we say from minus 1 to 0 it will be no vegetation, call it class 0 because I can only store values in rusters from 0 to 0.1. Let's call it bare little vegetation, but I can only put values there 0.1 to 0.25, it's like a medium dense, 0.25 to 0.4 and 0.4 to 1 click OK and I want the lower value to be part of the range and the higher value not to be part of the range. And let me choose here the integer 16 because we have some no data there that we want to use and I save this to a TIFF file and I call this vegetation classes. There we have it. Let's also style this layer, palette.uniquevalues, classify, here we are and then I'm also going to use greens and this gives us a nice classification of the area in terms of vegetation cover. I can change this like no vegetation, bare, low vegetation cover, medium vegetation cover. There we are.