Hello! I am using snap 11 to map fires and burn severity with S2 images. I have an issue. When i use images of 2025 the fire is not obvious even when i make the RGB IMAGE. BUT when i try older images from 2024 eg. They appear fine. Any proposals?
Creating a Burn Severity map requires using scenes close to fire period. In essence, if the map you want to create is from a 2024 Fire incident, you need to use a scene maybe a week or few days before start of fire for Pre-fire and a week or few days after the end of fire for Post-fire. If you use a 2025 image for 2024 fire, you will not get any burns considering vegetation growth that has now covered the most recent captured image. If you need more details on how to deduce the dNBR from the images, do let me know.
Not really my issue. What i mean is: i have a fire on July 2024. I take an image pre and after the result is fine. When a try a fire eg March 2025 and follow exactly the same way as in the fire mentioned i get no result.
Okay, I get you now.
Have you tried a longer date? Maybe the Pre and Post images you picked have too much cloud cover.
You can reduce cloud cover to about 60/70 when trying to download the images and see the available scenes.
Remember also that in mapping it, you need to resample all required bands to 10m (20m is also okay but you lose finer details), i.e B8, B8A & B12
You can alternatively visualize the image true colour on ESA Copernicus browser to be sure there is a burn because I am working on a 2025 Burn Map and I have produced it from my Pre and Post selection images, and concluded using QGIS.