Hi,
I am quite new to satellite imagery and currently exploring how I can use satellite data to detect straw/crop residue burning in a specific area and ultimately to track over years. Should I be looking at the Burned Area indices like Normalized Burnt Ratio (NBR) or more into NDVI? I think Burned Area Indexes are more for large area fires like forest fires, might not be suitable for agricultural fires? If NDVI is more suitable, how do I set the threshold so that the imagery shows burnt/unburnt area? I would like to know if there are other suitable indices that I should look into.
Hi! Welcome to the forum and thanks for the question. It already sounds like you are on the right track and have some knowledge on the right indices to use to detect your burnt areas. I don’t have enough domain knowledge to know if these indices are still suitable for agricultural fires but I don’t see why not.
To help you on your way, I’d recommend:
- checking out the custom scripts repository. There is even a section focussing on wildfires. This should give you plenty of resources to start experimenting with the data.
- Watch some of our webinars which have focussed on wildfires and multitemporal analyses
- Other than, you cannot go wrong with a literature review and see how the community has tried to solve the problem you are having.
As you are interested in the wildfires domain, I’d also like to share the summer school that we are organising this summer in Ljubljana, this is something I think you’d be interested in joining!
Good luck with your experimentation and data exploration.