I want to calculate deforestation using Landsat 2 images. Right now I am using the raster calculator to generate a NDVI image and then the reclassification or Iso Cluster Unsupervised classification tool. However, since they are old Landsat images, a lot of them have clouds and it messes with the final deforestation image (difference between classified images).

Is there a way I can get rid of the clouds using the same Landsat for different dates in the same year?

What I am thinking is that it may be possible to overlay the rasters and somehow get an optimal raster for the year, using input from different dates.

Does that make sense?

Is there another way?


Clouds are serious problem in optical remote sensing, but using other images from different dates to substitute the regions with clouds will reduce the accuracy of your work, especially if you are calculating NDVI which depends totally on dates (seasons).

However, it might be acceptable if the difference between cloudy images and non-cloudy images is just few days as you are still in the same season. The more efficient way to keep your analysis in safe side is when you classify your images you put a class named 'cloud' that can be excluded from your calculation, or mask the clouds from you images and do the NDVI and classification calculations with the masked clouds images.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.