I am currently looking at Tree Cover Change around a set of 288 villages in the Equatorial Forest of Africa using QGIS.

In order to do so I am using the GFCC30TC: Global Forest Cover Change (GFCC) Tree Cover Multi-Year Global 30 m V003 dataset. This dataset contains data for the years 2000, 2005, 2010 and 2015. Please find information on the data here:


I am basically using a Difference Image Approach, subtracting images band pairs for each year (2000-2005), (2005-2010) and so on to highlight changes.

However, the results I find are biased because the images for 2005, 2010 and 2015 contain "stripes" or "banding" leading to biased calculations. Here is an example in the picture below (raster of change between 2000 and 2005).

enter image description here

The problem for me is that pixels in the stripes take the value of 0 and thus bias my analysis.

Does anyone have any solution to overcome this issue?

The best for me would be either to vanish the stripes or to find a solution to set the stipes value to missing and to account for this in the calculation.

As I am not a GIS expert (by far), maybe some of you know any other satellite imagery database that do not have such stripe issue.


This sounds like it's using Landsat 7, which had a Scan Line Corrector (SLC) Failure. The Landsat website describes the issue in further detail and has resources for filling the gaps.

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.