I am trying to determine the change/reduction of shorelines using Sentinel 2 data using ArcGIS Pro 2.8. However, I don't have a lot of experience in raster analysis and I'm a bit at a loss... how would one approach this? I'd like to:

  1. Select/filter pixels to only the shoreline.

  2. Determine which pixels should be counted as "shore"/"land".

  3. Apply the analysis to images from six-month intervals to determine change over time.

enter image description here

  • 3
    What is magnitude of changes you expect? How to account for tide? Is resolution of your blurry images even close to expected changes? If you don't have answers, it will be a waste of time, machine learning or manual digitizing.
    – FelixIP
    Aug 17, 2021 at 21:00

1 Answer 1


This would call for a classification analysis. You could perform a classification analysis to determine waterlines (water vs. ground) for each year. There is a Classify Pixels Using Deep Learning tool available in ArcGIS Pro.

Then you can compare the rasters from each year (probably using Raster Calculator) to detect change. Alternatively, you could convert the classified raster to a polygon in order to compare polygons.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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