I just started playing around with stacking NDVI images from different years into one RGB value and am unsure how to read the photos or analyze them. The image on the left had R 2011 G 2011 and B 2003: I'm assuming the light grey and white means high vegetation/little change, the black is low vegetation/little change. Does the blue mean change from 2003-2011? How about yellow?. Would I use band math to subtract the bands to find the difference.

The image on the left has R 2011, G 2007, and B 2003. Again does the green in the image show most change in 2007 vs blue is most change in 2003? How do I calculate statistics from here?

Thanks for any help that can be given! Inline image 1 enter image description here


Basically you display your image stack in RGB, and by default R is for band 1, G for band 2 and B for band 3.

In your case, you have 2011 in R+G and 2003 in blue.

So you are right when you say the scales of grey mean "no change" and the brighter the more vegetation at both dates.

Now, let's look at the change: in additive color, R+G = yellow. This mean that yellow indicates high NDVI values in 2011 and low NDVI values in 2003. In other words, vegetation regrowth. Conversely, blue indicates a decrease of the NDVI. If you add a 3rd date (the maximum that you could visualize), you should also take care of the other combinations (R+G= magenta, G+B = cyan)

You can of course substract the NDVI of both dates to summarize the information in 1 band (easier to work with), but selecting a threshold for the change will be necessary. You could also select some training samples and run a supervised classification for the 4 classes.

As a remark, be careful with the interpretation of NDVI change, because it could also be due to clouds.

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.