3

Using as an example Landsat 8 bands:

Does pansharpening band enhace only visible range and therefore can I use it only for analysis connected with visible spectrum? Or I can pansharpen every possible combination of raster composition so I can get an image with NIR/SWIR/Coastal/Cirrus included? If the latter is correct, will the algorythm perform some not-so-accurate 30->15m interpolation or it's 'pure' 15m? I see in the specs of L8 that Pan band ranges from 0.50 to 0.68 µm what covers G and R band what leads me to the first hypothesis, but I have seen some other compositions that were pansharpened.

2

According to USGS

Pan-sharpening is a technique that merges high-resolution panchromatic data with medium-resolution multispectral data to create a multispectral image with higher-resolution features.

The above technique can be applied to visible spectrum as well as any band combinations. The above link provides a tutorial on how to perform pan-sharpening on 5,4,3 band combinations which are SWIR, NIR, and Red, repectively.

1

In theory, it doesn't make sense to pansharpen coarse resolution data with higher resolution if the spectral ranges don't match. However, that doesn't stop you from doing it - and it generally looks reasonably okay in most cases.

As such, in theory, your first 'hypothesis' correct, but in many cases it makes sense to ignore that, and simply force the processing in order to create a pansharpend product.

  • Thank you, I guess my analyses don't require terribly high confidence data and I can simply keep using it. – adamczi Mar 24 '16 at 16:14

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.