Modis takes daily images at coarse resolution (500m?) that are then lumped into 16-day composites.

LandSAT, let's say 8, takes roughly bi-weekly images at 30m (100m resampled to 30), and those are then lumped into 32-day composites.

Clouds aside, would you generally expect the two for the same time period to match? In other words, if you aggregated the Landsat pixels by mean, ought you to get the same value of the corresponding Modis pixel?

Probably not, right? Could you explain why? Is it the 16 vs 32 days? The world changes in two weeks? Equipment differences? Resampling algorithms?

Are there any Modis-LandSAT pairs of data that actually would match quite closely?

  • $\begingroup$ Note that one of the layers in MODIS NDVI composites is the day from the 16 day interval corresponding to the value that made its way into the product. Therefore you could select the pixels that were taken on the same date or near the date of the Landsat image. $\endgroup$ Apr 3, 2017 at 12:11

1 Answer 1


I can't comment yet, but I want to share this thought:

One challenge of the comparison is the co-registration of the data, so that really the same area on Earth's surface is compared.

In this article about the difference of MODIS and Landsat 7 NDVI, the authours put much effort in the co-registration to really compare the same area on the ground (but still there was a difference between the two sensors)


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.