Hi,
I was going through the landsat-c2-example.ipynb notebook and noticed that the NDVI is calculated directly from the red and nir08 bands as loaded via odc.stac - without applying the documented scale and offset.
My understanding of odc.stac is that it does not apply scaling by default, so the computation as in the example notebook seems to be using the raw integer DNs. Doesn't this bias the NDVI? Is there a specific reason why the example computes NDVI on the unscaled data - or have I misunderstood anything?
Thank you for providing and maintaining these examples, they've been super helpful for getting familiar with MPC!
Best
Maike
Hi,
I was going through the landsat-c2-example.ipynb notebook and noticed that the NDVI is calculated directly from the red and nir08 bands as loaded via odc.stac - without applying the documented scale and offset.
My understanding of odc.stac is that it does not apply scaling by default, so the computation as in the example notebook seems to be using the raw integer DNs. Doesn't this bias the NDVI? Is there a specific reason why the example computes NDVI on the unscaled data - or have I misunderstood anything?
Thank you for providing and maintaining these examples, they've been super helpful for getting familiar with MPC!
Best
Maike