Hi Nikos/devs,
Currently, I am Co-supervising a student who is using i.landsat8.swlst module for his bachelor thesis.
During that work we stumbled upon some minor issues where we would like to propose some changes in the module:
-
The k-flag does the opposite of what the description tells. It sets the computational region to the Landsat scene. I would say, it is more in line with other GRASS modules if the computational region is not changed by default. So I would rather change the description than the behavior (please find a suggestion in the diff below). However, maybe it is even more coherent with the rest of the GRASS modules if the flag is removed?
-
Currently, the module always overwrites an existing MASK and requires either the QAB band or a raster representing unreliable pixels which the module then uses as an inverted mask. I would like to propose to leave the (cloud) masking out of the module and leave that to other modules (i.landsat8.qc, r.reclass, r.mask…).
Please find attached a diff covering the suggested changes in i.landsat8.swlst plus some message and UI cosmetics.
Nikos, if you prefer a PR on github I can create one (don`t know how you sync gtihub and svn…)?
If you agree in the proposed changes I will also check and adjust the manual accordingly…
Kind regards,
Stefan
(attachments)
i.landsat8.swlst.diff (9.65 KB)