This is something that was mentioned in a few meetings: we could clip the SDM result using land cover before showing the final map.
This would help having cleaner maps.
Please debate the scientific validity of that approach:
Brian Leung
Sure - you can analyze any subset you want.
If you clip it, you are constraining the analysis to the land-use types under question (e.g., forests)...which could be a fairer analysis, if the other land-use types are irrelevant (i.e., it'd be easy to get a good fit, if you're including irrelevant areas, but possibly doesn't advance prediction much, since everyone would already be excluding them).
Timothée Poisot
For the reason that Brian mentions, I would maybe let the user decide, by picking from a list of land-use types to use for clipping?
François Rousseu
I agree that having an option for clipping could be very useful. In most cases, it would have to be species specific.
This is something that was mentioned in a few meetings: we could clip the SDM result using land cover before showing the final map. This would help having cleaner maps.
Please debate the scientific validity of that approach:
Brian Leung
Timothée Poisot
François Rousseu