Water Yield: per pixel interpretation.
I'm using the Water Yield model (in addition to other InVEST models) to determine which parcels within a given watershed provide more ecosystem services (ESs), including water production. One step of this study would be to assign levels of ESs to these parcels. However, the InVEST manual says that the output of this model should not be interpreted at the pixel level.
In this case, can I still use per pixel values and assign levels of water production to each parcel (some of them may contain only a few pixels) or I should calculate the average of water production per pixel within a subwatershed and assign this same level to all parcels within this subwatershed?