Does fogpy consider different cloud classes? Should it?

[gerritholl]

From Cermak (2006), §4.2.5, “an entity is a set of adjoined pixels belonging to the same cloud class”. I'm not sure what type of classes Cermak (2006) uses, but fogpy seems to only use cloud or no cloud:

https://github.com/gerritholl/fogpy/blob/2c9aa9436e4924f55879d52a9de14817d370df5a/fogpy/algorithms.py#L993-L994

In my interpretation this differs from what Cermak (2006) describes. But maybe the implementation is more practical than the theory in this case, because neither the DEM CTH approach (Cermak (2006), Equation 4.29, or https://github.com/gerritholl/fogpy/blob/2c9aa9436e4924f55879d52a9de14817d370df5a/fogpy/algorithms.py#L680-L683) nor the lapse rate approach (Cermak (2006), Equation 4.30, or https://github.com/gerritholl/fogpy/blob/2c9aa9436e4924f55879d52a9de14817d370df5a/fogpy/algorithms.py#L684-L691) will work if one type of cloud is immediately neighboured by another. But it does seem an aspect that leaves room for improvement.