dry river

[cmungall]

is a dry river still a river?

The definition says yes:
A river that has either permanently or temporally lost its water. MA:ma

But the is_as say no.

I suspect it’s the def that should be changed. perhaps this should even be a def for dry rived bed

Original comment by: cmungall

[pbuttigieg]

This is a really ugly spot.
Part of it comes from the very weird definition which messes with the idea of 'dry'. I think they probably mean 'reduced flow' or 'lentic'. The latter would violate the def of stream (and thus river).

I suspect it’s the def that should be changed. perhaps this should even be a def for dry rived \bed\

Agreed. I think we should obsolete classes like dry river and add it as a synonym for classes like 'dry river bed'.

[cmungall]

With my normal anatomical ontology prism through which I (advisedly or inadvisedly) view ENVO, I think of this as analogous to

cell = plasma membrane + cytoplasm
blood vessel = blood vessel walls + lumen/contents

With my ontologists hat on we need to clearly distinguish between the 'walls' and contents vs the mereological sum

With my practical hat on I concede it can be overwhelming to always name all members of the triad. E.g. in anatomy you end up with jejunum wall, jejunem, jejunem lumen repeated for all segments, and if you want to be really comlete jejnumem wall continuous with ileum wall, jejunum lumen continuous with ileum lumen etc....

[pbuttigieg]

in anatomy you end up with jejunum wall, jejunem, jejunem lumen repeated for all segments, and if you want to be really comlete jejnumem wall continuous with ileum wall, jejunum lumen continuous with ileum lumen etc....

Interestingly, I think there's value in this approach for ENVO: each of those class formulations using riverine entities would probably align with some working definition used out there. some would use river channel, others river stream (odd label, currently just river in ENVO, but suggests flowing water body), and others would need river channel continuous with river stream. It's a way to avoid 'sensu' labels.

I think I'll experiment with this for streams. If this is done for the superclass stream can it be auto-propagated to the subclasses?