A couple years ago, @Mateusz_Konieczny spotted a number of systematic names on tributaries of Darby Creek near Philadelphia, such as “Trib 00808 To Darby Creek”. These waterways were imported from PASDA, a statewide GIS dataset. These names seem too unwieldy to use as common names for these streams, if they even have common names. Then again, not many people would have a reason to refer to such obscure creeks by name anyways; whatever the field scientists use might be fine to put on a map, as long as we expand the abbreviations.
There wasn’t much feedback in Slack or this note about how to deal with these systematic names, but @dannmer resolved Way: Trib 00808 (28347307) | OpenStreetMap by splitting it up into ref=Trib 00808 and destination=Darby Creek. Does this seem like a reasonable tagging style that can be applied across the rest of the import? It’s unclear to me whether 00808 is a standalone reference number or whether it needs to be combined with “Darby Creek” to be unique. As far as I can tell, these reference numbers aren’t posted anywhere, so it might be misleading to use the main ref key versus something like ref:pasda.
It isn’t totally unheard of for names to have cryptic numbers in them, especially for obscure irrigation canals, peaks, bridges, on-ramps, and boundaries. It’s slightly more surprising here because the creeks are natural and run through a built-up area, but I guess either no one has managed to find out a more natural name for the creek or there was never an effort to name them all in the old days.
I’m in favor of the 2nd option with the tag being modified to ref=Tributary 00808.
An example of the way I’ve been adding creeks is: https://www.openstreetmap.org/way/1254514021. This was not an import, but a trace using satellite imagery. I used the destination tag on the waterway relation.