Personally I would think that if it occurred it would be unusual since the usual Roman naming practice seems to have been to name for an individual rather than a feature.  However, the Roman world covers a long time period and had a great deal of local variation so it wouldn't surprise me if such names were adopted unofficially.

Fully agree with T. E. Jones, it seems to me an answer that must be given casuistic form or to give a neologism like license of the language.

Thank you both! I did not mean to use it as a neologism, but trying to understand a medieval place name that recalls a classical name, but Timothy E. Jones' point is helping.

In that case it is legitimate whenever the reader is warned that it is a retrospective use of the medieval toponym.

Thank you

How certain are you that it recalls a classical name rather than it being a current medieval usage? 

It is a city name mentioned by classical sources, but lost as city name in medieval ages. It is only attested again in the 13th cent. with the form that I was suggesting before, but naming only a country area.

My question would be, "When did it become part of the Empire?". Certainly, the foederati, colonii, and civitas out on the periphery, such as the Rhenish Limes, had some funky town names, because they were incorporating the Gauls, Belgae, minor Germanic 'volks', and other peoples. They were aware of multiple cultural names for the same place (i.e. Latinic/Roman, Celtic, Germanic, and possibly even Slavic).