How are DOIs managed for ongoing time-series?

I'm familiar with DOIs been allocated to historic time series, to give academics a unique, citable identifier for datasets. Similar to how a DOI is allocated to a specific journal article.

As the International DOI Foundation says:

A DOI name provides a means of persistently identifying a piece of intellectual property on a digital network and associating it with related current data in a structured extensible way.

What's the procedure for DOIs for datasets that continue to grow: for example, a time-series of external temperature data for an airport? Is one DOI allocated to the ever-growing time-series. Is there a DOI for each year's worth of archived data? Something else?