Those who use Pan-STARRS data might bemoan the lack of a "u" band.
There is a quick fix, of sorts -- the POSSI-O band which is centered
on 4050A. While not exactly u, it is blueward of Pan-STARRS grizy,
and overlaps the g band only about 33%. The POSS-I survey was done in
the 1950's and it covers all sky down to declination -33, which means
it overlaps the entire Pan-STARRS sky which is down to declination
-30. POSSI-O can be calibrated to Pan-STARRS via the POSSI-E band
which is just Cousins R.

Lupton 2005 gives the following transform for SDSS r & i to Cousins R:


R = r - 0.2936(r-i) - 0.1439

and this works here because Pan-STARRS r & i are just the same as
SDSS. So get the R-E offset and add it to O, and boom, you've got
Pan-STARRS "o" to add to their "grizy" bands -- "ogrizy" now. This is
for the brighter objects, as POSS-I is about 2 magnitudes less deep
than Pan-STARRS.

There is one issue remaining, though, and that is that the POSSI-O and
POSSI-E bands must be correctly calibrated to each other.
Unfortunately, the USNO-B has bad offsets there, and USNO was the only
place which digitized the Galactic POSS-I plates. There is a history
there, USNO-A1.0 was USNO's first release which digitized the entire
POSS-I survey and it was well-calibrated. But shortly thereafter they
released USNO-A2.0 for the purpose of better astrometry, but some
nasty photometric offsets got introduced also, and the POSS-I coverage
was only down to declination -12 (replaced by UKST southwards of
there). These bad photometric offsets were passed onwards to the
USNO-B catalog, unfortunately.

A while ago I released the ASP catalog (2017,PASA,34,25) which
presents the whole POSS-I data (375M sources) calibrated correctly to
the USNO-A standard and also to the APM (which covered POSS-I away
from the Galaxy) standard. The calibrations are shown on the web
pages:

http://quasars.org/APM-USNOB-plate-calibration.txt for USNO-B, and
http://quasars.org/APM-USNOA-plate-calibration.txt for USNO-A1.0

So the ASP catalog gives the full POSS-I data with POSSI-O calibrated
correctly (or at least better than anywhere else) to POSSI-E. So that
data can be mined for "Pan-STARRS o", using the above technique.

Eric Flesch