In order to ease the comparison of fluxes between our source catalogs and reference catalogs, I propose that the catalog in a sky circle or pixel box instantiated from a reference object loader provide fluxes in the same units as our source catalog fluxes (Maggies, per RFC-322). Also, our reference catalog standards need to be clearly documented in some location outside the source code (pipelines.lsst.io?).

The on-disk representation can be in whatever units are most appropriate for that particular catalog. If those units are not Maggies, LoadReferenceObjectTask will convert before the catalog is provided to the user.

This is a change from current reference catalog practice, where LoadReferenceObjectTask's docstring says that all fluxes are in units of jansky:

    - *referenceFilterName*_flux: brightness in the specified reference catalog filter (Jy)

        Note: the function lsst.afw.image.abMagFromFlux will convert flux in Jy to AB Magnitude.

    - *referenceFilterName*_fluxSigma (optional): brightness standard deviation (Jy);

        omitted if no data is available; possibly nan if data is available for some objects but not others

    - camFlux: brightness in default camera filter (Jy); omitted if defaultFilter not specified

    - camFluxSigma: brightness standard deviation for default camera filter;

        omitted if defaultFilter not specified or standard deviation not available that filter

    - *cameraFilterName*_camFlux: brightness in specified camera filter (Jy)

    - *cameraFilterName*_camFluxSigma (optional): brightness standard deviation

        in specified camera filter (Jy); omitted if no data is available;

        possibly nan if data is available for some objects but not others

I believe the only impact of this RFC is direct users of reference catalogs: single frame calibration and meas_mosaic (and jointcal, which is what provoked this RFC). As such, this could be implemented before RFC-322, but is probably best implemented simultaneously with it.