Synopsis: DASCH’s photometric calibration against the ATLAS-refcat2 catalog uses g-band magnitudes, while the vast majority of DASCH photometric measurements are in the photographic B band. Lightcurves based on the ATLAS calibration can exhibit false long-term features stemming from the fact that they are not in the “native” system of the plates. The APASS calibration should be used for analyses relying on precise long-term photometry.
Details
DASCH currently provides photometry calibrated against two different “reference catalogs”. The apass calibration is based on the AAVSO Photometry All-Sky Survey (APASS), data release 8, which is based on B-band magnitudes. The atlas calibration is based on the ATLAS All-Sky Stellar Reference Catalog (refcat2), using its g-band magnitudes. (There are several other historic reference catalogs available, but they are no longer maintained and they should not be used outside of special circumstances.)
The definition of the B-band magnitude system used by the APASS reference catalog traces back to photographic plate collections like the Harvard plates themselves. This calibration yields excellent long-term stability in the resulting light curves.
The g-band magnitude system in ATLAS, on the other hand, is based off of measurements made with photoelectric detectors (Thuan & Gunn, 1976). The g-band filter profile is similar to that of the B-band, which in turn is similar to the spectral response of a blue emulsion, but it’s not identical.
Empirically, we find that DASCH lightcurves based on the g band calibration can demonstrate false long-term trends, discontinuities, and other such features. This appears to be due to changes in emulsions, and perhaps other aspects of the photographic observing apparatus, over time.
If your science depends on the analysis of long-term trends in lightcurves, do not use ATLAS. Use APASS.
This raises an obvious question: why do we offer the ATLAS calibration at all, then? The answer is that compared to APASS, ATLAS offers superior completeness, depth, and astrometric quality, thanks to its combination of various modern datasets, especially Gaia. For science cases that do not depend on long-term photometric stability, ATLAS may be preferable.