SIMBAD references

In synthetic versions of two broadband photometric systems, Johnson-Cousins and Washington, we find the dependence of atmospheric extinction corrections on colour and on macro features in the spectra of stars, such as the Balmer jump, as parameterised by T_eff, logg, and [Fe/H]. We use standard passbands, a mean atmospheric extinction law measured at ESO/La Silla, extended and modified by us, and the Kurucz library of synthetic spectra. The true broadband atmospheric extinction is far more complicated than any current reduction (transformation) methods consider. Hence all broadband magnitude systems are fundamentally unphysical - they contain not the extra-atmospheric magnitudes, but unobservable magnitudes whose relation to physical magnitudes is unknown, but may differ by 0.05mag or more for hot and cool stars. Hence, it is hazardous to compare them to any synthetic magnitude system derived from either synthetic spectra or spectral scans. These problems exist to a lesser degree in intermediate band systems, but narrow band systems are relatively immune from these complexities. We do not treat either kind of system here. If our results were incorporated into a photometric reduction program, and standard stars and program stars stars carefully selected by metallicity and luminosity class, a standard magnitude system could be established that would be directly comparable to synthetic systems. As a bonus, measurements of intrinsic flux variations at the millimagnitude level would become more secure. We describe our own operational photometric transformation program that incorporates only the linear part of the dependence on colour of atmospheric extinction. Our results and prescriptions are useful for aperture photoelectric photometry, but our implementation is aimed at CCD photometry of stellar populations.