2003AJ....125.2645C

We describe the technique that will be used to develop a set of on-orbit calibrators for the Infrared Array Camera (IRAC) and demonstrate the validity of the method for stars with spectral types either K0-M0 III or A0-A5 V. For application to SIRTF, the approach is intended to operate with all available optical, near-infrared (NIR), and mid-infrared (MIR) photometry and to yield complete absolute spectra from UV to MIR. One set of stars is picked from Landolt's extensive network of optical (UBVRI) calibrators, the other from the Carter-Meadows set of faint IR standards. Traceability to the ``Cohen-Walker-Witteborn'' framework of absolute photometry and stellar spectra is assured. The method is based on the use of either ``supertemplates'', which represent the intrinsic shapes of the spectra of K0-M0 III stars from far-UV (1150 Å) to MIR (35 µm) wavelengths, or Kurucz synthetic spectra for A0-A5 V stars. Each supertemplate or Kurucz model is reddened according to the individual star's extinction and is normalized using available characterized optical photometry. This paper tests our capability to predict NIR (JHK) magnitudes using supertemplates or models constrained by Hipparcos/Tycho or precision ground-based optical data. We provide absolutely calibrated 0.275-35.00 µm spectra of 33 Landolt and Carter-Meadows optical standard stars to demonstrate the viability of this technique, and to offer a set of IR calibrators 100-1000 times fainter than those we have previously published. As an indication of what we can expect for actual IRAC calibration stars, we have calculated the absolute uncertainties associated with predicting the IRAC magnitudes for the faintest cool giant and hot dwarf in this new set of calibration stars.