UBVRI photometry of star forming regions

This dataset contains UBVRI catalogs for the following star forming regions: alpha Per, IRAS 20050+2720, Lk Ha 101, NGC 281, Orion (not the ONC, but the less dense "hokey stick" region). The catalogs are merged form observations that were carried out over 4 nights from 2009-11-18 to 2009-11-21 at the Fred Lawrence Whipple Observatory. Optical photometry was obtained at the Fred Lawrence Whipple Observatory (FLWO) with the 1.2 m telescope and the Keplercam instrument. Keplercam is a camera with a Fairchild ``CCD~486'' detector. Each pixel corresponds to 0.336 arcsec on the sky and saturates at about 65,00 counts per pixel with a total field-of-view (FOV) 23.1*23.1 arcmin. We operated the chip in the default 2*2 binning. It is read-out in four amplifiers, each of them has a different gain and read-out noise. Observations were taken with staggered exposure times of 6, 60 and 300 s, so that photometry for brighter stars can be performed on frames with shorter exposure times. Keplercam is equipped with filters for different photometric systems, here we use observations taken in u', B_H, V_H, r', i'. The B_H and V_H filters belong to the Harris set, the u', r' and i' filters are part of a Sloan Digital Sky Survey (SDSS) set. We performed the data reduction in Pyraf, a Python interface to IRAF. Photometry is done with PSF-fitting in DAOPHOT as implemented in IRAF. Data and flatfield are bias-corrected, then the data images are divided by the flatfield. On those images we performed a source detection and an initial aperture photometry as starting point for the PSF-fitting. The PSF is fitted independently for each exposure and each amplifier, but is assumed to not to vary spatially. We approximate the PSF with a moffat25 function and calculate a look-up table of deviations from this function from the brightest, non-saturated sources in the field. We place a very conservative limit of 40,000 counts per pixel to ensure linearity of the detector. The full width at half maximum (FWHM) of the PSF is about 4.5 binned pixel, i.e. about 3 arcsec, but the PSF changes from exposure to exposure and in some cases it is significantly non-symmetric. This leads to notable artifacts in PSF-subtracted images and limits or ability perform a second source detection. Also, the tail of the PSF is very wide an merges into the background after 90 binned pixels or more for bright sources. Thus, our source lists are incomplete close to bright sources and therefore the completeness limit is non-uniform over the FOV. The absolute pointing accuracy of the FLWO telescope is insufficient to match source positions with the other datasets directly. Therefore, we first correct the WCS of each individual exposure with the IRAF task msccmatch, which minimizes the residuals between sources in our images and the 2MASS catalog. Sources are extracted independently for different exposures and source lists are merged with a matching tolerance of 3 arcsec, which is of the order of the remaining astrometric errors. Properties of bright sources, which are identified in exposures with different exposure times, are taken from the shorter exposure, if their instrumental magnitude is above a threshold value. To perform absolute photometry extinction and zero point of the instrumental magnitude are calibrated with observations of Landolt field 01, which was reduced analogues to the data itself. We fit the instrumental magnitudes in u', B_H, V_H, r', i' to the catalog values for UBVRI assuming an offset, a linear color term and a linear extinction term. Ten sources in the standard field are extremely red (1.15 > R-I > 2.0), but we do not find any break in the transformation law. We remove the faintest stars from our final sourcelist, so that the typical error on the flux is about 10 %