Calibration of the BATC survey: Methodology and accuracy

We describe in detail the extinction correction procedures used for the Beijing-Arizona-Taiwan-Connecticut Sky Survey (BATC Survey). The survey covers the spectral range 3200-9900 Angstrom by utilizing a set of 15 intermediate-band filters. These filters are specifically designed to exclude most of the bright and variable night-sky emission lines. We also present extinction coefficients for the filter passbands for typical photometric nights at the Xinglong Observing Station, Beijing Astronomical Observatory (where the observations of the survey are being carried out). Time-dependent, low-amplitude (similar to 1%), nightly extinction variation has been observed. Such variation is demonstrably independent of filter bandpass and air mass, with amplitudes ranging from similar to 0.01 to similar to 0.03 mag. The variation is plausibly caused by slowly varying (at similar to 1%) atmospheric extinction, possibly related to changes in air pressure/temperature/humidity that occur during the night. An iterative fitting scheme has been developed to take this time-varying component into account. We conclude that the survey can achieve its stated observational goal, namely, an absolute photometric calibration that is tied to the AB(nu) system to an accuracy of 1% in all filters.