Atmospheric dispersion and differential refraction will lead to a non-neglectable light loss varying with wavelength across the field of view during the integral time of multi-fiber spectral observation.
These effects will be more severe in telescopes with large field of view such as the Large sky Area Multi-Object Fiber Spectroscopic Telescope(LAMOST). We have calculated the light loss due to atmospheric refraction for LAMOST. To improve the efficiency, we could move individual fiber to correct its 5000 A monochromatic displacement during observation, the best number of this kind of track should be balanced between the displacement correction and the fiber positioner error. We have calculated the best track number by
monte-carlo simulation for different declination field and different position of the field. Also we have derived the fiber positioner error tolerance from simulation.
Flat-fielding is one of the most important data calibration procedures for wide-field multi-fiber spectroscopy of LAMOST (the Large Sky Area Multi-Object Fiber Spectroscopic Telescope). LAMOST’s unique optical design, wide field of view, significant telescope vignetting and unconventional enclosure structure present a challenge to its flat-fielding method. On the other hand, the spectrographs with its CCD detectors are well fixed on the ground, which implies that there is almost no flexure during spectroscopic observation and thus provides a favorable factor in flat-fielding. Taking into account generally-accepted principles for wide-field multi-fiber spectroscopy and the specific features of LAMOST, a scheme of flat-fielding is designed. It utilizes a combination of the multi-fiber lamp flat field and the offset sky flat field to calibrate all known telescopic and instrumental response ununiformity. The lamp flat field will be used to calibrate the CCD pixel-to-pixel sensitivity variation, fiber transmission as a function of wavelength, and possible spectrographic vignetting. Fiber-to-fiber throughput difference and telescopic vignetting will be corrected with the offset sky flat field. A few choices are proposed for the flat field equipment. The flat-division is described in the context of the data reduction pipeline.