The primary operational wavelength regime for DEIMOS is redward of 5000 Å, and most of our DEIMOS observing procedures are written under the assumption that red calibrations are appropriate. DEIMOS can also be used in the blue, but calibrations must be acquired in a different manner in order to obtain optimal wavelength solutions and flatfielding with the DEEP2 DEIMOS data reduction pipeline. This document describes the procedure developed by Evan Kirby of Caltech to achieve satisfactory blue calibrations.
A new DEIMOS script called calib_blue has been developed for obtaining arc line and internal flat lamp calibrations which are appropriate for use of DEIMOS in the blue. This procedure is optimized for use with the 900ZD grating at a central wavelength of 5500 Å. It takes separate arc and flat exposures for the red chips and the blue chips. The red-chip arc and flat images are not appropriate for the blue side and vice versa.
For arc lamp exposures, the red chips use the standard combination of lamps used in the main calibration script. The arc exposure for the blue chips use a combination of lamps optimized to get the most lines in the blue. Some lamps are turned off during the exposure so that they do not saturate. The active lamps will not be recorded properly in the FITS header. The user must modify the FITS header to reflect all five lamps that were on at some point in the exposure: Hg Cd Zn Ar Kr. If the spec2d pipeline is used for data reduction, separate arc exposures may be specified in the plan file using the BLUEARCNAME and REDARCNAME keywords.
For flat field exposures, the script acquires three images for the red side. Because the DEIMOS throughput is very low near 4000 Å, the script acquires six images for the blue side. These exposures will saturate the red side. As written, the spec2d pipeline will not accept separate red and blue flat field images. However, it is reasonably easy to modify the read_planfile.pro and deimos_mask_calibrate.pro files to accept different images. Interested users are welcome to contact Evan Kirby for examples.
The script requires about 23 minutes per mask to acquire the calibration exposures. If one is short on time, then the user may wish to modify the script to reeduce the time to warm up the arc lamps.