Spectroscopy: Arcs

An Argon lamp is used for wavelength calibration with UIST. Plots of Argon arc lines obtained with the UIST grisms are available from the table below. A 2-pixel wide slit was used with all data, except for the IJ, JH and HK grisms, where a 4-pix slit was used. Lamp lines are labeled in each plot (with VACUUM wavelengths). Arc line designations are also available from the CGS4 arc line web pages.

Short-H and short-K Grisms

Arc spectra taken with the short-K and short-H grisms exhibit unusual internal reflections, which are manifest as large circles with long-slit spectroscopy, and faint, shifted "ghost" lines in IFU data (see examples below). These features are not seen with the other grisms and are independent of the blocking filter used. These reflections seem to be associated with the calibration unit itself, since similar features are not seen from sky lines with long exposures. They should therefore not affect your data.

Short-H long-slit arc	Short-K long-slit arc
Short-H IFU arc	Short-K IFU arc

Wavelength calibration in the thermal IR

NOTE: short-L, long-L and M-band spectra should be wavelength calibrated with sky lines; using arc spectra will only give an approximate scale.

The optics associated with the calibration unit are quite bright in the L and M bands. Consequently, arc lines are routinely observed through the J- and H-band spectral blocking filters in second (short_L) or third (long_L and M) orders. However, these filters also introduce a wavelength shift, so that simply calibrating a target spectrum (observed through the L or M-band blocker in first order) with these second or third order J-/H-band arc lines gives an incorrect wavelength scale.

Estimates of the wavelength shift - between spectra properly calibrated with sky lines and spectra calibrated with argon lines observed in a different order - are given below.

Note that the orac-DR pipeline will calibrate IFU spectra using Argon arc spectra (long-slit spectra will be assigned an estimated wavelength scale). To shift spectra to the correct (to first order) wavelength the offset noted in column 2 should be applied. However, best results will always be obtained when sky lines are used for wavelength calibration.