[Note: IRCS has a low-resolution prism, which has the wavelength coverage of 0.8 - 5.0 um @ 52 mas (1.8 - 4.1 um @ 20 mas). If you are interested in it although there are several problems which should be solved yet , please contact to Dr. Hiroshi Terada.]
Grisms for use with the 20 mas pixel scaleThere is a single JHK grism for the 20 mas pixel scale, but only a single spectroscopic window can be observed at a time. The short wavelength cut-off is due to the order sorting filter and is listed at the approximately 10% transmission level. The long wavelength cut-off occurs at the end of the array when the transmission is still high.
Grisms for use with the 52 mas pixel scaleThese grisms provide the slit length of a maximum 20''.
*New zJH grism has 1.5 times higher wavelength resolution than previous JH grism.
Comparison lines dataFor wavelength calibration, please use Ar_0.79-4.05.dat which contains Ar lines from 0.79 to 4.05 micron. The Ar lines in short wavelength region (0.79-1.10 micron) are added into argon.dat of IRAF.
Corrections for finite seeing and slit width
Point sourcesThe sensitivity for point sources can be improved by extracting the spectrum along several spatial pixels. Apertures between 1 and 2 times the seeing FWHM are optimum. The following corrections should be added to the numbers in the above tables to determine the point source sensitivity in the relevant seeing conditions. Note that the first two rows represent image quality only attainable with Adaptive Optics. Please ensure you use the correct table for your choice of grism.
Extended sourcesIncreasing the slit width will improve the sensitivity of an extended source, if its surface brightness remains constant. Corrections to be applied to the numbers in the table above are given here. Sensitivity can be improved by binning in the spatial direction; again the improvement is 1.25 log N, for binning over N pixels.
If the source is too extended to be nodded along the 20'' slit, remember to double your integration time to allow for observations of blank sky.
ExamplesBoth examples are for the 52 mas grisms.
1. An observer wishes to obtain S/N=10 per pixel in a properly-sampled spectrum of a K=18 quasar with 500 km/s resolution.
The 52 mas K grism has a baseline sensitivity from the above table of 15.3 mag/arcsec^2. The desired resolution requires the use of the 0.30'' slit, which provides a correction of 2.4 mag from the above table on point source corrections, assuming typical seeing of 0.5''. Since the spectrum will be oversampled, we can bin into groups of 2 pixels for a gain of 1.25 log 2 = 0.4 mag. The 5-sigma, 1-hour sensitivity is therefore K=18.1 for a point source. The required integration time is therefore
[100.4(18-18.1) x (10/5)]2 x 1 hour = 3.3 hours.
2. An observer wishes to map a 6'' x 6'' region of a nebula in the Br gamma and molecular hydrogen lines with 0.6'' resolution to a sensitivity of 5 x 10-19 W/m2 (5 sigma) per spatial element.
With the 0.6'' slit, the spectral resolution will be ~320, so the lines will be unresolved. Each resolution element will cover approximately 22000(A) /320 = 69A, and therefore the emission line sensitivity corresponds to a continuum sensitivity of approximately 7.2 x 10-21 W/m2/A, or K=16.9. This is K=15.3 mag/arcsec2 when the seeing is 0.5" (spatial element = 0.5" x 0.5").
According to the above tables, the extended source sensitivity for the 0.6'' slit is K=16.05 mag/arcsec2, to which a further 0.75 mag can be added to account for 4-pixel binning in the dispersion direction. The exposure time in each slit position is therefore
[100.4(15.3-16.8) x (10/5)]2 x 1 hour = 15 minutes
and 10 separate exposures must be taken to cover the desired region. In addition, sky exposures will be required, so the total exposure time will be 300 minutes.