ESTIMATED SENSITIVITY AND COUNTS

Introduction

Listed below are estimated sensitivities for the SXD and LXD modes. The estimated sensitivities are about 0.5 mag brighter than those calculated by the flux calculator. The most likely reasons for the discrepency are a combination of the guider not keeping the source perfectly in the narrow slits and noise introduced by sky variability and division by flats and standards, since throughput, background and instrument noise are well known.

Limiting flux is proportional to SNR/(itime)**0.5

Limiting magnitude is proportional to 2.5log10[SNR/(itime)**0.5]

SXD Mode

In the short wavelength cross-dispersed mode (SXD) the maximum on-chip integration time used is 120sec to ensure good OH emission line subtraction. A one hour integration requires 15 cycles of object-sky (AB) pairs assuming the object is nodded along the slit and appears in both the A and B nod positions.

Plot of counts in DN/sec from 85 Gem (A0V, B=5.34, V=5.38) taken in SXD mode with 0.5 arcsec slit. The seeing was 0.8 arcsec FWHM at K. Note the blaze function of each order, A0V stellar absorption features and telluric absorption features. Since A0V defines stellar magnitude (ie. V-J=V-K=V-L'=0.0 etc.), counts may easily scaled to other magnitudes (allowing for changes due to slit width and seeing). Wavelength is plotted at the bottom of each order and pixel number at the top. The Flux (DN/sec) is the total counts summed along a spectral column (one pixel wide). Peak counts in a particular pixel is of course less. Use the box function DV STATS to sum along a column to check the total flux in your data.

LXD Mode

In the long wavelength cross-dispersed mode (LXD) the maximum on-chip integration time used is 30sec to ensure good thermal sky subtraction. A one hour integration requires 60 cycles of object-sky (AB) pairs assuming the object is nodded along the slit and appears in both the A and B nod positions.

LXD Point Source Sensitivity

			10sigma 1hour (mag) (DN/sec)
R	Slit	Seeing	K	L	L'
1500	0.5"	0.5"	14.2	11.9	11.4
1500	0.5"	0.8"	13.9 (0.2)	11.6 (0.3)	11.1 (0.7)

Plot of counts in DN/sec from 85 Gem (A0V, B=5.34, V=5.38) taken in LXD2.3 mode with 0.5 arcsec slit. The seeing was 0.8 arcsec FWHM at K. Note the blaze function of each order, A0V stellar absorption features and telluric absorption features. Since A0V defines stellar magnitude (ie. V-J=V-K=V-L'=0.0 etc.), counts may easily scaled to other magnitudes (allowing for changes due to slit width and seeing). Wavelength is plotted at the bottom of each order and pixel number at the top. The Flux (DN/sec) is the total counts summed along a spectral column (one pixel wide). Peak counts in a particular pixel is of course less. Use the box function DV STATS to sum along a column to check the total flux in your data.

Prism Mode

In the low resolution prism mode the maximum on-chip integration time used is 120sec to ensure good OH emission line subtraction. A one hour integration requires 15 cycles of object-sky (AB) pairs assuming the object is nodded along the slit and appears in both the A and B nod positions. Note that in this mode the resolving power R changes with wavelength since it is a function of prism refractive index. The average value is given in the table below. For the 0.3 arcsec slit R~350 at K, R~250 at H and R~150 at J. These resolving powers scale inversely with slit width.

Prism Point Source Sensitivity

			10sigma 1hour (mag) (DN/sec)
R	Slit	Seeing	J	H	K
~250	0.3"	0.8"	16.9	16.4	16.0
~150	0.5"	0.5"	18.3	17.8	17.3
~150	0.5"	0.8"	17.9	17.5	17.0
~150	0.5"	1.1"	17.7	17.3	16.8
~100	0.8"	0.8"	18.9	18.4	17.9