Spectroscopic Observing
This section describes basic spectroscopic observing setup and
techniques. The
subjects covered are:
We recommend familiarizing yourself with the Spectroscopy
Overview before reading through this section.
From imaging mode, OSIRIS is put into spectroscopic mode with the
SPMODE command. SPMODE performs the
following operations, with typical times for completion given in ()'s:
- Retracts the prefilter wheel and pupil mask mechanism (~2
minutes).
- Flips the diffraction grating into the beam, replacing the
imaging flat mirror (~5 seconds)
Once started, this operation cannot be interrupted, and it will take
the
same amount of time to switch back to imaging mode (via the IMMODE
command).
After selecting SPMODE, you will need to select the camera,
camfocus, grating tilt, slit, and filter required for the desired
spectroscopic mode as described in Table 1
of the Spectroscopy Overview.
To make this more convenient, SPMODE can take additional command-line
arguments to select the filter, camera, etc. in one line. For example,
typing:
SPMODE FILTER=3 CAMERA=0 CAMFOCUS=245 SLIT=0
will put OSIRIS into SPMODE, put filter 3 into the beam, select camera
0
and slit 0, and set the camera focus to 245. Which settings you
actually
use, however, will depend on the configuration you want (warning: these
values are only placeholders not actual examples).
[ Top of Page | Contents
]
You also need to select the grating tilt for the spectroscopic region
of
interest. Once selected, the grating tilt will stay fixed until you
change
it again. To keep calibration tractable, it is a good idea to keep the
number of different settings to a minimum, as dicussed in the Strategies section below.
Low-Res and X-Disp Modes
In Low-Res and cross-dispersed modes the grating tilts are fixed as
given
in the Wavelength Selection section of
the Spectroscopy Overview. The commands
provided for selecting these preset grating tilts are as follows:
- F3SPEC band
- Selects the optimal tilt for low-res long-slit spectra in the
given band (one of J, H, or K).
- XDSPEC
- Selects the optimal tilt for the cross-dispersed mode. The
order-sorting filter for cross-dispersed mode is integrated into the
X-disp slit.
In general, you should never need to make any adjustments to the preset
tilts for either of the two low-res spectroscopic modes (see
Low-Res Long-Slit Mode for details).
High-Res Mode
For High-Res spectra, you need to select the tilt by wavelength for the
order of interest using GRATING
commands as follows:
GRATING tilt
Tilts the grating to encoder value tilt. This requires that
you first determine the grating tilt corresponding to a particular
wavelength.
[ Top of Page | Contents
]
Because it takes 2 minutes to switch between full imaging and
spectroscopic
modes, a simple quick-look imaging configuration, called "Acquisition
Mode", is provided via the ACQMODE command.
This will replace the grating with the imaging flat (leaving the
grating
tilt unchanged). It takes approximately 5 seconds to switch
from
SPMODE to ACQMODE and back, compared to 2 minutes to switch
from
SPMODE to IMMODE and vice versa.
After selecting ACQMODE, you will also have to select a filter and
imaging mask appropriate to your spectroscopic setup. Like SPMODE,
the ACQMODE also accepts optional command-line arguments
to set the SLIT and FILTER to be used for acquisition imaging. The
typical
selections are described in Table 1 below:
Thus, if you are in Hi-Res long-slit mode, you need to select the
f/7 Imaging mask for ACQMODE target acquisition, etc.
If your target is bright at visible wavelengths, the J filter is
usually
the best choice for target acquisition since J has the lowest
background. If your target is very reddened, H or K will be best, but
since
ACQMODE works without a pupil mask, the background will be
higher
than in full imaging mode (this is why ACQMODE should not be
used for normal photometric imaging - it is only intended for
quick-look
spectroscopic target acquisition).
Table 2 below gives the approximate pixel coordinates (X,Y) of the
slit centers. The precise location of the slit center should be
always be verified. For a discussion
of object centering in the slit, see the
IRS manual.
Note that these pixel centers assume that you have the appropriate
camera in place for each slit mode (see Table
1 of the Spectroscopy Overview Section).
The typical target acquisition strategy is as follows:
- Switch to ACQMODE, selecting the appropriate imaging mask and
filter.
- Find your object, and move it to the nominal slit center.
- Setup the autoguider on nearby field stars (the guide probe is
warm and must not be in the OSIRIS field of view as it will greatly
increase the thermal background).
- Verify centering of the target after the guider "locks."
- Switch back to SPMODE and select the appropriate slit,
filter/grism, etc. for your spectroscopic observations.
For particularly bright targets, an optional step is to image your
target
through the slit while still in ACQMODE to verify that it is centered.
This is done by selecting the slit, and then "peaking up" the object by
taking a series of images with slight adjustment of the pointing (using
the
NORTH, SOUTH, EAST, and WEST commands as appropriate) so that the
maximum
broad-band light gets through. Once you are peaked up, lock the guider
onto the guidestar and switch back to SPMODE. In general, you will find
that if you can center the object at the location nominal slit center
in
ACQMODE, you will have the object well-centered in the slit.
Note that if your object is very bright, the first few spectra may
have
residual image artifacts due to the previous imaging observations. The
solution is take a few quick exposures before starting your science
frames to clean off the residual images. The section on the
HgCdTe Detector Array provides more information on the array
properties.
[ Top of Page | Contents
]
Yes, you should sky chop.
Point Sources
The most reliable way to sky chop when observing point sources is to
chop along the slit, so that you are always taking object and sky
spectra
simultaneously, giving 100% on-source dwell time.
In general, this strategy is most useful if the point source is
very bright
and easy to see on the array in a single integration. If you need to
build up S/N to see a faint point source, you can still use this "slit
chopping" trick, but you need to be very careful to make sure your
object
does not wander out of the slit.
Extended Sources
For extended sources that fill the entire length of the slit, you need
to sky chop away from the object to take separate sky frames, then go
back to the object, much in the same way as is done for imaging
extended
sources. The main difference is that in order to build up sufficient
S/N, you need to be certain of returning the object to the same
location
on the slit each time. You do this as follows:
- At the end of an object spectrum integration, disable the
autoguider using the -GUIDER command.
- Chop off to some distant sky position using the OFFSET
command.
- Take sky spectra of comparable integration time.
- Return to the object by using OFFSET with the reverse
vector.
- After the telescope has settled down, move the guide star back
into position, and resume guiding using the +GUIDER
command.
- Once the guider has locked, start the next object spectrum
integration.
[ Top of Page | Contents
]
Wavelength calibration can be done in one of two ways:
- Sky Spectra using the OH airglow lines.
- Lamp Spectra using the lamps in the 4-meter
guide/acquire module (He+Ar and Ne gas-discharge lamps).
All sky spectra will contain bright OH airglow lines that may be used
for
spectroscopic calibration. Adding together all sky spectra from a given
object will usually provide a good S/N reference spectrum. The only
problems are some blending of OH lines together, and the general
scarcity
of OH lines in the red half of the K-band. Both IRAF and XVista provide
lists of identified OH airglow lines.
A list
of the OH airglow lines in
the near-infrared can be found in Oliva & Origlia [1992, A&A,
193, 327].
Example OH airglow spectra are available (dw is the linear dispersion
in microns/pixel and linear fits have residuals of ~ 0.1 pixel):
- XD K-band,
SOAR telescope, R=1200 (IRAF
dispcor solution: xdskyk: ap = 3, w1 = 1.793121, w2 = 2.703399, dw =
8.898E-4, nw = 1024)
- XD H-band,
SOAR telescope, R=1200 (IRAF
dispcor solution: xdskyh: ap = 2, w1 = 1.344754, w2 = 2.030659, dw =
6.705E-4, nw = 1024)
- XD J-band,
SOAR
telescope, R=1200 (IRAF
dispcor solution: xdskyj: ap = 1, w1 = 1.075595, w2 = 1.626385, dw =
5.384E-4, nw = 1024)
- f/7 K-band,
SOAR telescope, R=3000
Example He+Ar and Ne lamp spectra are available:
The near-infrared region has many strong absorption features due to
various molecules in the atmosphere. One of the best ways to remove
these
features is to ratio the spectrum of the target object with the
spectrum of a
featureless source observed with the same instrumental setup and
airmass. In the J and K band, A stars provide good
atmospheric
standards, since they have only H absorption features at 2.17 microns
(Br-gamma)
and 1.28 microns (Pa-beta). In the H
band, A stars have many Br-series
absorption features that make them somewhat problematic. Note that
Kurucz models of A stars reproduce the near-infrared spectrum of
these stars reasonably well, so these may be used to correct for the
intrinsic absorption in the stellar atmosphere.
Another technique for
correcting for atmospheric absorption in the H band is to obtain
standards of both G and A stars as atmospheric standards. G stars have
relatively
few spectra features in this regime and can be used to correct for the
Brackett absorption lines in the A stars. Ratio the A star by the G
star, then fit the resulting Br spectrum. This fit can be
normalized and then divided into the A star spectrum. The G and A star
should be a\observed at the same airmass.
Spectroscopic flats should be obtained using the dome flat field lamps.
A series of exposures with the lamps off should be subtracted from a
similar
series with the lamps on. In the case of K band spectra, it may be more
advantageous to subtract only a dark frame as the total illumination
may have a
significant thermal component.
[ Top of Page | Contents
]
Multiple Spectroscopic Modes
Since each grating tilt setting requires a separate set of flat-field,
wavelength, and flux calibrations. The best strategy is to keep the
number
of different settings for a given program to an absolute minimum and to
take
calibration frames as often as possible.
Spectroscopic mode setup can get involved, if you are going to be
switching often between spectroscopic and imaging modes, it is a good
idea to learn how to define command aliases or scripts in Prospero
and use these for setting different modes. These can greatly simplify
your
setup issues, and eliminate loss of time due to careless mistakes late
at night. See the Prospero Observer's Guide for OSIRIS
for a
description of how to use aliases and scripts for simplifying
instrumentconfiguration.
[ Top of Page | Contents
]
Updated: 2010 August 26
