Observing with VIRUS
Below I provide basic properties for VIRUS, and then an overview of
planning, executing, reductions and analysis of VIRUS data.
This page contains the IFU layout, both
planned and current.
PROPERTIES:
VIRUS final performance is under analysis, and here are current numbers for planning:
- 3500-5500 AA, with resolving power of 670 at 3900AA, 850 at 4600AA, and 990 at 5200 AA.
- set of 3 dithers is typical (but not required)
- current line flux limit of 8e-17 ergs/s/cm^2 above 4300AA for point source in an 18-min dither set at S/N=10
- continuum of g=23 AB in 100AA at S/N=50 (similar for u=23) in 18-min dither set. We will update with full curve.
- layout of current and the 50 expected IFUs are given here .
- reductions are automatic every morning and produce flux calibrated spectra. Extraction scripts at given positions are available.
PLANNING:
The main planning tool is called shuffle. The observatory will run
shuffle for any of your shots, and it might be helpful for you to use
it as it helps to understand the layout. Also, you might need to check
the output if have specific locations within an IFU where you would
like to setup on an object, or if you want to use a specific guide
stars. For example, you may want to get more than one IFU with a
specific target, and in that case, I suggest running shuffle and
checking the output images.
First, it is important to distinguish dithering (used to fill fiber
separation) and tiling (used to fill IFU separation). Within a given
IFU, the fibers are separated by 1/3rd of their diameter, allowing for
a three position dither to given a nearly compelte fill. There is
software at the telescope that will dither to high accuracy. The
current accuracy of the dithered positions is below 0.05
arcseconds. Dithering is the default mode, and you should specify that
you want this in the phase II.
There is no automated software for tiling. Within hetdex, we currently
write a routine for individual cases if an observations needs to be
tiled. It is straight-forward to set up tiles and you would need the
PA of the observation and the corresponding RA,DEC positions of the
active IFUs. The PA is easily obtained with a run of shuffle and the
RA,DEC positions of the IFUs using pyhetdex's add_ra_dec.
Shuffle is installed on TACC, and you first neeed to set up the
configuration files (done only once in a directory) and then you can run
shuffle. Thus, the steps are:
shuffle_config
and copy over the expected focal plane layout for the fall trimester. This is kept at:
cp /work/00115/gebhardt/maverick/sci/panacea/shifts/fplanefall18.txt fplane.txt
do_shuffle -v RA DEC rad_off east/west IFU ra_offset dec_offset Name
where:
- shuffle_config and do_shuffle is installed in python in ~gebhardt/anaconda2/bin
- RA,DEC (RA in hours, DEC in degrees) are the positions of where you want the "IFU" field centered
- rad_off is not used anymore, but it designed to "shuffle" the field; set to 0
- east/west should be 0 for east, 1 for west
- IFU is the specific IFU that the RA,DEC will be centered on. Suggest using 000
- ra_offset, dec_offset are applied offets one may want around the RA,DEC
- Name is the target name that will be used for output
There is a descriptive webpage for shuffle at ?. The output you want
is contained in subdirectories (should be obvious). The important
numbers are the RA,DEC of the focal plane center and the PA of the
observation.
EXECUTING:
For the Phase II, you should first work with Matthew Shetrone to help
with any specific Phase II questions. The strong suggestions are:
1) input the RA,DEC of focal plane center (IFU=000) in the Phase
II. Of course, you can input the RA,DEC of a specific IFU and then
state in the phase II that you are NOT using the focal plane
center. This will work, but it easily causes confusion, mainly for
yourself.
2) specify whether you want the standard dither set in the Phase II.
3) specify the positional accuracy of what you need for the initial
setup. This will determine whether the astronomers take a setup image
and tweak the position, or just go ahead and start guiding where the
telescope pointing setup on.
REDUCTIONS:
Reductions to the point of sky-subtracted, flux-calibrated images have been
automated. The scripts work for both standard dithered and non-dithered
sets. The products are multi extension fits files that contain a variety of
reduction steps, including sky-subtracted images.
There are post-processing steps that require accumulation of data before
execution. These are renomalization of individual fibers, full-frame sky
subtraction, flux calibration, and astrometric solutions. Each of these
steps are automated as well. To access the final calibrated spectra, one
runs an additional scripted step. It would be helpful to provide feedback
on how best to integrate both steps.
The reductions can be run by any individual with a tacc account, although
we are working towards a system of having a central directory for all
products of all observations.
Individuals should expect and plan to have access to calibrated products
within days of observations.
There is a complication, not resolved yet, for objects that fill the focal
plane of the telescope. For example, large nearby galaxies (Andromeda, M81
group, etc.) will pose a difficulty for sky subtraction. We have plans to
address how to do this properly (since there are many individuals who want
this), but it has not been implemented yet. It would be great to get
additional work on this.
ANALYSIS:
Analysis is defined as anything post sky-subtraction. There are scripts
that will extract fibers at a given position, and thereby shape, and
provide spectra in ascii files. You can choose to grab all data that has
been taken at a given position or for a specific observation. To execute
these scripts, one types:
~gebhardt/bin/rsp1a2 RA DEC radius wave_center wave_range name1 name2
where:
- RA is in degrees
- radius the size around the RA,DEC one want to extract
- wave_center is where an emission line will be attempted
- wave_range is the extent extracted
- name1 and name2 create a subdirectory called name2_name1 for the output
The script will also always extract the full spectrum, irregardless of the
wave_range. We are working on a webpage that describes these scripts, and
their output, in detail.