|
Australia Telescope: Mopra Telescope |
 |
|
|
||
|
Note added 26 May 2009
Kate Brooks will be on extended leave from July 1 2009 to February 1 2010. During this time James Urquhart will be the Mopra Astronomy Research Scientist.
The Mopra 22-m antenna is part of the Australia Telescope National Facility (ATNF), operated by the CSIRO. It is used together with other AT antennas (the six 22-m dishes at Narrabri, and the 64-m Parkes dish) as part of the Australian Long Baseline Array. It is also used on its own, for single-dish observations, particularly at millimetre wavelengths.
The telescope is at latitude -31° 16' 04''.451, longitude +149° 05' 58''.732 (east of Greenwich), and its height above sea level is 850 m.
As part of a major upgrade program, a new millimetre receiver system covering 3 bands between 16 and 116 GHz and a new backend system have been installed at Mopra. The backend produces two 8-GHz bandwidth IFs from each polarisation, andis capable of observing several spectral lines simultaneously.
Information on applying for telescope time is available online in the ATNF Guide. Be sure to first check the current status. Observing at 3-mm is best in the APR term only (winter months). For the OCT term (summer months) ATNF encourages medium-large programs (e.g. large-sky maps, surveys, large samples) that require a time allocation of at least two weeks with 12-hour daily observing blocks.
Observing is carried out remotely from the control room of the Australia Telescope Compact Array located at the Paul Wild Observatory near Narrabri. Observers must carry out their own observations. Full remote observing from anywhere in the world is available.
If you have a query about the Mopra telescope then send an email to mopra-sciops @ atnf.csiro.au .
The receivers use Indium Phosphide (InP) High Electron Mobility Transistor (HEMT) Monolithic Microwave Integrated Circuits (MMICs) as the amplifying elements. The receiver systems require no tuning and include noise diodes for system temperature determination. (Note that the 3-mm system also requires an paddle/vane measurement for the system temperature determination).
The following table summarises the parameters for the three receiver systems.
MOPS is a digital filterband. The 8-GHz broadband signal channels from the mm receivers are transmitted to MOPS, installed in the Mopra equipment room, via broad band analogue fibre optic links. Each 8-GHz IF is then split into four 2.2-GHz bandwidth channels and down converted to the base band. Every 2.2-GHz base band IF signal is digitised by a 2-bit 4.4 Gsample/sec Indium Phosphide digitiser. The digitisers were developed at ATNF and fabricated at TRW in March 2000, under the CSIRO Special Executive program.
MOPS offers two configurations:
Wideband Mode
In Wideband Mode continuous coverage over 8.3 GHz is achieved using 4 overlapping sub-bands.
Each sub-band is 2.2 GHz wide with 2 x 8096 channels (1 x 8096 chans for each polarisation, xx and
yy). Each dataset contains a total of 64 768 channels (4 x 8096 x 2). The label for this mode is
mops_2208_8192_4f. You may choose a binning mode identified as mops_2208_1024_4f (not popular but will help save disk space since it reduces the sizes of your data files).
Zoom Mode
In Zoom Mode it is possible to observe up to 16 zoom bands (often referred to as "windows")
simultaneously. Each zoom band is 137.5 MHz wide and has 2 x 4096 channels (1 x 4096 chans for
each polarisation, xx and yy. Up to four zoom bands may be selected in each of the 4 overlapping
sub-bands. The label for this mode is mops_il_zoom.
The following diagram depicts a typical arrangement of bands within
the 8 GHz frequency range.
At any frequency within these ranges the full 8-GHz frequency coverage is possible. The gap between 38 to 44 GHz cannot be processed as one frequency setting.
For example, when tuned to 38 GHz your frequency band will cover the range 34 to 42 GHz. When tuned to 44 GHz the frequency band will cover the range 40 to 48 GHz. However, one cannot have a band containing 38 to 44 GHz.
Additional Technical Information
1. Corrected antenna temperature (Ta*)
This is the power detected
by a telescope expressed in terms of an equivalent blackbody
temperature. Ta* should be corrected for systematic effects of the
Mopra system (receiver noise, back scattered radiation, side lobe
contributions, ohmic losses etc, all of which are incorporated into a
Tsys measurement ) as well as the affects as atmospheric attenuation,
Tau. This temperature scale is essentially the Mopra temperature
scale and is what comes out of the MOPS box.
2. Main Beam Brightness Temperature (Tmb)
This is the calibrated
antenna temperature after correction for the main-beam efficiency of
the telescope. Tmb will be the true source brightness temperature of a
source that just fills the main beam. This temperature scale is
essentially a telescope-independent temperature scale. It allows one
to compare data taken with Mopra with data taken with, for example,
APEX.
To convert Ta* to Tmb you need to divide by the antenna efficiency. This is just a constant factor which we estimate to be 0.5 at 3-mm and 0.6 at 7-mm and 12-mm (we still need to better tune these values but they are good enough for a first-order correction). For details on the antenna efficiency in the 3-mm band refer to the article by Ladd et al. 2005.
An alternate method for measuring tau is to use a skydip procedure whereby the gated total power (GTP) is measured at several different elevation settings. The change in GTP with elevation yields a measurement for tau. The skydip method is time consuming (~ 10 mins to complete) and is only suitable for very stable weather patterns (cloud-free skies). Skydip measurements are not straightforward with the Mopra system and are currently not supported.
Currently for Mopra there is no atmospheric correction made in the 7- and 12-mm bands, either online (with paddle or skydips) or offline (using an atmospheric model). For 12-mm and 7-mm Mopra observations the effect of the tau correction to the line signal strength is expected to be less than 25% and combined with the other flux uncertainties (pointing, antenna efficiency etc) this is not significant.
Note: For ATCA the tau correction at 12- and 7-mm is made offline during the data reduction process. There is a Miriad routine that uses a simple atmospheric model to determine tau and then scale accordingly the 7-mm and 12-mm visibilities. No online paddle measurement is made. There is a command in Miriad that lets you play with the atmospheric model to gauge what the tau correction is for certain temperature/pressure/humidity readings. The task is called "opplt". For Mopra, the plan is to eventually incorporate the miriad opacity model into ASAP so that corrections can be done for the 7- and 12-mm bands.
For additional information take a look at the presentation from the Mopra Training Day held in March 2008 -Temperature Scale and Flux Calibration.
Of course in bad weather (e.g thick clouds) even spectra with short integration times will suffer from baseline ripples because of the change in sky continuum levels as the clouds pass overhead. In this situation keep the time difference between the on-source and off-source observations as short as possible (e.g. 30 sec) or consider stopping all observations until the weather improves.
Online Approach - Median Reference Spectrum
One technique that seems to have some success involves a non-standard
way of doing position-switching whereby two spectra at different
reference positions (also called "sky" or "off" positions) are taken
and then averaged together to form a median reference spectrum. This
median reference spectrum is then used in the quotient calculation. The
reasoning behind this method is that the ripple is exasperated by the
difference in continuum levels between the target position and the
off-source (reference) position. By using two different reference positions and
then combining them together to form a median reference spectrum there is a
better chance of matching the continuum levels with that of the
target. Unfortunately this method involves quite a bit of fiddling
with the schedule files and the data reduction.
Currently the schedule preparation tool on the web only permits the following sequence: REFERENCE/TARGET/REFERENCE/TARGET or REFERENCE/TARGET/TARGET/REFERENCE. So the way to proceed is to decide on your two reference positions (e.g. offset on either side of your target), then make your schedule files with the web preparation tool but only enter one of the reference positions. Save the output (a text file) and then edit directly the text file to include your second reference position (e.g. REFERENCE#1/TARGET/REFERENCE#2/TARGET)
Data reduction for position-switching mode is done with the software package ASAP. To perform a non-standard quotient calculation on the data requires some python knowledge . A tutorial is available online that may help you (Tutorial 6). The tutorial uses Parkes multibeam data to extract a median reference spectrum.
Note that you will still be able to reduce your data with ASAP in the
standard way (with no median reference spectrum). It is always worth
using this standard method first for a quick look at the data and to
experiment with the lagflag procedure (see next paragraph). You may
find that the lagflag procedure is sufficient for your data. In which
case you do not need to create a median reference spectrum and
calculate the quotient manually. Unfortunately you won't know this
until you have already taken and looked at your data, which is why it
is a good idea to take data using two different reference positions,
just in case.
Offline Approach - Flagging the Lag Spectrum
Another way to reduce the affect of the baseline ripple is to fourier
transform the spectrum and flag the data in frequency space. It is possible to do this in ASAP using the "lag_flag" command. A tutorial is available online
that may help you
(Tutorial 5). The best result is found using the following input parameters:
c.lag_flag(12,11.9999,unit="MHz")
To determine the settings for MOPS zoom modes you will need to use the program mops_vis. Currently this is only running on the local machine pyroclast. To run the program type "mops_vis" at the command line. This will open a GUI in which you can enter the transition lines you wish to observe and the correction for the target velocity frame. A selection of zoom windows will then be displayed from which you must note down the MOPS settings (numbers from -59 to +59). These numbers should then be entered into the on-line schedulers, in the windows for the "MOPS Zoom Bands". If you want to prepare your schedule files before arriving at the Observatory (strongly recommended) then leave these windows empty and enter the values directly into the text file at a later time, once you have access to mops_vis.
For more information on using mops_vis take a quick look at this presentation.
To install mops_vis on your own computer you could try your luck with these two files: mops_vis.gz and mops_vis_broad.gz. They require Perl/Tk installed to run.
The schedule is intended to mimic the operations on the GUI - you enter some new fresh parameters (eg, a new source position), and launch operations (e.g.,a scan). Any GUI command can be included in the schedule. Once a schedule has been invoked ("start") the schedule maintains control of TCS, shutting outmost of the user-selectable functions of the GUI.
A schedule contains a loose mix of parameters and actions; these are grouped in UNITS. The recommended procedure is to make each unit self-sufficient; each unit may contain a number of actions, and each action will be governed by all the parameters between the start of the unit and the action. TCS reports the number of actions and units in a schedule; the schedule pointer increments on data collecting actions, not units. It is therefore in practice simpler to write schedules containing precisely one action per unit (though unit-based incrementing may be introduced in the future).
When the user requests TCS to start an observation by clicking on the "start TRACK"/"start SCAN" button on the main action panel (without using a schedule), TCS reads all the parameters from the GUI and uses these to define the observation. When a schedule is used, by clicking on the "Start sched" button, TCS first reads all the GUI parameters, then steps through the schedule from the nominated starting point. Any parameter defined in the schedule will therefore override the GUI value. Further, the value displayed on the GUI will be immediately reset to the value defined in the schedule, so the new value persists indefinitely (until explicitly changed). Thus the GUI should represent accurately the parameters of any observation currently in progress. (If this fails to occur it is a reportable bug!). A further important consequence of this model is that parameters not specified by the schedule will default to the value displayed on the GUI. This is important to bear in mind, particularly at the start of an observing session. Unintended settings on the GUI inherited from a previous observing session can be disastrous if not detected by careful checking!
An observer can use TCS without ever using a schedule. For example, a series of repeated or similar simple observations of a list of sources might be accomplished more easily from the GUI alone, loading the source details from a user-prepared catalogue. This is the most interactive but also the most flexible style of observing.
The next level is to write a "minimalist" schedule containing only those commands which change from one observation to the next (e.g. source details) and use the GUI for fixed parameters or those which change only occasionally (e.g. correlator configuration etc).
Alternatively, an observer can "play it safe" and include in the schedule file every relevant setting in every "unit". This gives protection against inheriting unwanted settings. It also has the advantage that the schedule files can be started midway through (simply specify the start "unit" number in the TCS).
A hybrid scheme is to define all or most of the relevant parameters in the first "unit" of the schedule, and thereafter specify only changes to parameter settings (rather than repeating them for every integration). This can be a very convenient mode of observing, but beware of re-starting a schedule midway through after an interruption - any GUI parameters that have been changed in the meantime may not get reset if you skip the first integration of the schedule.
The easiest way to create a schedule file is to use the on-line scheduler tools. There are two different schedulers - one for Position Switching and one for On-the-fly mapping. Links to the latest versions of the schedulers are on the main Mopra webpage. Both schedulers use the hybrid scheme (defined above). These schedulers create a text file that should be first saved to your local machine (filename.sch).
It is possible for users to create their own schedule file (without using the on-line scheduler tools). Refer to the TCS User Manual for more information on individual commands.
Once your schedule file is complete it must be uploaded to a local machine called bigrock, which runs the Mopra Telescope Control Software (TCS). On bigrock there is the directory /nfs/online/local/tcs/sched/2009. In this directory you must first create a directory using your Mopra project number (e.g. M007) and then upload all your schedule files to this new directory. Please note that there is no read/write protection or guaranteed backup of observer schedule files on bigrock.
If you have a large target sample you may prefer to use a catalogue file, which lists all the source names and coordinates. The catalogue file can then be used in conjunction with a generic schedule file, which does not include entries for the source name and coordinates. Here is an example of the format of a catalogue file. Here is an example of generic schedule file that does a simple position-switch observation with a 3-mm paddle calibration.
Currently there is no generic schedule file that does on-the-fly mapping. For this observing mode a centre position must be specified in the schedule file. (A catalogue file cannot be used.)
For position-switching data use ASAP. A tutorial for ASAP beginners is now available (May 2008). More tutorials will be added.
An ASAP-GUI interface has been developed by Cormac Purcell and is available for download. This GUI is useful for quick-look data reduction of position-switching data. It also has the capability to export the reduced data. To start the GUI type "python -i QuickData.py". Load in your data files with the "Browse" and "Load" buttons. Select the IF you wish to work on with "Current IF". First average in time the data for each of the two polarisations (2 blue boxes) and then average the two polarisations (2 green boxes).
For on-the-fly mapping use Gridzilla/Livedata. New users should use this worksheet.
Mopra observers must stay on site at the Paul Wild Observatory in the Narrabri Lodge. Accommodation can be booked through the on-line form. Information about getting to Narrabri and the Narrabri Lodge are available here. For new observers we recommend you arrive at the Observatory at least one day before your observations start so that you have time to finalise your schedule files and become familiar with the online observing system. There are computing facilities at the observatory for you to use for data reduction.
Observers are welcome to visit the ATNF Headquarters in Sydney before and after an observing run. On-site accommodation is available and must be booked through the on-line form. See the Visitor's Guide for further information about getting to ATNF Headquarters. Colloquia are usually held on Wednesday afternoons starting at 3:30 pm. If you would like to deliver a colloquium then contact the colloquium organiser Patrick.Weltevrede @ csiro.au. An informal Director's morning tea is held each Wednesday at 10:00 am in the office of the AAO or ATNF Director.
Please note: You may only remote observe with Mopra if you have prior approval by Phil Edwards or Balt Indermuehle.
Information on how to request and setup remote observing is available via the ATNF Remote Observing Guide
Mopra user support is part of ATNF Science Operations, headed by
Philip Edwards. Kate Brooks and Balthasar Indermuehle are the primay
Mopra support people. Send all your help requests to mopra-sciops @
atnf.csiro.au preferably at least one week before your observing run.
Kate Brooks (on leave from July 1 2009 to February 1 2010
Room 77A
Radiophysics Laboratory
Cnr Vimiera and Pembroke Roads
Marsfield NSW
Kate.Brooks @ csiro.au
Tel: +61 2 9372 4649
Fax: +61 2 9372 4310
Please note I work Mondays, Tuesdays and Wednesdays only
Balthasar Indermuehle
Locked Bag 194
Narrabri NSW 2390
balt.indermuehle @ csiro.au
Tel: +61 2 6790 4093
Fax: +61 2 6790 4090
James Urquhart (Mopra Astronomy Research Scientist from July 1 2009 to February 1 2010)
Room 84
Radiophysics Laboratory
Cnr Vimiera and Pembroke Roads
Marsfield NSW
Kate.Brooks @ csiro.au
Tel: +61 2 9372 4360
Fax: +61 2 9372 4310
Mopra Control Desk
Tel: +61 2 6790 4048 or 4088
ATCA Control Room
Tel: +61 2 6790 4033
"The Mopra radio telescope is part of the Australia Telescope which is funded by the Commonwealth of Australia for operations as a National Facility managed by CSIRO."
Where possible, authors are requested to include the term "Mopra" in the ABSTRACT of their papers. This is to facilitate electronic searches for publications that include data from ATNF observatories. Please inform Christine van der Leeuw (Christine.VanDerLeeuw @csiro.au) of any publications which include ATNF observations.
The ATNF welcomes any articles featuring Mopra data as contributions for the ATNF Newsletter.
The Mopra antenna is situated about 20 km west of the town of Coonabarabran, which is nearly 500 km north-west of Sydney. It is on the eastern edge of the Warrumbungle National Park, and about 4 km due east of the Siding Spring Observatory. It was called the Mopra antenna after Mopra Rock (elevation > 1000 m) which is about 2 km south-west of the telescope and can be seen as you approach from Coonabarabran.
The Mopra site consists of the antenna and a control building, which also contains the on-site accommodation. More information is available in the Visitor's Guide.
All requests to visit the Mopra Antenna should be emailed to ATCA and Mopra site manager Brett Hiscock (Brett.Hiscock @csiro.au).
| © Copyright CSIRO Australia, 1994-2009. Legal Notice and Disclaimer, Privacy Statement. |