The long baseline group (LBG) forms one of the “study groups” of the Technical Advisory Group and is dedicated to long-baseline commissioning and development.
The LBG is led by a management team (MT) consisting of James Anderson, John Conway, Neal Jackson and Olaf Wucknitz (head). This team has in a first telecon defined a strategy for our next steps.
Additional members may join the MT if necessary, but we try to keep this group small to be more efficient. Being a member of the MT is not required in order to contribute to long-baseline efforts or to have influence on the strategy.
We will discuss our work as much as possible on the USG forum under the Long Baselines board. There are sub-boards on the following issues:
Emails shall be avoided as much as possible because of their non-public nature and the difficulties in archiving. Please activate email notification for those forum (child) boards you are interested in!
While the forum is a good place for having discussions that otherwise would have been on email, it is not that convenient for uploading large documents/memos. In order to facilitate this, a wiki has been created at http://www.astro.uni-bonn.de/~wucknitz/wiki/doku.php/lbg:start . Wherever possible, an announcement on the user forum should be made linking to these larger documents. Most parts of that wiki are publicly readable. In order to make changes, however, you have to register by sending a request to Olaf Wucknitz. You are welcome to do this and add your ideas and results!
Membership of the study group is open to all those willing to contribute effort to the commissioning of long baselines. As a starting point the MT has collected a list of people we believe are interested and able to contribute. This list is kept on the wiki. It will be updated whenever necessary, including to add new members and removing inactive ones. Being on this list does not buy data rights or authorship on papers.
NOTE: Since the discussions and progress reports of the LB study group will be placed on the user Forum it is not necessary to be a member of the study group just to follow developments and keep abreast of the commissioning state of long baselines. Members of the LB study group should be limited to those involved in testing, writing software and contributing algorithm ideas. If you do not want to be asked by other members of the team to contribute to LB commissioning efforts, there is no advantage being on the list.
The MT has identified two main problems we have to solve to make progress with long-baseline commissioning:
Fringe-fitting is the more difficult one, but it can be avoided for bright sources for which phase solutions as function of time and frequency are sufficient. This is therefore more a long-term problem. DFR is more urgent, because it prevents us from applying the standard short-baseline calibration strategy of using XX and YY correlations without full polarisation calibration. DFR mixes X and Y in an unpredictable way.
In order to avoid having to solve for DFR explicitly, the strategy for the near future is to work in a circular basis (R/L) instead. In most regions of the sky, FR can then be described as additional phase terms that are independent in R and L so that separate calibration of RR and LL works in the same way as XX and YY on short baselines. The RR-LL phase differences provides the DFR signal as by-product. The difficulty is that a model for the dipole response is required for the conversion to circular, and that corrections to this model are difficult at later stages.
The circular basis is needed as workaround for the moment, being used as long as BBS cannot successfully solve for DFR. This temporary solution will allow us to gain experience with ionosphere stability on long baselines, find the strength of sources we can detect fringes to, form a list of calibrators, and make first images. The long term aim is, however, to process LB through the pipeline using linear polarisation signals.
The conversion from linear to circular was already applied in Olaf's ad hoc software. Using better dipole beam models, it will now be implemented in a stand-alone program that can be used by anybody. This work is currently being done by James Anderson, Tobia Carozzi, Masaya Kuniyoshi and Olaf Wucknitz.
Once this works (goal end of May), we can start with more regular observations (hopefully at a regular time every week so that international stations can plan their stand-alone activities around this). These test data will be calibrated and imaged in CASA/AIPS/difmap or BBS. We try to establish this monitoring program even before the converter becomes available to allow us to solve possible technical problems as soon as possible and to provide the necessary data for the development. Main goal of the monitoring is to gain a better understanding of instrumental and ionospheric parameters. See Long Baselines - Observations & monitoring.
We have to fix the DFR calibration in BBS in order to allow the analysis of long-baseline observations with the standard pipeline. This requires tests with the current BBS versions and later fixes of the code.
Fringe-fitting has to be implemented in BBS to make long-baseline observations (not only of the brightest sources) possible on a regular basis. Volunteers who know about programming with casacore and for BBS and NDPPP, or are willing to learn, will be needed for this. We are in the process to acquire funding for dedicated positions for this purpose.
If you have experience with the standard pipeline, tests with long-baseline data would be most welcome. As mentioned above, the most serious problem at the moment is DFR. For all but the brightest compact sources, low S/N will be an issue, but sources like 3C147, 3C196, Taurus A and others should work fine.
Tests with MeqTrees or other more advanced software will be required for a better understanding of those parameters that we cannot currently calibrate with BBS.
Because scheduling international observations is more effort than pure Dutch observations, we have to archive important data sets either in raw or pre-processed (NDPPP) form. Storage will be available in Jülich, but this has to be organised and coordinated with the general LOFAR archiving efforts.
Other and more specific requests will be posted on the wiki and the forum under Long Baselines - Announcements.
Please announce your activities there when you start working on any long-baseline issue!