2024 10 16

Attending: Leah, Nadia, Neal, Hrishi, Roland, Frits, Jort, Marco B., Tom, Jurjen, Étienne, Sagar, Emanuele, Henrik, Jonny, Matthijs, Aleksandra K.

Apologies: Gulay

Software updates

• pipelines: No major updates, dd_selection still in process of being integrated. Benchmarking with new DP3 Ateam clipper showed some discrepancies between the python script and new DP3 when using a data set larger than the test data set.
• facetselfcal: Reinout is working on things, but nothing new now. We are considering re-factoring facetselfcal.py in the coming months because it's getting a bit out of hand! Splitting things up into modules etc. Will need a test dataset to validate/check benchmarks.
• lofar_helpers: No updates.
• containers: No updates.

Polarisation
No particular updates. Ongoing processing of jet in OJ[some source]. Andromeda data - have pol errors, but pol calibration wasn't included in facetselfcal. Looks like facetselfcal works on individual sources.

LBA

Henrik: extending calibration strategy to work on international stations. This also works for poor ionospheric conditions. Also working on other bright sources - e.g. M87, 3C sources. The tough step is to make this work for in-field calibrator sources. Starting to look at this with test data. There is also commissioning data with NenuFAR.

• Neal: memo on expanding to in-field calibrators, not looking great; what is your take? Henrik: saw this but using full data not LBCS data. Interesting to compare the different tests.

Jort: long baselines with 3C sources. Difficult to calibrate the most distant stations. Can get good solutions on German stations. Are there any tricks to extend solutions to longer baselines?

If you want to give it a try, contact Henrik/Jort!

LoTSS reprocessing

Now have 1 field with final postage-stamp images, another field where sources have been split out. Goal is v0.1 catalogue by end of year with around 5 pointings to start working out how to put the final catalogue together.

Improved strategy for DD calibration for wide-field imaging [Jurjen]

Slides: https://docs.google.com/presentation/d/1a-50mF-GqtpbkeLyLLEJGBWeW1ZpmgbKCA8gSnWEFt0/edit?usp=sharing

Questions:

• Étienne: are you invoking nerevar within facetselfcal for these tests? Apart from the cohjones/kafca solvers I think this should be the only thing DDF pipeline does that facetselfcal doesn't in terms of calibration and imaging Answer: we've just compared the outputs [Frits: good to have nerevar but bit of a different philosophy to fix problems by using different chain rather than weighting]
• Marco: Is it true that you have to do 25 subtracts, one for each facet? Answer: no, subtract everything first and add back individual facets
• Marco: once you have your facets, do you average taking into account the size of the facet? Answer: yes! There is a workflow to do this
• Frits: do the final dd solutions have remote stations? Answer: the ones that are reset have the good solutions from the previous step [Henrik: why not constrain them to be zero? Answer: tried it, but better to reset]
• Roland: in terms of nerevar, in terms of performance, what are the current parameters to make it run optimally? There is the ddf version implemented with a flag, use the same parameters as the calibration.

Technical Round Table

deferred for today

AOB

• date of next meeting: 7 November
• Étienne: lots of observations with NenuFAR + LBA - Henrik going through to see which converge with LiLF. Trying to get this started. Once we have an idea of what converges or what doesn't, would be great to get input from later group! This will help finish NenuFAR-LSS commissioning.