It's a beautiful time. The covid measures have been released, enabling us to meet each other in person much more frequently at the office. It's great to have these spontaneous conversations with colleagues again, giving new energy and inspiration. In the coming period, we will explore how to combine the good things from working at home and at the office.
At ASTRON, you can now witness the test and verification of the LOFAR2.0 hardware. As the design of most LOFAR2.0 parts has been completed, the focus of the development teams is now on testing and procurement. So far, the results are impressive! Besides the activities described in this Newsletter, the first steps towards the procurement of CEP6, LOFAR’s central pre-processing cluster, are being taken. The goal is to replace CEP4 by CEP6 in the second half of 2023.
Development and validation of TMSS is almost done and we are eager to get TMSS up and running and to phase out the old systems. It is hard to imagine the enormous improvement that TMSS provides for those not involved in specification and scheduling. TMSS is going to make operations and maintenance much more efficient. The TMSS team managed to deliver a system that preserves LOFAR’s flexibility, while being much more user friendly. Well done!
Compliments to all who contributed to the impressive results presented in this edition of the Newsletter, and of course: Happy reading!
On Monday April 11th we held the first LOFAR2.0 Large Programmes workshop to discuss the 20 submitted Expressions of Interest (EoIs) we received. Representatives from the 20 submitted EoI science teams, SDC and the LOFAR2.0 programme were present for this full-afternoon online session. Boudewijn Hut gave a highly motivating summary of the great progress in LOFAR2.0 development. Jason Hessels then led a discussion about the scope of the LOFAR2.0 Large Programmes. We discussed the potential for commensality, which is critical for maximising the amount of science we can do per observing hour. We also discussed the logistics of running the programme, processing the data and storing it in the long-term archive. It is clear that the community has ambitious scientific plans, and we need all the storage and processing we can get if we want to take full advantage of LOFAR’s long baselines. A follow-up discussion on April 20th is scheduled to discuss the details of the imaging setup. These workshops are a key step towards defining the full LOFAR2.0 Large Programmes portfolio.
A group of scientists uses the LOFAR telescope to gain more understanding of thunderstorms and lightning flashes. LOFAR’s frequency range, large number of antennas and transient buffer capability make LOFAR one of the best instruments in the world to study lightning. For lightning observations, LOFAR is used in a special mode-of-operation that is different from the modes that are used for most astronomical observations (that is, the “imaging mode” or “beamformed mode”). The LOFAR2.0 upgrade is primarily aimed at these typical modes of operations. In 2021, the transient buffer functionality was descoped from LOFAR2.0 to save costs. Without the transient buffer, lightning research would no longer be possible with LOFAR2.0. The lightning-science community would lose the unique instrument that allows them to study lightning flashes at very high spatial resolution.
Brian Hare, a lead scientist of the lightning research with LOFAR jumped into action. He was awarded an ERC Starting Grant for his project “LIFT”. The project includes adjustments that are needed to make sure that LOFAR2.0 will allow this type of research in the future. The figure shows an overview of the science requirements. This reflects the need and is used to optimize the system for this type of research. Some preparatory work is done now and in the following months, after which the adjustments can be made.
With these adjustments in place, LOFAR will keep its special modes of operations for this type of research. A larger scientific community is served with valuable data and LOFAR will be used to reveal the answer to the questions above.
Since our last report in these newsletters, we have focused on two things, mainly.
First, we have finished a large stack of documents intended for the tendering of the newly developed LOFAR2.0 hardware. Almost all major boards will be tendered this spring and so it was important that the technical documentation, drawings, component lists, etc. were all in sync and up to date with the latest changes we have made to these boards. In a large effort that went on for several weeks we managed to get it all lined up and ready for usage as documentation for the tender. Nico Ebbendorf is now finishing the high-level overview and requirement documents and then, after a final review, the tender can be published.
Second, after writing the technical documentation, the team moved on to prepare our DTS test station outside. To get a better grip on organising this effort, we decided to handle it as a ‘real’ first station rollout. We identified all parts and components necessary to add to the system (incl. cables for power and network), and their properties (connectors, color, length, etc.). Based on this inventory we make sure we have all components and cables ordered and available. Detailed 3D-drawings of the cabinet layout will help decide what goes where during installation and how to connect it with other components. A complete inventory will be set up this way. The result of this process will help us with future installations, for instance when we move the DTS equipment to CS001, this summer.
Of course, these have not been the only team accomplishments in the period since the last newsletter has been distributed. A (incomplete) summary of other activities is given in this list.
- Measurements on the susceptibility of the subrack to received and self-generated RFI. This was done using ASTRON’s EMC room with new equipment setup for the Wireless Data Lab initiative. A lot has been learned about calibrating the EMC measurement equipment, and about the subrack behavior. Results indicate that the subrack is well within the specified requirements.
- Measurements of the phase (and thus delay) behavior of the subrack with increasing temperature. These were done using the Climate chamber. Also here, results indicate excellent phase stability (with 1 degree for low band, 2 degrees for high band) over a large temperature range up to 50 degrees Celsius.
- The analog beamforming (‘tile beamforming’) has been tested with the RCU2Hs. Final tests have to wait for DTS-Outside to be completed, but we have at least demonstrated that through the software and the RCU2Hs we sent the same weights to the HBA tiles as LOFAR1.
- The digital beamforming (‘station beamforming’) is almost ready for testing with the SDP firmware on the Uniboard2s. The amount of weight numbers to be calculated and sent to the firmware is more challenging here, esp. for the interface software, the translators.
- We have finished and tested the firmware data ring development for correlator and beamformer mode, for up to 8 FPGAs. There has been a Daily Image showing the celebration of this milestone.
- First demonstration of a fully working monitor, archiving and alarming chain in software.
- Complete the automated boot sequence of a subrack, including failure reports and failure actions.
This list of accomplishments is already overwhelming, and even more has been achieved that I don’t have the room for to mention here. There should be no doubt that the LOFAR2 team is pushing hard to establish a working DTS-Outside system that is ready to run the test program.
WR-LEN unit for station
To enable easy mounting of the WR-LEN hardware in a LOFAR station a 2U rackmount has been made at ASTRON by S. Kuindersma. In here we mount the WR-LEN and the optical multiplexer we need to combine the RX and TX signals onto a single fiber to connect to the Central Core.
When we roll-out the timing hardware we can assemble and calibrate all the station components in the lab and send it as a unit to a LOFAR station, where it can be installed easily.
If we ever have a need to replace hardware at a station this unit can easily be replaced with another unit after which we can work on it in the lab.
In the pictures we see the optical multiplexer installed at the left and the WR-LEN on the right.
The program is working out the decision on the hardware to be purchased for the TD upgrade. The procurement preparations will start immediately after. In the meantime, the TD team focuses on preparing a detailed plan for the rollout.
- The LOFAR observations are soon to be managed by TMSS, the Telescope Management and Specification System. TMSS will replace MoM as an interface for checking specifications. It has been developed over the past two years and we are very soon going to start running the first project with TMSS with more projects to follow in the coming months. If Telescope Operations and Science Data Centre Operations approve it, the first project will run with TMSS starting on April 25.
- Over the past months we have worked on scheduling, on combining beamformed and imaging observations, and most importantly, on validating the data and metadata, first for the LoLSS LBA Survey strategy. After a final approval, the system will be ready to run the LoLSS survey in production. Development and validation will continue to accept all observing strategies (the LoTSS survey, pulsar timing, etc.) and add the user administration (currently ongoing), dynamic scheduling and move reporting of daily observations from JIRA to TMSS. The goal is to run all Cycle 18 observations with TMSS, and, when ready, with dynamic scheduling. Switching from JIRA to TMSS for reporting will take place at a later stage.
- TMSS supplies one interface and an integrated process for specification, scheduling and reporting. It is also an update to modern web technology, which can relatively easily be expanded to support additional use cases such as COBALT2 and LOFAR2 stations and be interfaced with additional services. The system supports a selection of standard observing strategies, each requiring in itself only a minimal specification. A spreadsheet-type interface is available for a quick specification of many observations. A direct system interface will also be available for experienced power users for specification and for following the progress of their projects. Through dynamic scheduling and integrated reporting, we aim to optimise the operational procedures and make operations more efficient. The remaining time can then be used to improve the instrument further.
- We are excited to start using the new system. Keep an eye on the daily images for more updates in the coming weeks.
Radio sources associated with the nuclei of active galaxies (RLAGN) are often morphologically complex objects that can consist of multiple, spatially separated, components, such as jets, hotspots and radio lobes. Before we can find the radio object's corresponding optical host-galaxies and infer its physical size and luminosity, we need to group these components together.
Radio detection software, like PyBDSF, was not designed to group these well-resolved spatially separated components together, so up till now, we are grouping these components by hand.
However, applying this manual process to the thousands of well-resolved RLAGN that appear in the images from the LOFAR Two-metre Sky Survey (LoTSS) is a fearsome, time-consuming process, even when delegated to many persons.
Using a machine learning approach, we casted the association problem into a classification problem and trained an adapted Fast region-based convolutional neural network (Fast R-CNN) to mimic the human expert annotations from the first LoTSS data release. This architecture, originally designed to detect the location of everyday objects like cats and dogs in FaceBook-images, now broadens its horizon by inspecting LOFAR images and can help us to improve our radio catalogues.
After more than 10 years of continuous operations of LOFAR, the HBA tiles face degradation that concerns mainly the frontend boards. DANTE will ensure that the HBA continues delivering science for the years to come.
As we prepare to start our activities in May, our team has been busy working on the requirements and the architecture of the HBA front end. The study of the architecture had the objective ensuring that the design of the new HBA front ends can be used later on for a HBA dual beam tile. Now, we are confident we can start developing the new front ends without precluding the dual beam option.
We have also been working on the definition of our project management plan and defining a way in which we work together, how do we communicate our results and keep our stakeholders updated of the progress.
DANTE will have a kickoff meeting in May, soon we will be able to share pictures in this newsletter.