LOFAR2.0 is getting its final shape and the new station hardware is now tangible. In May we celebrated the inauguration of the Dwingeloo test station (DTS). As you will see from the contributions in this Newsletter, the first test results of the DTS are very encouraging! On July 1st the tender to procure the LOFAR2.0 station hardware has been published. This will be followed shortly by the tendering of the Timing Distributor hardware, which is also ready for production. We are of course extremely interested in the outcomes given the challenging economic situation. A lot of work went into these milestones, and a compliment is well deserved!
In this Newsletter we share the highlights of the LOFAR Development program. This program delivers new LOFAR capabilities enabling super-cool science projects, makes LOFAR more efficient and easier to operate. An example of the latter is the TMSS system that is now used in production. TMSS will gradually replace the old MoM system, making life for the operators, science users and software maintainers much easier. Well done TMSS team!
I stop before I spoil all the other highlights in this Newsletter as well. I hope you enjoy this 18th edition of our Newsletter. Happy reading!
Solar Eruption through LOFAR2.0 Eyes
After rigorous lab-tests of the newly developed components of the LOFAR2.0 Station, we continue their verification in the outdoor Dwingeloo Test Station (DTS). Most of the DTS tests use stable and predictable signals and sources. However, in our test also unexpected events are captured. Recently, we were happily surprised that one of our test observations included the detection of a Solar eruption with both the LBA and HBA and at the same time!
With the Dwingeloo Test Station, we observe the sky almost every weekend for stability and long-duration tests. As a coincidence, LOFAR was used by Pietro Zucca for observing Solar eruptions during the weekend of the 21st of May. He recorded several eruptions in that timeframe. Knowing the exact times of these eruptions, we looked at our recorded DTS data and found that our test station also detected the eruption! In our case, only a single LBA antenna and HBA Tile was used and therefore the sensitivity is limited. However, the bright eruption at 14:33 UTC is also clearly visible in our data. It is even visible at frequencies above 100 MHz, nicely showcasing the LOFAR2.0 improvement of simultaneous recording in both the Low Band and the High Band!
This result is obtained with a limited system. We will convert core station CS001 to the first full-scale LOFAR2.0 Test Station. That station will have all LBA antennas and HBA Tiles connected to the LOFAR2.0 backend, which is needed to verify all interfaces of the station. It will also boost the sensitivity and further improve the result shown above. With today’s result, we know that we can observe bright phenomena such as the Solar eruption and we are on the right track to improving the LOFAR Telescope!
In the last newsletters we introduced the new test station on the ASTRON premises, called the Dwingeloo Test Station (DTS). Finally, after lots of preparation and work by the whole team, we have taken DTS into ‘production’ as the LOFAR2.0 test station.
To mark the opening of DTS and the fact that we have reached an important milestone in the project, a festive opening celebration has been organized on May 10th. The new ASTRON director, Jessica Dempsey, was present to witness the opening event, as were many ASTRON colleagues, colleagues from INAF, the crew of a local TV station and a newspaper reporter. We were happy to see that our news item was quickly picked up by national and international news outlets. A picture compilation of the opening event can be seen below.
After these celebrations, the scrutinous testing of the prototype hardware in DTS has commenced. This is organised in weekly cycles as we want to make sure that the system is used in the most effective way. Over the weekends we run long sets of observations and analyse all data (antenna data and station monitoring) immediately thereafter.
The first results (see the figure for an example) show that the system is performing in a stable manner without any real issues, so that certainly looks hopeful for the weeks to come. More detailed analyses and more focused observations will learn us more about the system’s behavior and whether we can fulfill all the requirements.
According to the current planning, we run the weekly DTS test cycles until end of August, and then we will physically move the DTS hardware to the LOFAR1 station CS001. That will effectively start the LOFAR2.0 Test Station (L2TS) phase of the project.
At first, the L2TS will run with DTS hardware. We intent to expand that to a full LOFAR2.0 station as soon as we can. That requires additional manufacturing of boards for L2TS, for which we have placed most of the orders. For the RCU2s, we have set out an invitation to offer with several potential manufacturers; the total cost all RCU2s for L2TS demands us to do so. We will decide on a manufacturer in early July. Delivery timescales are highly uncertain in the current market, but we hope to be able to extent L2TS to a fully equipped LOFAR2.0 station early in 2023.
The program made a final decision on the hardware to be purchased for the Timing Distributor upgrade; this is to rollout White Rabbit (WR) switches (left image) in all LOFAR stations. The alternative option was to use WR switches only in stations that would be distributing timing signals to other stations, and for the ones considered end-nodes, White Rabbit Len modules (right image). This setup was considered because it represented an important reduction in hardware costs. However, after some issues encountered during testing and considerations regarding how the timing distribution system interfaces with the rest of the LOFAR system, the confidence on the reliability and flexibility of the WR-Len modules was questioned and the decision was changed.
Now we are preparing the tender documents; to be published during the summer. The activities in the project are naturally slowing down as we are now in the procurement phase; the main focus is to have the detailed rollout plan in place and make sure to cover all the logistical aspects. Only after receiving the quotations from the manufacturers, we will have a clear idea of the timeline for the rollout, that will be executed in 2 steps: the first one for the remote stations and the second for the core stations.
Telescope Manager Specification System (TMSS)
- TMSS is live! The first projects of LOFAR observations are now managed by TMSS. The LOFAR LBA Survey has used TMSS since April 25th. Overall, operations have been quite successful for a new system, thanks to extensive testing and bug fixes beforehand. Some remaining issues had to be resolved or required some additional workarounds for the time being. This is all expected.
- We have celebrated going live on our TMSS Review / demo meeting on June 10th where we gave a system overview. For this day we invited the developers from our ILT partners for the day, our first live meeting since corona, and of course had a remote connection with our Indian developers.
- Commissioning is ongoing to accept other observing strategies. Due to other obligations, this could not all be performed before the start of Cycle 18, but it is expected that soon all projects will run with TMSS.
- Development is funded till the end of October. Development is focused on:
- Fixing issues in running production observations and reducing the response time of using TMSS.
- Supporting the scheduling set editor for efficient support (current sprint)
- Being able to run operations without the need of MoM
- Some modes that are not fully supported yet may be unavailable for a while until support is added, e.g., the prefactor/LINC pipeline
- Finish development of a basic form of dynamic scheduling
- Move reporting of daily observations will from JIRA to TMSS.
- Finish project, cycle and failure reporting.
- Partial COBALT2 support
- We want to support fully the COBALT2 functionality that can currently be specified
- In scope are:
- Support for long baseline observations (doppler correction)
- Support for spectral line observations (spectral leakage fix)
- Support for 8-bit raw beamformed data
- Support for multiple simultaneous beamformers
- Out of scope are:
- Pulsar pipeline support for multiple pipelines and/or repointings
- As the complexity is high and resources are still limited, this may not all be completed by October.
Brief update on two exciting software projects
(i) AARTFAAC is an all-sky radio telescope and transient-detection facility. It piggybacks on raw data from the LOFAR antennas. The PADRE project improves the AARTFAAC processing pipeline by detecting transients in real time and with low latency. The PADRE pipeline will scale to support more antennas and increased bandwidth, through algorithmic optimizations and the use of GPU acceleration. The image shows the improved PADRE pipeline where the correlation to imaging steps now takes only 6 seconds. (ii) The CORTEX project will show how data driven pipelines can make use of new techniques (e.g., auto-tuning and AI) to produce optimal results. For LOFAR, CORTEX will make widefield, international baseline imaging a commodity. The image shows the results a new facet calibration technique applied to deep LOFAR imaging of the merging galaxy clusters A399-A400 (de Jong in prep).
DANTE started officially in May with a kickoff meeting at ASTRON, where the project team shared an overview of the development plan. DANTE is organized in two phases; the kickoff meeting marked the start of Phase 1 that will deliver a new design of the HBA front end (AHBAFE) and will prepare for the following phase. Phase 2 will deliver the design of a dual beam capable tile. At the end of the meeting participants could provide feedback to the team. After processing the comments, several actions were identified by the team and they are being followed-up.
In terms of activities: requirements for the AHBAFE were defined before the start of the project, the reviewing process is being now finalized. In addition, we are currently working on the design optimization. The sketch shows the design of the printed circuit board that will host all the components that make the frontend, with the proposed configuration. The idea is to optimize the design to provide maximal performance, is suitable to be in the field for years to come, and that reduces manufacture complexity and maintenance. All that remaining at lowest cost possible. One of the priorities at the moment is to have a complete BOM (bill of materials); the list of components needed to produce the prototypes, so we can place orders as soon as possible, knowing that in the current situation prices and waiting times for components might be driving the timeline of our development.