Rapthor is a pipeline to process LOFAR data and produce high-quality images. Specifically, it performs direction-dependent calibration in a fully automated self-calibration loop.
The image is a result produced by Rapthor, and has a maximum colour scale of 3 mJy/beam, and thus many sources that are of sub-mJy/beam brightness are visible. It is made from 12 MHz bandwidth of an HBA observation, and reaches an expected noise level of 150 µJy, which for such an integration time and bandwidth is approximately the best that can be achieved with LOFAR NL.
In the future, LOFAR users can directly request the high-quality images from Rapthor when proposing observations, or users can run (and modify) the pipeline themselves. Rapthor makes use of various modern pipelining paradigms, such as for example the common workflow language (CWL). This makes it easy to maintain and support development, and reuse parts of our pipeline for different science cases. Development of the pipeline is helping to push our existing processing software (e.g. DP3, WSClean, PyBDSF) to new limits, and these improvements are already eagerly used by various (expert) users outside of Rapthor. Finally, it also provides a platform to perform new high-performance computing research, and various projects are underway to come up with better and faster algorithms, for example by making use of GPUs.
By releasing our software, we are telling the community that the pipeline is ready to be used by others. Such users can take a look at the release notes and the generic documentation. Our current release targets the processing of Rapthor HBA NL data, but this is just the beginning — in the future, this will be extended to support long baseline and LBA data and to make use of all LOFAR2 capabilities.