Advancing foreground removal techniques and calibration methods for the next generation of radio telescopes

The current generation of radio telescopes aiming to make a first detection of the Epoch of Reionisation have each come across unforeseen obstacles, particularly when it comes to the conspiracy of foregrounds removal and calibration of the telescope. These obstacles have all largely been overcome though advanced methods of foreground removal which take into account the calibration step would advance this further. This project will explore developing those methods, using current LOFAR data to assess obstacles and solutions and embedding any successes within the SKA EoR Key Science Project data pipeline.

This project will result in a broad knowledge of simulation, instrumentation and data analysis. The first part of this project will focus on developing an understanding of the reionisation signal and sufficient coding and HPC skills, through producing various simulations and understanding the parameterisations behind them. These simulations will be used to test various foreground methods (new methods to be determined). The second part of this project will assess the validity of the methods on LOFAR data, possibly incorporating a calibration-foreground removal feedback loop in order to couple the two approaches more robustly. The use of LOFAR data and the application of different foreground removal methods will allow further early science to be extracted from the data, for example the overall topology of reionisation.

Any instrumental effects which appear within the LOFAR data will be incorporated into SKA instrument simulation software OSKAR in order to make the SKA simulations as accurate as possible.

The third part of this project will involve developing the successful methods/pipeline routes into a modular data pipeline within the SKA EoR working group in order to both design the Key Science Project and eventually process the early commissioning data. The application of these methods on early SKA commissioning data will ideally result in some extremely early science results, namely a deeper redshift and lower noise signal detection.