Radio Interferometric Baseline-dependent Averaging Studies And Implementations

The Square Kilometer Array（SKA）is the largest radio interferometer telescope under construction in the world,and its immense amount of visibility data poses a considerable challenge to data transmission,storage and scientific processing at the back-end.Therefore,the study of visibility compression methods has become a focus of SKA Science Data Processor（SDP）development.Baseline-dependent averaging（BDA）is a technique that averages the visibility data associated with the baseline to reduce the volume of data effectively.The vast majority of the visibility data obtained in the observation of radio interferometer are from short baselines.Therefore,by averaging the short baseline data through a suitable range selection,it is possible to maintain a sufficiently high imaging accuracy with a significantly reduced amount of data.BDA technology is one of the most likely data compression techniques to be chosen for SKA massive data processing.In this paper,the BDA visibility data compression method and implementation are systematically investigated for SKA-SDP data processing requirements,and the final imaging quality is evaluated through observation simulation and experimental validation methods.The work includes the following aspects:1、The principles of the BDA algorithm are studied intensively,and the BDA algorithm is implemented using the Python language and different dependency pack-ages.The paper evaluates the performance of the different implementations,gives the corresponding performance analysis and determines the optimal BDA implemen-tation.The results show that the BDA processing module implemented in this paper can reach a processing speed of 13 GB per minute in a single process,and that its processing power for visibility data will be further improved in a parallel environment.2、Based on the Radio Astronomy Simulation Calibration and Imaging Library（RASCIL）,the paper implements simulated observations of the SKA1-Low,and com-bined with the BDA module implemented in last work,a fully functional BDA opera-tional test environment is constructed.3、Based on the above work,the paper systematically analyzes the effect of BDA processing results on the final image quality under different settings of the combined averaging interval of visibility data(Δt_max)in full-scale SKA1-LOW observations,and gives the quantitative effect ofΔt_maxon the image quality through simulation data.Overall,this paper has achieved a BDA function with good performance that can meet the data simulation requirements of the current phase of the SKA-SDP and ensure the development of scientific data processing in the current phase of the SKA.Meanwhile,the paper gives a quantitative analysis of the imaging quality and clarifies the variation of imaging quality under different conditions,which is a valuable reference for subsequent SKA scientific research.