The Effelsberg-Bonn HI Survey

Abstract

Since Summer 2008 a new L-band 7-Feed-Array is operated for astronomical science at the 100-m radio telescope at Effelsberg. This receiver will be used to perform an unbiased, fully sampled HI survey of the whole northern hemisphere observing both the galactic and extragalactic sky in parallel - the Effelsberg-Bonn HI survey (EBHIS). The integration time per position will be 10 min towards the area of the Sloan Digital Sky Survey and 2 min for the remaining sky. The sensitivity is chosen to be competitive with the Arecibo ALFALFA and GALFA surveys which are restricted to a much smaller portion of the sky. The use of state-of-the-art digital Fast Fourier Transform spectrometers based on field programmable gate arrays - superior in dynamic range and allowing fast dumping of spectra - makes it possible to apply sophisticated radio frequency interference (RFI) mitigation.<br /> The EBHIS survey will be extremely valuable for a broad range of scientific disciplines ranging from the study of the low-mass end of the HI mass function in the local volume, as well as, environmental and evolutionary effects on the HIMF, the search for galaxies near low-redshift Lyman-alpha absorbers, to the analysis of multiphase and extraplanar gas, HI shells, and ultra-compact high-velocity clouds.<br /> The thesis focuses on developing the data reduction pipeline and software. The program to perform the stray-radiation correction was already available, other tasks like RFI detection, gain-curve correction, intensity calibration, gridding, and source detection were completely redesigned. The implementation of the algorithms was done in the programming language C++, using multi-threading techniques to significantly improve the computation speeds on multi-core or -processor platforms. This aspect is crucial, as the total amount of recorded data expected during five years of observing will exceed several Terabytes. <br /> The software was tested extensively on simulated data, not only revealing the impact of RFI signals on the results, but showing several potential bias and selection effects which need to be considered during the scientific analysis of the data. <br /> Finally, several test observations using the new instrument were carried out at different stages of the development of the receiver. These measurements were used to evaluate the status of the 7-Feed-Array and the FPGA backends, and to test whether the data reduction pipeline provides viable results. As a cross check two of the astronomical sources, Leo T and NGC 2403, were re-observed with the old 21-cm single-beam receiver (but using the new backends). The data were analyzed and basic physical properties derived.