School of Physics - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemMeasuring the Epoch of Reionization Signal with Murchison Widefield Array( 2021)With the emergence of the first stars and galaxies at ~100 million years, the Dark Ages ended. The ultraviolet radiation from the first structures started to reionize the surrounding neutral hydrogen, creating ionized bubbles. Gradually the bubbles grew and merged until the intergalactic medium was completely ionized. This transition was a critical phase in the evolution of the Universe, called the Epoch of Reionization (EoR), spanning the redshift range of z ~10-6. Understanding the critical physics happening in this epoch answers many questions about the structure formation and evolution of the Universe. The observation of the neutral hydrogen 21-cm signal is an excellent probe for tracing the ionization process and studying the underlying physics happening therein. Technically, the EoR experiments utilise radio interferometry which provides the required sensitivity and resolution in the frequency range of interest. The sensitivity of current instruments allows us to statistically measure the power spectrum of the EoR signal, while the next generation of instruments is under development with the ultimate goal of direct imaging of EoR. This work measures the EoR signal at a redshift range of z~6-7 with Murchison Widefield Array (MWA), a radio interferometer located in Western Australia. However, detection of the EoR signal is a challenging procedure due to the low amplitude of the signal (~10mK), bright foregrounds (up to 4-5 order of magnitude brighter than signal), ionospheric distortions, Radio Frequency Interference and instrumental effects. An integration time of ~1500 hours MWA EoR data can potentially detect the EoR signal with an S/N of 14 [1]. However, signal detection is currently limited by the aforementioned systematics. They contribute to the power exceeding the thermal noise level. Therefore, detection of the signal requires the development of different strategies to overcome these challenges. This thesis analyzes the EoR data from MWA while developing, improving and providing insight into systematic mitigation approaches to obtain a more precise measurement. The MWA EoR observing program is targeted on three different EoR fields: EoR0, EoR1 and EoR2. In this work, we measure the signal from two fields. First, we calibrate the EoR data from the EoR0 field and develop a data quality metric for refining the data. As a result, the first deep measurements of power spectrum with MWA, using the RTS+CHIPS pipeline, with ~32hr integration is obtained. The lowest upper limit is Delta^2 <= 2.5x10^4 mK^2 at k=0.14 (hMpc)^-1 and z=6.5 which is consistent with previous results from other instruments. Next, we explore some strategies to mitigate the foreground contamination which is a major obstacle in detection of the EoR signal. We modelled the Galactic Diffuse Synchrotron Emission, the dominant foreground at the redshift of interest, over the MWA field. However, since the MWA is a wide field experiment, it requires a full sky model. Therefore, we explored another strategy, i.e. developing a weighting scheme for baselines based on the severity of their contamination. Another accomplishment in this thesis is the analysis of EoR data from the MWA EoR1 field which has a different foreground containing the bright radio galaxy of Fornax-A with a total flux density of ~500Jy at 189 MHz. A precise model of Fornax-A is essential for effective foreground removal. Using the imaging capability of the analysis pipeline and available shapelet fitting tools, the model of Fornax-A in our sky catalogue is improved. While improving our analysis algorithm, we made an effort to mitigate contamination in our measurements by detecting systematic signatures in the data and excluding them. We explored various features in the data, hunting for the source of systematic signatures. We also recognised the visibility noise RMS as a metric to distinguish the more contaminated data within a refined dataset. Eventually, we obtained the upper limits on the EoR signal power spectrum from the MWA EoR1 field at three redshift bands centered at 6.5, 6.8 and 7.1 with the lowest at z=6.5 of Delta^2 <= (73.78 mK)^2 at k=0.13 h Mpc^-1, from ~14 hr data integration. Although it contains a shorter integration time relative to the previous EoR1 analysis[2] (~19 hr), the limits are lower (~1.26 times), thanks to the improvements in the analysis algorithms, foreground modelling and data refinement strategies. We also compared the analysis results from EoR0 and EoR1 fields. It is shown that, due to the improvements achieved in this work, EoR1 can potentially lead to lower limits (at least on large scales) which warrants analyzing longer integrations from this field. In the final chapter, suggestions for further systematic mitigation and obtaining lower limits are provided.