Master's Thesis projects @ Uni of Heidelberg

[1] The impact of galaxies on the ultraviolet background

At redshifts 5<z<6, roughly 800 million years after the Big Bang, it is believed that light from the first galaxies is finishing to ionise the diffuse hydrogen in the intergalactic medium. The Universe then becomes fully transparent to UV light after this process known as "reionisation". Even though it is nearly a consensus that early galaxies are responsible for reionisation, very little direct evidence exists linking them to a decrease in neutral hydrogen in their environment. Using the new data released by the groundbreaking observational program "XQR-30", we can now start to look for such direct connections.

The student will use spectroscopy of the Lyman-alpha forest - the intergalactic medium seen in absorption in the foreground of distant quasars - to look for a decrease in neutral hydrogen in the surroundings of early galaxies. Conveniently, the locations of the galaxies have been pinpointed by the imprint of the metals in their environments. The detection of a signal linking lower levels of neutral gas to the presence of galaxies at the end of reionisation would be the most direct proof so far of early galaxies' role in the process, and even the *lack* of a signal would be very useful in steering future models.

After a signal is detected (or not), the student will have the opportunity to deepen the project either in the modeling direction (comparing the signal with reionisation simulations) or further in the observational direction by using the lack of galaxies and the higher-frequency Lyman-series to identify true regions of neutral hydrogen in the data.

Main skills involved:

> Physics of reionisation, the IGM, and the first galaxies.

> Data analysis: spectroscopy and large data catalogs.

> Any coding language, but a common one would be preferred, such as python, C++, or idl.

> Statistical analysis.
> Applied cosmology (basic, course not required).

[2] When does primordial Helium become re-ionised? 

Background: Primordial Helium is created during Big Bang nucleosynthesis, and makes up about 25% of baryons in the intergalactic medium to this day. This primordial Helium quickly recombines, existing in the form of neutral Helium atoms for billions of years. While Hydrogen quickly gets photo-dissociated into its ionised form by the first stars, neutral Helium survives for much longer. This is because only accreting supermassive black holes (quasars) are capable of producing en masse the high-energy photons required to ionise primordial Helium. The timing of Helium reionisation is not currently known, but it is suspected to coincide with the "golden age" of quasar activity around a redshift z~3-4.

This project: Our group has recently pioneered a new way of detecting the process of Helium reionisation directly by looking at subtle variations in the temperature of the intergalactic medium. Photo-dissociation leaves the gas in a highly heated state, and since quasars are very rare sources, the resulting temperature fluctuations can be sufficiently extreme to be detectable. The method involves accurately measuring the optical depth of the Lyman-alpha forest on large scales, and confronting observations with state-of-the-art numerical simulations. We then modify the amount of temperature fluctuations in the simulations to match observations. Our first attempt (soon to be published by a former student) ruled out the presence of temperature fluctuations at 3.7<z<4, implying that Helium reionisation needs to happen before or after this time. Now with the eBOSS sample, we can repeat this successful method at 3<z<3.7 and hopefully actually detect the signal. Even a non-detection would be extremely interesting, since it would have strong implications for the properties of the quasars which drive Helium reionisation.

Skills acquired during the project:
Spectroscopic analysis
Cosmology: large-scale structure, intergalactic medium, Lyman-alpha forest, temperature fluctuations
Working with large databases: eBOSS database of quasars
Applied machine learning: quasar continuum reconstruction methods (mostly applying already-trained methods, but development is possible)
Cosmological simulations analysis and post-processing

Requirements: Basic familiarity with python; good grade in Cosmology course (better than ~2.5)

[3] Hunting for gravitational redshift at the highest redshift

Under the framework of General Relativity, light escaping from regions located close to a black hole's event horizon is expected to experience gravitational redshift toward external observers. Detections of gravitational redshift have been claimed in light coming from the accretion structures around quasars, which are the most massive and brightest black holes. Specifically, light originating in ionised gas clouds very close to black holes in the ``Broad Line Region'' have been claimed to emit Fe~{\small{III}} photons at wavelengths shifted towards the red. Could this be gravitational redshift?

The challenge in measuring Fe III shifts comes from having to know, independently, the distance (and therefore redshift) to the quasar itself $-$ this has not been done with good precision in previous studies. In this project, a student will analyse for the first time a collection of spectra of quasars in the early Universe ($z>6$) for which distances are known precisely from observations of host galaxies. They will then compare the detected redshifts in a sample of later-time quasars where distances are similarly known. Since gravitational redshift from escaping a BH gravitational well should get stronger for larger BH masses, we will check if the observed Fe~{\small{III}} redshifts correlate with BH mass. If they do, it will mean that we have detected gravitational redshifts in the first billion years of the Universe, and will provide a new way of measuring black hole mass. If no correlation is found, this would disprove previous claims of gravitational redshifts. 

In practice: This project uses spectroscopic analysis, including emission line modelling. Other skills include working with large databases and quasar physics, such as measuring black hole masses. \\

Requirements: None. Astronomical Techniques and familiarity with python are referred. \\

Difficulty rating: Hard. While the individual analysis steps are not complex, they must be conducted very carefully, and the project represents a fairly large amount of work. On the other hand, a publication of the results would be highly impactful even with a null result.