The Monthly Media for May 2022 is from student Jennifer Shi who recently completed a vacation research scholarship at our Swinburne node. Jennifer was measuring key properties associated with known galaxy-quasar pairs, under the supervision of Dr Rebecca Davies, to better understand the origin of the physical processes behind gas that flows in and around galaxies.

Quasars are the brightest sources of light in the Universe and they can be used to study the properties of gas flowing in and around of galaxies when they are positioned such that the galaxy is between our telescopes and the distant quasar, as shown in the below image.

Light from the extremely bright quasar travels through a galaxy that by chance sits along the line of sight from an Earth telescope to the quasar. As the quasar light passes through the gas of the galaxy, the gas absorbs some of the light based on what the gas is made of, how it’s moving around the galaxy and the distance between the galaxy and the quasar. This affects what we see of the quasar light from Earth, as the gas absorbs some of the quasar light. Image Credit: Michael Murphy.

When quasar light passes through the outskirts of a galaxy, a series of absorption lines are produced when some of the photons (particles of light) are absorbed. These absorption lines can be thought of as a cosmic fingerprint – it encodes many important properties such as chemical composition, velocity of the gas and distance to the galaxy. The critical step is to identify the galaxies that have absorbed the light from the quasar.

In order to identify galaxies that have absorbed quasar light, Jennifer analysed ‘datacubes’ from the Multi Unit Spectroscopic Explorer (MUSE) instrument on the Very Large Telescope, in Chile. These datacubes, or integrated field spectroscopy data as they’re officially called, contain information on two spatial dimensions (the image) and one spectral dimension (the colour). This means each pixel contains a full visible spectrum of light information.

Jennifer filtered through MUSE datacubes in search for line-emitting galaxy candidates. Once a promising quasar-galaxy pair was identified, she then measured the distance between the galaxy candidate and quasar (this distance is the “Impact Parameter”), redshift (galaxy-Earth distance) and velocity.

The absorption lines produced by the compound Magnesium II (MgII) is often used as a tracer for gas in galaxies as it is common in the types of galaxies Jennifer was looking for and is easily observed in the quasar spectra. By measuring the strength of the MgII line (its Equivalent Width) and comparing it to the distance measured between the quasar and galaxy, she was able to compare her results to previous surveys looking for the same thing, as shown in the image below.

The strength of the MgII signal against the distance between the quasar and galaxy. The blue line indicates the expected correlation between these two parameters (as in Nielsen et. al. 2013), such that as the distance between the two objects increases, the strength of the MgII signal decreases. The dots are Jennifer’s results, which do not show this correlation. This suggests there is something else going on with the data.

The above figure plots the expected correlation between distance and signal strength, along with the results from Jennifer’s research. Her results do not match up with the expected results, suggesting that something else is happening in the data. The original sources that Jennifer filtered through were very weak, making it easy to find a signal where there is none (known as “noise”). This is likely to be the cause for the mismatch between Jennifer’s data and the expected result, and is a very common (though rarely discussed by the media) occurrence in science.

Jennifer’s first attempt to sort the true data from the noise was to look at the difference between the velocity of the absorption lines supposedly from the gas around the galaxy and a particular emission line (known as the Lyman alpha) from the galaxy. She hypothesised that the candidates that have the smallest difference would best follow the expected curve. This was not to be, as seen in the plot below.

There is no correlation between the difference in velocities and the expected curve. A different method for separating data from noise will need to be employed.

Jennifer’s next step is to look at the statistics of the quasar-galaxy pairs. She aims to estimate the expected number of candidates in each of her datacubes and compare it to the number of candidates she actually found.

Keep an eye out for her progress in a future Monthly Media!