February 27, 2017

Using Population Models for…Astrophysics?


NASA, ESA, H. Teplitz and M. Rafelski (IPAC/Caltech), A. Koekemoer (STScI), R. Windhorst (Arizona State University), and Z. Levay (STScI). (2014). “Hubble Ultra Deep Field 2014.”

Galaxies are gorgeous structures containing billions of stars. But they come at a formidable price: a super massive black hole. Invisible, mighty, and mysterious, black holes are space regions where nothing—not even planets, stars, or light itself—can escape from inside it once its powerful gravitational pull has swallowed them up.

How can astrophysicists track these alarming—and invisible—regions? This question is at the heart of Yannis Liodakis’ research at the University of Crete. As part of his doctoral work, Liodakis has come to CDS as a visiting scholar so that he can learn more about how data science techniques can solve questions about black holes.

At last Wednesday’s Research Lunch Seminar, Liodakis explained one strategy that astrophysicists have been using to locate black holes. Around every black hole is an accretion disc, a ring of space matter that has been caught by the black hole’s gravitational pull, but not yet consumed by it. Accretion discs contain smaller masses called quasars, which emit strong electromagnetic fields and, most importantly, light. These light jets are called blazars, and their powerful beam can be seen from Earth.

Blazars are important, Liodakis stated, because calculating how bright they are then enables us to calculate its velocity. After calculating a blazar’s velocity, we can then not only calculate its distance and map its location, but also predict where the invisible black hole is, because blazars exist on the accretion disc surrounding the hole.

But a major problem is that the light jets blazars emit to Earth oscillate due to the quasar’s electromagnetic field. In other words, the blazar’s light bends at varying degrees as it travels towards Earth, meaning that the light we would see through a telescope is inaccurate.

Astrophysicists must therefore calculate how much the light rotates as it travels to Earth to produce an accurate measurement. Additionally, they have to remember that there are competing time scales involved: although the universe is always expanding, the blazar’s light jet is also compressing time while it travels to Earth. To account for all of these complications, Liodakis and his team have recently been experimenting with population models that are typically used by data scientists.

Population models are powerful because they can show us how highly complex factors of a given topic interact with each other. For example, ecological population models are usually concerned with calculating how population size, age distribution, and physical environment affect a given group of organisms.

Applying population model techniques for astrophysics has allowed Liodakis and his team to produce sharper calculations about how bright blazars are because they can account for all of the elements of the problem—like oscillation and competing time scales—in a single model.

Liodakis’ innovative work demonstrates how interdisciplinary data science strategies can become. From astrophysics to ecology, who knows where data science will be needed next?


by Cherrie Kwok