This website was never intended to be a science blog, but since science is an essential part of my life – with all its ups and downs, kinks and quirks, bores and funzies – I shouldn’t neglect it here! Today, a paper of me and my dear collaborators from Iran, Elham Hasani Zonoozi and Hosein Haghi appeared on astro-ph.
Elham and Hosein sit in Zanjan, Iran, but they’ve spent quite some time at the AIfA in Bonn. I once even shared an office with Hosein for about half a year. So, I’m happy that our collaboration and friendship is still ongoing, even though we’re by many means as separated from each other as you can be on this planet – one of the many examples where science shows that borders vanish once you start interacting with people from the other side.
The paper was a joint project with Pavel Kroupa in Bonn, Matthias Frank in Heidelberg, and Holger Baumgardt in Brisbane, Australia. Our aim was simple but challenging: given the large amount of high-precision observational data (coming from Keck and the Hubble Space Telescope) on the 11 billion year old star cluster Palomar 4, can we infer the configuration it was born with? Our approach is similarly straightforward: set up a ball of stars with McLuster and integrate it forward in time using Sverre Aarseth’s Nbody6. Whichever configuration for the ball matches the present-day observations best after 11 Gyr of dynamical evolution is the winner. This would be a perfect case for a huge parameter-space study, in which you’d go systematically through all of the quantities that set your initial conditions. Unfortunately, each computation takes about one month, so we had to stick to a careful try-and-error approach.
While the physical distance to Zanjan from New York is a mere 10,000 km (6,000 miles), the computers we used were located in Brisbane, and were part of Holger’s amazing GPU cluster. The geodesic distance from New York to the GPU cluster is 15,500 km (roughly 10,000 miles), which is close to being as far as it gets on Earth (20,000 km). In comparison, Palomar 4 is one of the most distant globular star clusters orbiting around our home galaxy, the Milky Way. Its distance from us is 3 billion billion kilometers. It’s as far as it gets in the Milky Way. Light from its stars takes about 300,000 years to reach us – whereas it takes just 8 minutes from the Sun (and 0.05 seconds from Brisbane).
What we find is technical but very interesting. By looking at the distribution of masses of stars within the cluster and their distance from the center of the cluster, we see that Palomar 4 must have been born as a very extended ball of stars and – and that’s the surprising discovery – with a very odd configuration for the masses of the stars. Massive stars must have been much more abundant in Palomar 4 than in young star clusters we observe today. And these massive stars must have preferentially been closer to the cluster center.
It is not obvious why a globular cluster like Palomar 4 should have been born with an overabundance of high-mass stars, with a segregation of massive stars towards the cluster center, and with a very large extent. Models of star formation and cluster formation will have to explain this result. Even more so, because for another globular cluster, namely Palomar 14, we found a very similar result in our pioneering numerical study. Back in those days (2011), our simulations of Palomar 14 were the first-ever simulations of a globular cluster over its entire lifetime. It’s motivating that we were able to now do the same with a second globular cluster, but also puzzling that we arrived at pretty much the same conclusions. What’s going on with these outer-halo globular clusters?