Visualizations

Numerical simulations do not automatically produce visualizations of the data they generate. In most cases, you will get a binary-formatted output file, and, in order to access the data buried in this file, special software or some home-brewed code is necessary. Ultimately, to make it visually appealing and understandable, the data needs to be processed and transformed into plots, graphics and movies. This is actually a big and essential part of the art of science – the communication of the results to the scientific and non-scientific audience. This implies your co-workers and your colleagues worldwide, but also your parents or the kids from the high school next door.

R136S00i
Simple false-color visualization of a star cluster from an N-body simulation. Dark and extended points correspond to massive and luminous stars (from Küpper et al. 2011, MNRAS, 417, 2300).

Here are a few fairly old visualizations of mine from some of my projects on tidal tails of star clusters. In order to produce these gif movies, I took hundreds of snapshots from numerical simulations of star clusters forming tidal tails. From the projected positions of the stars in the simulations (below are face-on and edge-on projections), I created mock observations by folding their positions with a point-spread function (a 2d Gaussian), mimicking the aperture of a telescope. The corresponding color and brightness of each star I generated from its temperature and luminosity. Hot, luminous stars appear white, whereas cold, dim stars appear reddish.

Dissolving star cluster on a circular orbit (from Küpper et al. 2010)
Dissolving star cluster on a circular orbit (from Küpper et al. 2010, MNRAS, 401, 105)

Depicted in this movie is a star cluster that loses stars while it is orbiting within a galaxy. Star clusters do this all the time. They can’t help but fall apart. Some do this very quickly, others can resist the tidal forces of their host galaxy for billions of years.

Dissolving star cluster on an eccentric orbit with eccentricity 0.25 (from Küpper et al. 2010, MNRAS, 401, 105)

In this study we were interested in the appearance of the tidal tails – the streams of stars emanating from the star cluster. We found that there’s a significant change in the appearance of the tails when the cluster is not on a circular orbit around the galaxy, but on a rather eccentric one.

M050R8E05
Dissolving star cluster on an eccentric orbit with eccentricity 0.5 (from Küpper et al. 2010, MNRAS, 401, 105)

More eccentric orbits result in more irregular looking streams. But there is actually much order in this chaos, and understanding this order has been a goal of my research at Columbia University in New York City. Below is a different way of looking at the above data. In the upper panel, the number of stars along the tails is plotted in bins of a fixed size, and in the lower panel the mean velocity of stars with respect to the cluster is shown within each bin.

Density profile and velocity profile of a dissolving star cluster on an eccentric orbit with eccentricity 0.5
Density profile and velocity profile of a dissolving star cluster on an eccentric orbit with eccentricity 0.5 (from Küpper et al. 2010, MNRAS, 401, 105)

Clearly, there are several patterns here, which we understand very well now. I developed a method that helps us to understand the trajectories of stars within tidal tails, and also enables us to model them with very little computational effort. Below is an example of such a streakline model (red line) compared to output of one of the N-body simulations shown above (black points).

1204_sim
N-body simulation of a star cluster forming tidal tails shown in four different orbital phases. The stars are shown as black dots, where the star cluster itself is cut out. The red line shows a streakline model, which visualizes the trajectories of stars within the tails (Figure from Küpper et al. 2012, MNRAS, 420, 2700).

It turned out that these patterns can be used to measure the weight of our home Galaxy, the Milky Way. My recent research focusses on the question why some tidal tails look completely different from the ones shown above. Below you can see a simulation of a star cluster on a chaotic orbit within the same host galaxy as the ones above. Chaos, as might be expected, is obviously affecting tidal tails in a severe and seemingly unpredictable way.

Pal13
Simulation of Palomar 13 on an eccentric, chaotic orbit (from Küpper et al. 2011, MNRAS, 413, 863)

 

Advertisements