Galaxy Builder

Galaxies are complicated. We need your help to build models of them.

This project lets you create a model of a galaxy image, which we can use learn about its structure, and how it assembled over cosmic time.

More than the sum of its parts

Galaxies can be broken down into multiple components. Most galaxies can be built with a bulge (a kiwi fruit-shaped bit) and a disc (shaped like a dinner plate). Many also have bars and spiral arms. Being able to separate the light from these different components will allow us to study the star formation history of each separately, as well as helping us understand the galaxy's dynamical history (such as any galaxy mergers it has been through).

For example, the presence of a "classical" bulge (one that's very centrally concentrated, and likely does not rotate) in a galaxy suggests that it has undergone some kind of merger event, with the brightness of the bulge relative to the galaxy disc tied to the strength of the merger.

How do researchers analyse galaxy structure?

Previous Galaxy Zoo projects have produced a wealth of new science, but most have been limited to simple multiple-choice questions. To obtain precise mathematical descriptions of galaxy structure scientists make use of computer fitting algorithms to automatically find the best model for a given galaxy.

The problem with computational fitting is that when galaxies become complicated the computer struggles to find the best model, as it is generally just trying to maximise some goodness-of-fit statistic. This results in scientists having to check the results and often control things manually, which isn't scalable with the size of modern surveys. A recent study by Gao and Ho found that just performing a detailed structural analysis of only 10 galaxies was unfeasibly time-consuming using current tools. The 10 galaxies they looked at can be seen here:

Caption: Gao & Ho (2017), Figure 1

 

One of the take-home messages from their paper was that it is important to properly model secondary (complicated) components, and future morphological analyses should attempt to take them into account.

This is where the strength of citizen science comes in: by letting you build models by hand you can come to decisions using previous experience and physical intuition about how the galaxy should be modelled, and by combining your answers with computer fits we'll be able to get much better results than either humans or a computer alone!

Acknowledgements

We would like to thank Research at Google for part-funding development of this project through a Faculty Research Award. Tim Lingard also acknowledges funding from STFC Studentship ST/N504245/1.