Round #2 science is complete! You can continue to classify, but we no longer need further classifications on these images. Learn more here.

We hope to have more data in 2014. Meanwhile there are other Zooniverse projects that need your help:

Classify Archive Data

We're on a collision course with the Andromeda Galaxy.

Help researchers understand the awesomeness of the Andromeda galaxy, because one day we'll be in it...

Adler Planetarium
GLORIA
PHAT
Hubble
Zooniverse

...and a special thanks to Campfire NYC for the awesome URL

About PHAT

The Andromeda galaxy is the closest spiral galaxy to our own Milky Way. For a hundred years, Andromeda (also known by its Messier Catalog identifier, M31) has played an important role in shaping our view of the Universe. In the early 1920's, Edwin Hubble's observations of Andromeda confirmed for the first time that galaxies lie outside of the Milky Way, and that Andromeda must contain billions of stars. Today, Andromeda is a template for understanding how spiral galaxies form and evolve.

The Panchromatic Hubble Andromeda Treasury (PHAT) survey (public webpage here) opens a new window on Andromeda. This four-year Hubble Space Telescope (HST) project has recently finished mapping one-third of Andromeda's spiral disk at six wavelengths ranging from the near-infrared to the ultraviolet (Dalcanton+ 2012). The HST images have exquisite resolution, allowing PHAT to reveal more than 100 million stars in Andromeda. This beautiful data set is the heart of the Andromeda Project.


The Andromeda galaxy with the PHAT coverage footprint overlaid (top). On the bottom is a small section of the PHAT data showing some of the millions of stars visible in the HST images, along with two star clusters (Image Credit: Robert Gendler & Zolt Levay).

Hunting for Stellar Clusters

Star clusters are collections of hundreds to millions of stars that were born at the same time from the same cloud of gas. This shared origin makes star clusters unique tools for understanding how stars form and evolve. Additionally, they are useful for studying the major chapters in the history of galaxies. But before Andromeda's star clusters can unlock these secrets, we need the help of Citizen Scientists to find the clusters. Not just the big bright ones, but the small faint ones as well. This is the goal of the Andromeda Project.

Star clusters vary greatly in terms of mass, size, age, and local environment. As a result, star clusters can appear quite different from one another depending on the properties of the clusters and where they are located in the galaxy. This makes the process of identifying clusters tricky and difficult to automate. From the first year of PHAT data, a team of eight astronomers searched through each image, manually identifying star clusters by eye. Using less than 1/5th the total PHAT survey area, we cataloged about 600 star clusters (Johnson+ 2012). With the Andromeda Project, we hope that you will help us find the thousands of star clusters hiding in the rest of the survey!

Because the appearances of star clusters vary so much, it is important for us to learn what kinds of clusters we can actually see. For this reason, we have inserted realistic synthetic clusters with known ages, masses, and sizes into some of the PHAT images. By identifying both real and synthetic clusters, we will learn what types of clusters are undetectable in Andromeda. This information is critical for understanding the age and mass distributions of the clusters by allowing us to determine whether certain populations of clusters do not exist or if they are simply avoiding detection.

After you help us to find these star clusters, we will use several techniques to determine the age and mass of these objects. In some clusters, we can resolve individual stars that allow us to determine the age, mass, and other aspects of the object. In other clusters, where individual stars are too faint or too close together, we can use the summed light from a cluster to determine its properties (Fouesneau+ 2012, in prep.). With these ages and masses in hand, we can use these clusters to study a host of interesting topics: rapid and rare stages of stellar evolution, the structure and scale of star formation, the evolution of cluster populations, and how Andromeda has changed over billions of years.

Distant Galaxies Peeking Through Andromeda's Stars

Star clusters are not the only interesting things in the PHAT images. Although Andromeda's stars dominate the images, parts of the galaxy are transparent enough that we can see distant galaxies that lie far beyond Andromeda. These "background galaxies" are excellent tools for studying the structure of Andromeda's interstellar gas and dust because they can reveal details about these dark dusty structures at small size scales. The background galaxies can also masquerade as stellar clusters, so we need your help to carefully sort these objects.

Science Team

The Panchromatic Hubble Andromeda Treasury science team is to be found all over the world. Listed here are the team members involved in bringing the Andromeda Project to life and who will be taking your classifications on the site and turning them into results.

Julianne Dalcanton

Julianne is the Principal Investigator of the PHAT project, and a Professor of Astronomy at the University of Washington, where she is privileged to work with awesome students every day. She is an avid snowboarder and loves her pitbull.

Benne Holwerda

Benne is a postdoctoral fellow with the European Space Agency studying dust in other galaxies, often using background galaxies to highlight the dust. He likes to join in when his 2.5 year old daughter points out the Moon or his 7 month old son has a fit of the giggles.

Cliff Johnson

Cliff is a PhD student in the Astronomy department at the University of Washington in Seattle. His thesis work focuses on understanding the formation and evolution of star clusters in Andromeda. In addition to playing with Hubble data, Cliff cheers for the Seattle Sounders and tries to go sailing whenever he can break free from work.

Bill Keel

Bill Keel is a University of Alabama astronomer, Galaxy Zoo team member, and all-round galaxy aficionado. He appreciates viewing the sky so much that he's also an amateur as well, having found all the Messier objects without computer assistance.

Anil Seth

Anil is the leader of the star cluster team of the PHAT project, and is a professor at in the Physics & Astronomy department at the University of Utah. He likes to drag his 2.5 year old daughter on hiking and cross-country skiing trips, and is currently busy learning to be an amateur astronomer, not just a professional one.

Evan Skillman

Evan is a professor of astronomy at the University of Minnesota. He is currently leading a team to bring the PHAT project to a planetarium near you. Although he is an avid fan of dwarf galaxies, he will work on larger galaxies (if they are close enough).

Development Team

The Andromeda Project was built by the Zooniverse, which has web teams in Chicago (USA) and Oxford (UK). Support for the development of the Andromeda Project comes from the Alfred P. Sloan Foundation and the European Research Council's GLORIA Project.

Brian Carstensen

Brian is a web developer working on the Zooniverse family of projects at the Adler Planearium. Brian has a degree in graphic design from Columbia College in Chicago, and worked in that field for a number of years before finding a niche in web development.

Amit Kapadia

Amit joined the Zooniverse team at the Adler Planetarium in September of 2011. He comes from a background of physics and mathematics. He has worked for various astronomy outreach groups including three of NASA's Great Observatories.

Chris Lintott

Chris Lintott leads the Zooniverse team, and in his copious spare time is a researcher at the University of Oxford specialising in galaxy formation and evolution. A keen popularizer of science, he is best known as co-presenter of the BBC's long running Sky at Night program. He's currently drinking a lot of sherry.

David Miller

As a visual communicator, David is passionate about tellings stories through clear, clean, and effective design. Before joining the Zooniverse team as Visual Designer, David worked for The Raindance Film Festival, the News 21 Initiative's Apart From War, Syracuse Magazine, and as a freelance designer for his small business, Miller Visual. David is a graduate of the S.I. Newhouse School of Public Communications at Syracuse University, where he studied Visual & Interactive Communications.

Michael Parrish

Michael has a degree in Computer Science and has been working with The Zooniverse for the past three years as a Software Developer. Aside from web development; new technologies, science, AI, reptiles, and coffee tend to occupy his attention.

Robert Simpson

Rob is a researcher and web developer at the University of Oxford where he works as part of the Zooniverse team. He is PI of the Andromeda Project's sister site: the Milky Way Project, which is looking at another very nearby galaxy (ours!).

Arfon Smith

As an undergraduate, Arfon studied Chemistry at the University of Sheffield before completing his Ph.D. in Astrochemistry at The University of Nottingham in 2006. He worked as a senior developer at the Wellcome Trust Sanger Institute (Human Genome Project) in Cambridge before joining the Galaxy Zoo team in Oxford. Over the past 3 years he has been responsible for leading the development of a platform for citizen science called Zooniverse. In August of 2011 he took up the position of Director of Citizen Science at the Adler Planetarium where he continues to lead the software and infrastructure development for the Zooniverse.

Julia Wilkinson

Jules has a background in managing and working with volunteers and is happy to be a volunteer herself now having been involved with the Zooniverse since its beginnings. She also moderates Moon Zoo and Solar Stormwatch forums and has a soft spot for a good star cluster. When not moderating she can often be found waiting for a gap in the clouds or walking the dog.

Guide

The primary purpose of the Andromeda Project is to identify thousands of star clusters in the Andromeda Galaxy, using images taken with the Hubble Space Telescope as part of the Panchromatic Hubble Andromeda Treasury project. Click on the links above to learn:

  • What star clusters look like.
  • How to identify distant galaxies seen through the disk of Andromeda.
  • How to flag unusual artifacts in the images.

You can mouse over any of the images in the Guide to see them in black and white.

If you'd like to learn more about the science we are hoping to tackle with this project, click on the "About" or "Blog" tabs at the top of the page.

Star Clusters →

Star clusters

Identifying star clusters:

Star clusters are the most compact groupings of stars visible in the PHAT images. Star clusters have an amazing range of properties, ranging from scrawny newborn clusters to old massive globular clusters that are among the oldest objects known in the Universe. In this page, you will learn how to identify star clusters.

The Images

There are two types of images that can be used to find star clusters. The first are color images, created by combining two separate exposures: one taken through a blue filter (centered at 475 nm) and the other taken through a red filter (centered at 814 nm). The second type are black and white images, available by clicking the “B/W” button, which show an inverted version of the blue filter image. In this guide, scroll over any color image to see its B/W version. These single filter images are sometimes useful for picking out very faint clusters or to separate stars in possible clusters.

What do star clusters look like?

All star clusters are dense groups of stars. However, the exact appearance of a star cluster depends on five factors: (1) mass; (2) angular size; (3) age; (4) dust extinction; and (5) local environment. Let's take a look how each of these influence a star cluster's appearance.

Left: A massive cluster, containing ~10,000 times the mass of the Sun. Right: A low mass cluster, which is 10x less massive than the cluster on the left.

1. Mass

Stellar clusters have a wide range in mass. The most massive clusters have millions of stars and are easy to find. The least massive clusters only have a few hundred stars, with just a few of these being bright enough to appear in our images. These clusters are therefore difficult to spot.

Three clusters with similar masses but different sizes. The most extended cluster is the most difficult to find.

2. Angular size

The typical star cluster is about 20 light years across, which in our images is about 30 pixels. However Andromeda contains clusters whose sizes differ by more than a factor of ten. If two clusters contain the same number of stars, the cluster with a larger diameter is harder to find, because its light is spread out over a larger area.

Clusters are typically quite compact and will not cover most of an image; for examples of the size scale of objects you should be identifying as clusters, look at the pictures on this page and in the examples section.

Stellar clusters with different ages, ranging from a few million years on the left, to a hundred million years in the center, to a few billion years on the right.

3. Age

In a very young cluster, the brightest stars will be massive blue stars. As time passes, these massive stars use up all their fuel and die, leaving only the lower mass fainter stars. Eventually, even these lower mass stars begin to die and evolve into red giants. Old star clusters therefore appear fainter and redder than young clusters with a similar mass.


Left: one of the most highly reddened clusters found so far in M31. Right: a close pair of clusters, where one is reddened by dust and the other is not.

4. Amount of dust

Clouds of cool gas and dust are frequently found in galaxies like Andromeda or the Milky Way. This dust can block some of the light from a cluster, particularly short wavelength (bluer) light, making the cluster appear redder.

Clusters found in different galactic environments. From left to right, the clusters are found in the central galactic bulge, midway out in the star forming disk, and in the outskirts of the galaxy.

5. Environment

The PHAT images cover the disk of Andromeda from its busy center to its tenuous outskirts, spanning a distance of more than 50,000 light years. In the inner regions, the density of stars is extremely high, while in the outskirts the distribution of stars is much more diffuse. Exactly how easy it is to identify a cluster will depend on the density and colors of the normal "field" stars, compared to the density and colors of stars within the cluster.

Tips for identifying star clusters


Examples of fuzz around a large star cluster.

Look for the "fuzz" in the clusters

Star clusters contain stars of many different brightnesses. When you look at an image, you can typically only see the brightest of these stars. Thus, even a star cluster that has only 3-4 bright stars can have hundreds of fainter ones that individually go unseen, but which blend together to form a component of diffuse unresolved light. Because of this, one reliable signature of a star cluster is an enhancement in the background light, giving them a “fuzzy” appearance. This fuzz may be less apparent in the more extended diffuse clusters, where more of the faint stars can be detected individually.


Examples of single bright stars.

Avoid single bright stars

A bright single star can sometimes be mistaken for a star cluster, particularly if it is surrounded by a 'halo' of light scattered by the telescope optics. The brightest of these can often be recognized by a small cross feature crossing through them, also caused by the telescope optics. In contrast to these single bright stars, a true star cluster will be tightly grouped with at least three or four individually visible stars.

Other Cluster Notes

Example of an artifical cluster (left) next to a real cluster (right).

Known clusters

Some of the clusters that you identify may have been found previously. There are a number of catalogs of star clusters that have been created from lower resolution ground-based observations. These catalogs include many of the brightest clusters, but miss most of the lower mass clusters that you can identify. We know this because in our initial work on this dataset (Johnson+ 2012), our team found up to four times as many clusters than had been previously cataloged. Most of the clusters you identify will never have been identified before.

Synthetic clusters

In addition to real clusters, we have inserted some highly-realistic synthetic clusters into the images. We use these objects to understand which kinds of clusters can be detected in the images, and which will be overlooked. Knowledge of these limits will be essential to deriving how Andromeda's system of stellar clusters evolves over time.


An example of an HII region, which can be identified by the blue nebulosity around the stars.

HII Regions

Some of the youngest stars reach high enough temperatures that they emit very high energy ultraviolet light. This light can ionize the surrounding gas, which then glows at a number of distinct wavelengths. The glow from these "emission lines" is particularly strong in our blue image, leading to a diffuse blue nebula (see below). This nebula is known as an "HII Region" (pronounced "H-two"), because it contains ionized Hydrogen. If you see tight groups of stars associated with these blue nebulae, you should mark them as clusters. However, not all HII regions contain or are associated with clusters, as is the case for the example presented here.

Background Galaxies →

Background galaxies

Distant galaxies that lie behind Andromeda’s disk are often seen in the PHAT images. For those who have participated in Galaxy Zoo, these galaxies will look familiar. There are disky, spiral galaxies (like Andromeda) as well as rounder, smoother elliptical galaxies. We want you to identify these background galaxies so that we can study dust on small scales in Andromeda.

On the left you can see some examples of background galaxies. The leftmost image shows a dramatic pair seen edge-on and face-on. Also shown are six typical background galaxies, showing a range of sizes, luminosities, and shapes.

Note that some of the smaller, rounder galaxies can be easily confused with stellar clusters, particularly in crowded fields with many stars. In general, background galaxies will tend to be redder, smoother, and less round than typical stellar clusters. Background galaxies will also have a larger fraction of diffuse light.

Artifacts →

Artifacts

Image artifacts you should not mark:

The PHAT images are beautiful, but they also have some imperfections. These fall into two categories: (1) the imperfections we don’t know about and that we could use your help in finding, and (2) the ones we do know about, but that you should be aware of.


Examples of chip-gaps, caused by the small space between the two CCDs on the Advanced Camera for Surveys. Within the gaps, you will sometimes see black stripes, individual bright pixels due to cosmic rays, and stars with unusual colors or morphologies.

1. Chip-Gaps

The images for this project were taken using HST’s Advanced Camera for Surveys (ACS) instrument. This camera uses two separate CCD chips that have a small gap separating them. We fill in this gap using overlapping images, but sometimes these overlapping data do not exist, or the overlapping data are imperfectly aligned. This causes diagonal streaks in the image, but we are aware of their locations and therefore they should simply be ignored. In this image there are examples of chip gaps.


An example of a field with an image edge. The blank, black portions of the image are regions of sky that were not imaged by the Hubble Space Telescope

2. Edge Fields

Our current round of data includes a lot of fields near the edges of our data. This means that you'll see images where a big portion of the image is blank, like the one to the left. This is not because the galaxy ends beyond that point, but just because our images don't cover that part of the galaxy. But we want to make the most of the images we do have and so we're trying to make sure that no part of our data goes unsearched for clusters, even the raggedy bits at the edges.


Examples of cosmic ray artifacts. These streaks occur when high-energy particles interact with the camera’s detector.

3. Cosmic Rays

Most of the images you will see were created by combining multiple exposures of the same area of the sky. This helps us see fainter objects, and also allows us to get rid of cosmic rays. Cosmic rays are bright dots or streaks that are created when an energetic particle (mostly protons) fly through the detector. These cosmic rays are abundant above the Earth's surface, and individual Hubble images are filled with them. In the current round, some of our data only has one exposure in each filter, and therefore many streaks from cosmic rays, like that shown to the left, remain. These look like either blue or red in the color images depending on whether they hit the telescope during the observations through the blue filter or red filter. Please ignore these and do your best to find any clusters that are in the field!


An example of CTE "streaks". Problems on the CCD lead to "bleeding". These features are most evident in images with smaller numbers of stars, and become weaker farther away from the chip gap.

4. Charge Transfer Efficiency (CTE) Imperfections

The cameras used to take the images are charge-coupled devices or CCDs. When a photon hits a pixel in these cameras, it creates a free electron. These electrons are stored in the pixel until the exposure is finished, and then the charge from each pixel is transferred to its neighboring pixels, and finally out to one corner of the chip where the electrons are converted into a digital signal. This process is typically very efficient, moving >99.999% of the electrons correctly. But as the CCDs age, the pixels wear out a little, causing a little inefficiency in the way that charge is transferred. This inefficiency shows itself as small streaks in the image; some examples are shown in this image. This is especially noticeable in the least dense parts of our images, and particularly in the black and white images.

Image artifacts you should mark:

There are a number of artifacts that we would like your help in finding. These are features whose presence cannot be predicted, but that can corrupt our measurements if we do not know about them.

Examples of saturated stars.

1. "Diffraction Spikes" Around Bright Stars (common)

Very bright stars cause an artificial "cross" shaped pattern that is caused by the support structure of Hubble’s secondary mirror. When we try to identify individual stars in Andromeda, we mistake these bright crosses (known as "diffraction spikes") as a line of stars. You can help us improve our stellar catalogs by marking the full extent of the crosses with the "Cross" tool.

Two examples of linear features.

2. Satellite/Asteroid Trails, & Other Linear Artifacts (somewhat rare)

There are a number of artifacts that can show up as linear artifacts in our images: (1) Satellites and asteroids that pass through the line-of-sight of the telescope during our exposures are seen as lines in our image that appear in just one color; (2) an artifact known as “dragon’s breath” occurs when a bright star lands on the edge of the image, scattering light into the frame; (3) diffraction spikes can be large enough that they show up in images adjacent to the one containing the bright star that produced them. Please mark any of these features using the "Linear" tool.

An example of an "eyeglass" ghost caused by a internal telescope reflections of a bright star elsewhere in HST's field of view.

3. Ghost Images (rare)

In addition to diffraction spikes, bright stars can also produce additional reflections within the telescope or camera, causing out of focus “ghost” images. These images often look like a ring of light or eyeglasses. Please identify these using the Ghost tool. You also find ghosts-of-ghosts, as you see in this image. These should be marked as well.

Examples →

Examples

Like many Zooniverse projects, this site uses the answers from multiple volunteers to decide what objects are present in any one image. A single wrong answer will not cause the science team a problem because several other people will look at each image as well. That is not to suggest that we don't want everyone to try their best to mark images accurately - we do!

To help resolve confusing cases, and to give you a better feel for what to expect on this site, here you can find many examples of images classified by the science team.

















Classify →