July 12, 2017  

#GZoo10 Day 3

We’re ready for the final day of the Galaxy Zoo 10 workshop at St Catherine’s College in Oxford; it’s been great to have so many people following along on the Livestream – yesterday’s talks are still up, and today’s schedule is:

09:30 Alice Sheppard (Forum Moderator 2007-2012)
10:00 Brooke Simmons (UCSD)
10:20 Nic Bonne (Portsmouth)
10:40 Coffee
11:00 Coleman Krawczyk (Portsmouth)
11:20 Mike Walmsley (Edinburgh)
11:40 Carie Cardamone (Wheelock)
12:00 Karen Masters: Summary (Portsmouth)

We’ll be blogging these talks as they happen here but you can also keep an eye on the twitter hashtag for updates too!

Our first speaker this morning is Alice Sheppard who was here from the beginning as a forum moderator on the original Galaxy Zoo site. She’s talking about the past 10 years and how she got involved with Galaxy Zoo site. She was very keen to get involved in the project and help classify galaxy images. After finding images that weren’t easily classifiable, users started to email members of the science team to ask them what to do. After this happened many times, the team realised that a place where users could talk together and interact with the team about classifications would be really useful. So the Galaxy Zoo forum was born! Alice was one of the first people to sign up and was asked to moderate the forum – she (along with other moderators) even started welcoming each new user who signed up with a friendly “Welcome to the Zoo!”

Alice is now talking us through some of the findings made by the users. These discoveries including the Green Peas, which when first spotted by the users they immediately started investigating what they were using the links to the science survey site. In the original Galaxy Zoo there was also no button for an irregular galaxy, so users started collating their own collection of irregular galaxies! But what makes users keep coming back to the Galaxy Zoo forum time and time again? One success story was the Object of the Day – the moderators even crowd sourced the users to find good images!

Alice has discussed some suggestions for future engagement with users online – always give people room to chat; whether it’s about astronomy or not at all! Remember: as good as you think you citizen science system, tools, tutorials etc are – the volunteers will teach you how to do it better!

Next up this morning is Brooke Simmons talking about what she’s calling probative outliers. The things that tend to break the mould and challenge our world (or Universe!) view. She starts with the bulgeless galaxies – those that look like pure disks – but are hosting growing super massive black holes in their centers. This is weird because the most accepted theory is that super massive black holes grow in mergers of galaxies BUT mergers also grow bulges – so how did these bulgeless things grow their black holes? Brooke is showing us some beautiful follow up observations of Galaxy Zoo SDSS images taken with the Hubble Space Telescope that will help to try and figure this out.

Thing is, Brooke only has about 100 of these galaxies – but not for lack of trying! They just seem to be really rare. If we could actually cover the southern sky in the same way that SDSS covered the northern sky, Brooke would be very grateful! Could we also use trained machines to pick out the weird outliers as well? In which case, Brooke thinks we need to adapt the next iteration of Galaxy Zoo to be both machine as well as user friendly. This will mean leaving things behind but let’s not be afraid to make changes!

Next up, we have Nic Bonne from Portsmouth who’s going to tell us all about making luminosity functions using the data from Galaxy Zoo. So what’s a luminosity function? It’s basically a count of the number of galaxies at different luminosities (or brightness) which can give us clues about how the Universe formed and evolved.

Luminosity functions are especially interesting if you start making them for different morphologies or colours of galaxies. This is what Nic has been doing using the Galaxy Zoo 2 classifications. He’s found something a bit weird though – that the galaxies classified as smooth seem to be more numerous than those classified as featured at the low luminosity (low mass) end. Bringing the colour of the galaxies into this picture as well shows you how similar red featured and red smooth luminosity functions are. Nic says there’s a lot more to do be done with this work though, including using Ross’s new debiased classifications to improve the sample completeness and investigating how the luminosity functions change shape for different kinematic morphologies.

Next up is Coleman Krawczyk who’ll be showing us some of the initial results from the Galaxy Zoo 3D project. This project asked users to draw around the features of a galaxy on an image so that researchers could pick out the spectrum of that particular feature using MaNGA data. Users were asked to either mark the centre of the galaxy or draw around any bars or spiral arm features.

Users could also choose which classification task they would prefer to help out with. This meant that the easy classification task of marking the centre of the galaxy was finished within a couple of days – whereas the spiral drawing task took 6 weeks for classifications to finish. Coleman has now reduced these classifications and has made “maps” for every galaxy marking which pixels are in which features. He’s now started making diagnostic plots to map the star formation rate in the different features of classified galaxies. Turns out we’re going to need more classifications in order to do the science we want to, so this project could have new data coming soon!

Next up is Mike Walmsley, who’ll be joining the research team as a PhD student in October.

He’s not yet done any work with Galaxy Zoo but as part of his Masters research he looked at doing automatic classification of tidal features in galaxies. His goal was to write a code that trained a machine, using a neural network, to detect these tidal features. He also figured out that masking the main galaxy light in the image makes it easier for the machine to spot tidal features. So does this method actually work? It identifies tidal features with ~80% accuracy – which is actually a much higher quality than other automated methods! He’s hoping to apply these methods to new and bigger surveys during his PhD.

Next up is Carie Cardamone from Wheelock College talking about her work building on the the discovery of the Green Peas by Galaxy Zoo volunteers.

So what are the Green Peas? First up their name describes them pretty well because they’re small, round and green. Carie originally wanted to study them because she thought they might be growing super massive black holes, but it turned out instead they have extremely high star formation rate for their relatively small mass. There has been many further studies on these objects so we now know a lot more about them, but one thing we still don’t really know is what galaxy environment they live in. Carie is trying to quantify this but the first problem was that she didn’t have enough Peas! There’s only 80 in the original GZ2 sample but now with the better analysis tools Carie has been able to select 479 candidate Peas. Analysing this sample and comparing it to a sample of well studied luminous red galaxies, the results suggest that peas are less clustered. i.e. the Peas have fewer galaxy neighbours.

We’re now coming towards the end of the meeting (sad times guys) and to remind us all why we’re here and what we talked about, Karen Masters is going to give us a summary of the past couple of days. She’s first pointing out how great we are as a team and the impact the research has had on the galaxy evolution community. Fitting for the 10 year anniversary is that we have 10 published papers with over 100 citations!

Karen has noticed a couple of themes from the past few days that she’s summarised for us. The first is that we have to keep engaging with the Galaxy Zoo community on Talk. The second is that we shouldn’t be afraid of change – let’s not get hung up on how it’s always been done and think about how best to do it now. The third is that galaxy’s are messy and we need to think carefully how we use the classifications. The fourth is that the users will always give you what you ask for – so be careful what you ask! But sometimes you get more than you asked for and end up with a wonderfully collaborative research team!

3 responses to “#GZoo10 Day 3”

  1. Jean Tate says :

    Could you say a few words about blueberries and red peas?

    Reply
    • ccardamone says :

      Hi Jean!
      I think maybe asking about the “Red Nugget’s” that astronomers found in the Sloan Sky Survey. They are compact but red galaxies. The galaxies, like peas, are also analogous to high redshift galaxies, however they represent a very different type of galaxy – more massive and likely rather passive (not forming stars). Since they were discovered there have been many follow-up studies There is a great story about them at the CFA – https://www.cfa.harvard.edu/news/2013-24 and one at astrobites (https://astrobites.org/2013/04/14/digging-up-red-nuggets-in-local-elliptical-galaxies/).

      There was a recent article about Blueberry galaxies (Yang et al 2017), these are much more similar to the peas being extreme emission line galaxies, analogous to high redshift star forming galaxies. The study searched for lower mass, closer galaxies (so they didn’t appear green in the Sloan images as the peas did because OIII wasn’t bright in the SDSS ‘r’ band).

      Hope this answers your question. If you have more – feel free to reach out over the green peas thread on talk!

      Carie

      Reply
  2. Christine Macmillan says :

    Nice to see you all, and thank you for the Livestreaming of the meeting.

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

Gravatar
WordPress.com Logo

You are commenting using your WordPress.com account. ( Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. ( Log Out /  Change )

Cancel

Connecting to %s

%d bloggers like this: