Third Galaxy Zoo paper submitted
Update : Paper now available.
It has taken over six months and a lot of work, but we have finally decided the third Galaxy Zoo paper is ready for submission to a scientific journal, and to be made public. The paper was already in reasonably good shape a month or so ago, but given the number of interesting results it contains, we wanted to be sure the presentation is as clear as we can make it. We therefore had yet another round of internal review by the coauthors, and elicited comments from a couple of other colleagues with links to the project. The response was very helpful and encouraging, so I decided to take a bit more time to improve the analysis further and perform some extra tests. We’ve now got a paper we are all really pleased with. We hope that all of you, and the journal referee, agree with us!
The paper is long, 30 pages in total. Even so, we’ve tried to make the paper as readable as possible by shifting some of the material to appendices (additional sections at the end of the paper). Much of the length of the paper is due to the large number of figures. Again though, we’ve tried to make these easier to absorb by combined multiple plots into single figures and maintaining a consistent style.
The paper has been submitted to the Monthly Notices of the Royal Astronomical Society (MNRAS), one of the world’s principal astronomy journals, and the same publication that the two previous Galaxy Zoo papers (by Kate and Chris) have been submitted to. The paper will now go for peer review. Hopefully we will have the referee’s comments back in a month or so, then there will probably be a few changes to make before it is accepted by the journal and published. As this process takes so long, and this work is so timely, we have decided to make the paper public before it is accepted by MNRAS, so other researchers can see our work as soon as possible. The paper has therefore been added to the astro-ph archive. It will be available for anyone to read from Tuesday 3rd June.
I’m taking a break for a fortnight, but then I’ll write a few blog posts, following on from this one, explaining all the results of the paper in less technical terms.
Update : Paper now available.
Merger hunting one month on
I (Waveney) joined Galaxy Zoo a few months ago and was a relatively quiet soul, who didn’t do very much…
A month ago I was sitting at a computer, and in theory looking at job ads, when I saw Chris’s request – Merger Hunters Needed. I read what was wanted and thought it would be fun to do, but the method requested was awful. One of my part time businesses is running a small scale internet services company and I like writing code to do fancy things. So I offered to provide a web page to do the merger hunting. I was busy during the day, but in about 4 hours in the evening I got the first program working, put it on one of my websites on my server and let the world know via the forum and the blog. I became famous overnight. The next day, Chris wanted to raise awareness and offered to add it to the Zoo front page and send out the newsletter, I requested that the front page be changed, but that the newsletter should wait a couple of days (just in case).
The following day, I was getting a lot of favorable comments and Chris mentioned it in the Blog, and Alice put it on OOTD. A couple of people commented about performance – this initially perplexed me as the program had almost no impact on the server. Then I found a limitation in the SDSS server about repeated requests from the same address, and I realised that this was the cause of the problem. I re-wrote fetches of the data and wrote my first ever SQL to pre-fetch the necessary info from the server. No sign of the front page change.
Usage grew, comments came and I started adding features, those useful features that I wanted like Zoom, Back and direct searches of the forum, requests for Invert, and then grid lines, and and and. I also leant a lot more SQL. Chris requested I have a “Basic” display and an “Advanced” display – done within the hour – still no sign of the front page or newsletter.
Prompted by the speed that one Merger Checker was working, and following discussions with Chris, the system was extended with a lot more data for images with lower merger ratings from the original Zoo. This required a complete re-write of the underlying data management – it took a few hours. Usage grew – still no sign of the front page or newsletter.
I was seen at the Winchester gathering, had it been a month before I would have largely been an unknown – now I was infamous. Usage grew – still no sign of the front page or newsletter.
Usage was steady, but at current rate it was going to take several months to get useful answers, I wrote to Alice requesting it be on OOTD again as it would be months before Chris had the front page changed – what happened – the front page changed almost as I finished sending the message. Users grew, but the rate of image classifications, didn’t change much. Oh well we will have to wait and see. Still no sign of the newsletter.
On Wednesday just after 7pm I noticed a sudden jump in users – 1 or 2 was not usual, but 5 in an hour was unheard of, clicked the status again and it had gone up 5 more! I asked on the forum and somebody said the newsletter was out! The next couple of hours where frenetic as the Merger hunters climbed from 200 to 700, and 20,000 images were seen. I kept an eye on the server – it has handled that sort of load before, and it did this time the only casualty was a logging program that runs in the middle of the night, when the system is usually quiet…
As I write this what started as a quick 4 hours a month ago has become the leading use of my talents for the last few weeks. 1200 users have seen 200,000 images and all sorts of useful results are beginning to be found. Oh and I am still looking for work – anybody want an experienced system/network designer and/or real time programmer?
Galaxy Zoo meeting is over
With the help of coffee and biscuits, we’ve managed to get through the afternoon and the meeting is now over… to be continued in the pub in a few minutes.
Observing round up
If you’ve arrived via the Galaxy Zoo newsletter, you’re probably wondering where the posts about our observing trip are. They’re here:
Bill started things off by writing about our quarry for this Zoo-inspired hunt.
I then got excited about where we were, and about our first images. And about the telescope.
Later in the run we were seeing more and more perfect pairs, all of which will help us track down the galaxies’ dust.
Despite a few problems we pressed on, and started to produce colour images for the first time.
Since we got back, work has continued – and I’ll leave it to Bill to give you the latest.
Join us online tomorrow
The UK based Galaxy Zookeepers are gathering in Oxford tomorrow to discuss, among other things, our plans for Zoo 2, and you can join us (virtually) via UStream. We’ll start around 10 am BST tomorrow, and the stream will be on throughout the day, and you’ll be able to view an archive after the event too.
Another Zoo meetup, and a Zoo celebrity meets some of the team
We’ve all read the story of Hanny’s Voorwerp, and about two weeks ago we were lucky enough here in the UK to have the Voorwerp’s discoverer come and visit us again, meeting some of the team and catching up with some fellow volunteers at another Galaxy Zoo meetup. Read More…
Dude, where's my Mars Polar Lander?
Remember the Mars Polar Lander? It was a mission sent to Mars and land at its South Pole but was lost during the landing. The engineers don’t quite know why it failed and would like to know to avoid similar crashes in the future. To do that, they’d love to see the wreck left behind (if any).
The folks from the HiRISE camera onboard the Mars Reconnaissance Orbiter have put high resolution images of the area on Mars taken by HiRISE where they think MPL crashed on their blog hoping that people will spot the wreck. It seems that human brains once more might beat modern computers when it comes to pattern recognition tasks. Want to give them a hand? Go here to download the images. If you do find it, remember to classify it “Star/Don’t know”.
There’s a nice news story about it in Nature here which also mentions Galaxy Zoo and has some comments from me.
The Story of Galaxy Zoo
Some of you on the forum might remember Devin Powell asking to interview some of you to write the story of Galaxy Zoo. The resulting article is now online here. It’s a wonderful piece:
Amateur Hour
In a modern twist on the centuries-old tradition of the backyard astronomer, thousands of galaxy gazers around the world have plugged in to become keepers of the Galaxy Zoo.
http://www.krieger.jhu.edu/magazine/sp08/f1.html
Enjoy!
Edit: the magazine cover is wonderful too!
http://www.krieger.jhu.edu/magazine/sp08/cover.html
NGC 3314 gets a counterpart
Not that I’m about to change my avatar or anything, but I think this is one of the best pairs we observed from Kitt Peak. First pointed out in the Galaxy Zoo forum by GwydionM, this features a face-on spiral almost exactly in front of an edge-on spiral. Like NGC 3314, it gives a rare chance to see the dust content of a galaxy almost all the way from outer regions to the nucleus (limited there by the accuracy with which we can extrapolate the profile of the background galaxy inward). This also drives home a point which isn’t always obvious from pictures, and is especially insidious when looking at books where the pictures tend to be all about the same size. Similar-looking galaxies can span a wide range of sizes, even among spirals which don’t come as small and faint as spheroidal or irregular galaxies. We don’t yet have good redshifts for both; the Sloan data give the single value z=0.067 for what must be the blended light of both, probably meaning that they are at similar redshifts so the size comparison in this picture is pretty close to reality. This image come from the first fruits of the next stage in processing our data, one which leaves them ready to analyze. To sample red light, we used an I filter rather similar to the SDSS i band. For the particular CCD we used, the skyglow that it sees causes a pattern of interference fringes from light reflecting within the chip. This can be calibrated and subtracted only using data on the night sky itself. We combined images where the target galaxies were at different places on the detector while rejecting objects that were at a certain pixel value only once (that is, things on the sky) to leave, ideally, only the interference pattern. I’m still tweaking until we decide that we’re close enough to that ideal… 
If you look closely, you can spot the heavily reddened core of the background galaxy behind the spiral arm to the lower right of the foreground galaxy core, and note the darker absorption next to those spiral arms. (I’ll be watching to see how this image shows up – this is the first time the regular Zookeepers let me have the keys to the blog, and I’m feeling my way around. It looked sort of odd in preview).
Galaxy Zoo 