by Doctor Science
Greetings, fellow Citizens of the Republic of Science!
Galaxy Zoo began more than 3 years ago. It's a crowdsource project to have regular people -- citizen scientists, as they say -- do image-processing for large-scale projects. I worked with them for a while -- especially on what are now called pea galaxies -- then dropped back as I did more political work and blogging.
Now I'm going through one of my periodic realizations that I'm wasting a lot of time doing mindless games (sudoku, fer chrise sake), and thinking maybe I could use those brain cycles -- FOR SCIENCE!
At this point, Galaxy Zoo has expanded to a zooniverse of projects, some by the original team and some using the Galaxy Zoo API. This is what I think of the current projects:
- At Galaxy Zoo itself you can help classify galaxy images from the Hubble Space Telescope. I found the tutorial video a bit confusing, but the actual classification is easy and fairly mindless. There are pretty colors, and you typically get to look at things like a spiral galaxy 2.5 billion light years away:
- Galaxy Zoo: Supernova is nearly up-to-the-minute science, sorting through possible supernovae within hours of detection so that the astronomers can get a jump on them: nothing is more valuable than data from a supernova that hasn't yet reached its peak brightness. It's easy enough and quite scientifically useful, but I find it too boring to really do much.
- Galaxy Zoo: Mergers is about computer simulations to model galaxy collisions. There are two things you can do: rate simulation results for their similarity to a given real-world target, and grow your own. My appreciation for how *hard* it is to come up with a reasonable similuations grows. However, I don't find this interesting enough to be really relaxing, if you know what I mean.
- Moon Zoo is my current favorite. There are two things to do: Boulder Wars, which is basically a bubble-sort for "boulder" (=loose surface debris) density; and Crater Survey, where you mark craters above a certain size, and also stick notes on other features of interest.
Obviously, this is not the right choice if you're looking for *colors*, but otherwise it's quite seriously fascinating. All the images you've looked at recently can be seen in "My Moon Zoo", on a map so you can see where you've been working:
The images you work on aren't scattered randomly, they tend to focus on areas that have some particular scientific interest. Many of the ones I've worked on are near Aristarchus:
which turns out to be a geologically (selenologically?) unusual and significant region, with a lot of peculiar topography. But even without that, I love knowing I'm looking at another *planet*, getting to know it in such detail.
Maybe you-all can help me with one thing. Sometimes when I look at an image of craters my eyes say "yep, those are indented". Other times, depending on the angle of the light, my eyes tell me "those are outdented bumps":
Turning my head sideways doesn't seem to help. Is there any way to re-boot my visual system, so I see craters as craters?
- We've all encountered plenty of references to ships' logbooks in fiction and history -- they've been kept as standard, daily records for hundreds of years. The UK Royal Navy, notably, has thousands upon thousands of old logbooks in its archives, very few of which have been examined since they were deposited. Old Weather is a project to transcribe hand-written logbook entries from Royal Navy vessels of the WWI period, building a database of historic weather data from all over the world's oceans. You "sign on" to a ship as a cadet, and work your way up the chain of command as you transcribe entries (and learn to read particular people's handwriting). For me, this is not as relaxing as some of the astronomical projects, not least because it engages my imagination: what were they seeing? how did they feel? It's very *human*, shadowing someone from a century ago -- while gathering data for climate models that may predict weather a century from now.
- I've looked at Solar Storm Watch, but haven't been able to develop a search image to use on the video data. This is another real-time project, trying to spot solar storms as early as possible, but I find the tutorial opaque. Maybe you can do better than me.