May Astrobite, or “How Is a Pixel Like a Bucket?”

Just a quick post tonight to point out my most recent Astrobites article which came out on the 22nd. This one was very interesting to write. It's about a paper which I read when it first came out back in February on the arXiv. (It's pronounced “archive,” and it's a website where most papers in physics and astronomy and several over sciences are hosted freely available; it's undoubtedly revolutionized the areas it serves by making it easier to communicate results, and I can't imagine trying to do research without it.) I'd been stockpiling recent papers that looked interesting for a week or two before sitting down to write, but none of them really seemed to call to me, till I finally remembered this interesting paper I'd read about CCD systematics.

CCD stands for charge-coupled device, which is the technology behind most digital cameras nowadays. Astronomers adopted them very rapidly back in the 1970s soon after they were invented, and they're responsible for a very wide variety of astronomical research since then. Despite coming up on fifty years old, the authors of the paper I wrote about managed to find a new, never-before-seen form of subtle systematic errors in sixteen out of twenty-two instruments they investigated. The thing that really blew my mind while browsing the abstract and got me to read the paper? They noticed an effect that was proportional to the number of 1's in the binary representation of the value of various pixels in the image.

If you just said “What‽” out loud like I did upon reading that, check out the paper! It's really well written and does a good job of explaining their findings with some really good, high-quality graphs. If you don't know what that means or why it sounds so weird, maybe check out my astrobite—I spent several hours wrestling with an analogy involving grids of buckets and sprinklers in an attempt to render the technical details more approachable, so hopefully I've explained it there in a way that makes sense.

Basically, the top part should be a flat line at zero, not…this.

The results of this paper, while not necessarily highly problematic, are likely to be very far reaching and will affect a lot of people and their science, so now that it's been officially published as of May 11th I expect we'll start seeing some more papers popping up on arXiv related to the issue it reveals. (arXiv allows people to upload “preprints” of papers that have been submitted to journals and are in the process of peer review, which is how I was able to read it back in February.) That's it from me for now though, a hui hou!