I have spent a good portion of the last few days trying to take telescope alignment data.The first night, we had network issues and weren't able to start until about 9PM. The process should start at 7, when the sky is still a bit bright, so that you can see the different mirrors of the telescope properly. This acts as your template. Once the sky gets dark, however, it makes it alot harder to get a good template. In prinicipal you can use the moon to light up the mirror facets, but the moon was too low the first night to do so. Last night, the weather was pretty bad, cloudy and such, so we didn't even bother setting up the alignment tool, and finally, tonight, I got lucky. We only managed to get a template image once it was fairly dark, but it should still be fine.

The way the alignment works is really interesting. You begin by pointing the telescope at a star. In prinicple, all the mirror facets should be lit up if you are at the focal point of the telescope and looking at the dish. However, some mirrors will be misaligned and they won't be lit up. If you move the telescope up, down, left or right with respect to the target star, eventually you can find a position so that a misaligned mirror facet will be lit up. Then, the angle difference between the star and your new pointing direction is related to the angle that the mirror is misaligned by simple geometry. 

A coworker of mine, Andrew McCann, who developed this method, has published a paper. The arXiv print can be found here. If anything, the figures are interesting.

I had to change dorms this morning. I have relocated to one of the sub-summits of Mount Hopkins, to the location of an automated telescope array called mEarth. I have the dorm to myself, which is both a blessing and a curse. It does get mighty creepy staying up on the Ridge at night, and the bed in the mEarth building is pretty tiny. 

mEarth has 8 16" telescopes. Note that this image is not my own; I pulled it from  rcopticalsystems.com