3.57am. Just had an unnecessarily long session in the computing lab, deriving spectra from the raw observational data we gathered. A discussion followed in the cafeteria, where we went back to our aims and objectives, and quickly went through whether we felt we'd achieved them. For all the glitches and issues we'd had along the way, I felt I had to stand up for what we'd achieved, so where things hadn't necessarily gone perfectly, I stood up to be counted, and explained why we'd done things that particular way.
A perfect instance of this happened, and it's something I need to remember should I choose this project to write up. During the observations in the dome, Peter (the nominated Project Manager for this project) arrived, and I asked how the data analysis was going for the other subgroup in the computer lab. Peter said that the pixel counts for our bias and flat frames had been significantly higher than those taken yesterday. I was sure that we had done nothing wrong and I felt a pang of professional pride, with a temptation to start defending the bias and flat frames that we'd taken.
However, in order to rule out a silly error, we took some new bias and flat frames. These came out with very similar results, and it wasn't until the conversation in the cafeteria that we realised the reason. Sunday night's bias and flat frames were taken with a telescope whose CCD had been cooled to -10C. Tonight's CCD on tonight's telescope had been set up to cool to 'only' -5C, something that we noticed straightaway but hadn't realised the implications.
Those implications are that if the CCD is not cooled as far, more electronic noise will be imaged, and so the pixel counts on those bias and flat frames will be much higher. Having realised this it was not an issue, since these frames are literally subtracted from the results that we get.
Interestingly for all Derek's overbearing habits, it was fairly clear to me that he didn't really know what results to expect on this project. I noticed that every time I wrote something down, he wrote it down too. At one point I wrote down that a star had CNBA classification (short for 'Can Not Be Arsed') in my notes, and he wrote it down too without comment.
One thing was clear, that the spectrum from Saturn was the wrong way around. I realised quickly that this had been my fault, I hadn't realised that it was possible to select a source in the wrong order, but once pointed out, it was obvious.
So, we went back to the computers and spent ages getting it the right way round again. Lots of people kept piping up with 'useful suggestions' as to how we could reverse just one of the sets of data, none of which worked. Too many cooks, etc.
In the end, Roy went back to MaximDL and resampled Saturn's spectrum the right way around. Eventually this produced a new spectrum, and thanks to Margaret (the tutor helping to run this activity for our group), we discovered that one of the spikes was due to the presence of methane. Since the only light from Saturn is reflected light from the Sun, and the Sun does not contain observable amounts of methane (for the purposes of this project at least), we realised that the methane was a signature from the surface of Saturn itself. Amazing to think that with a basic 12" telescope, we'd detected methane on another world. Quite mind-blowing.
People are starting to get fed up tonight, but it has been really useful, in ways I had not expected.
Heading for the coach soon.
No comments:
Post a Comment