The plan was to make another attempt at observing a Kepler planet. I've tried a couple a few before, but for the most part, the Kepler host stars are too dim or the transits to shallow for me to have a reasonable shot. Kepler-10b seemed within reach because 1) this gentleman has done it with somewhat similar equipment and 2) my observations of Hat-P-13, which is similar magnitude, yielded some very respectable results. Even so, I knew going in Kepler-10 with it's very shallow transit depth would be a tough target!
Since I had some time during the day, I also thought this would be a good time to try and piggyback a small refractor to acquire simultaneous data. I'll go into more detail on my setup in a later blog, but basically for this session I had a CPC1100 imaging with a QSI520 and an R filter, an Orion 80st imaging with a QSI583 and a luminance filter and guiding with a lodestar in a Giant Easy Guider. 3 imagers being controlled at the same time! After some fiddling to align the Orion to roughly the same field of view as the CPC, everything worked rather well! I'm not sure how often I'll do this though...the Orion has quite a bit of coma and vignetting with the QSI583--but still plenty of usable area, so maybe...
The Data
The CPC/QSI520 acquired data at a cadence of 120s, 1x1 binning and cooled to -20C. The Orion/QSI583 ran at 60s, 1x1 binning cooled to -20C. In retrospect, I should have ran them both at 120s. I started data acquisition when Kepler-10 was only 16 degrees above the horizon. I'm generally pretty lazy and go to bed as soon as everything is up and running. With the start of acquisition that low in the sky, I was worried about how much brightening would happen as it rose. I've lost runs int he past where the target star brightened into saturation during the course of the night!
The Results
I'm sorry to say I did not observe a transit.
Yeah, looks pretty ugly. It's possible a better selection of aperture and annulus would yield better results, but probably not the factor of 3 or so they'd need to get up there with my best--and certainly not the almost order of magnitude they'd need to match this!
But there's still some interesting things to talk about in this data run...
First, there is a general improvement in the scatter from start through to finish (well, if one ignores the bit in the middle which we'll discuss in a moment). Early on, there's a lot of scatter int he data. Near the end (which is roughly 7 hours later, by the way) the magnitude variances are looking more compact. This is primarily because the initial data was at 16 degrees above the horizon while the final data was closer to 80 degrees. This is a nice little demonstration of the effects of atmosphere! Low down on the horizon, the noise is much higher--scintillation, light pollution, and atmospheric extinction. I'd guess extinction and light pollution dominate scintillation at my location but I could only guess which of those two contributes more.
Second, if we look at some raw data, we can see some interesting effects as well.
This is a plot of "magnitudes". The magnitudes are non-normalizad measures of flux plotted on the same scale as magnitudes (each "magnitude" is 2.5x the previous "magnitude"). The data can be normalized to absolute magnitudes, but all we're interested in here are relative intensities and did I mention I'm lazy?
So what's interesting about this plot is that magnitudes change all over the place! In the first third of the evening, there's a general downward trend in magnitude as the star gets brighter as it moves higher in the sky and the light gets to travel through less air mass. But then what happens? The stars get dimmer again? That's most likely passing clouds. If one were outside looking up, maybe there would be a some nice halos around brighter objects, but the clouds weren't thick enough to disrupt the auto-guider--or if they did--not for very long. This is the reason for the large scatter just after predicted mid-transit in the first plot above.
What's interesting is the CPC has a narrow field of view. Any clouds passing through the field tend to cause nearly uniform extinction across the entire field. Certainly some of the increase in scatter is due to passing clouds dimming the comp and target stars differently at different times, but the bulk of the scatter likely arises from reduced signal to noise. I say that because this transit of CoRoT-1 showed a very similar light curve for individual stars where passing clouds caused large changes in observed magnitudes and yet, the scatter across the entire data set is pretty uniform. Looking at the CoRoT-1 transit plot and model fit, one would be hard pressed to identify when the passing clouds happened.
I'll try and remember to blog about it more in the future, but this shows the power of differential photometry. With differential photometry, the goal is to measure differences between two sources in the same frame. Since the sources are int he same frame, there is a much reduced chance of differences in the way the sources were measured.
Maybe Next Time??
So, no positive transit observation this time. Still not a complete loss--in fact, the results were very much what I expected going in--Kepler-10b is likely out of my current reach. I do now know I can piggyback the refractor (that data is interesting as well, but out of time here) and control everything at the same time. That's a good thing.
Now I'll have to think a bit more about how to improve for next time!
