Observations of an Exoplanet!

Last Wednesday night we lucked out with clear sky conditions and relatively good seeing to observe something amazing: the transit of an exoplanet!

HD80606b is a planet that is four times larger than Jupiter, orbiting one star of a binary star system 190 light-years away from earth. Every 111 days, this planet passes in front of its star as viewed from earth.

a binary star system in Ursa Major

approximate location in the sky for this exoplanet system

These stars are not visible to the un-aided eye at 9th magnitude. This image was combined from 15 second exposures.

The dip in the measured light from an exoplanet transit is very small – about 0.011 magnitudes in this case. Could we possibly measure anything to the accuracy needed to recognize it? A challenge to accept!

First, this binary system presented a special advantage for the measurement. The planet transits across HD80606 (magnitude 8.93). The binary star companion HD80607 (magnitude 9.08) is only 23 arc-seconds away and is of constant brightness. It should be possible to compare the brightness of the two stars. Atmospheric effects, seeing conditions and other sources of error would likely affect both stars in the same way. This measurement error would cancel out.

A separation of 23 arc-seconds also presents a problem. Accurate measurements are required for BOTH of the stars. We need to have a clear separation (no light) between them. How do we prevent the light from each star from spilling over into each other?

First, the image scale on the CCD detector had to be larger. We removed the telecompressor from the telescope setup to narrow the field of view. The full 3000mm focal length of the LX200 was used.

Second, the focus needed to be tweaked as much as possible. This took a long time!

Lastly, the atmosphere had to be stable. On one very cold prep night those two stars were dancing all over the place! But last Wednesday night, between two pressure systems, the air was steady and clear!

A few other techniques were planned as well. For measurements like this, the “pros” use monochrome CCD cameras with a band pass filter. Monochrome cameras have more sensitivity and controlled light spectral response. Our Starlight express camera is a “one-shot color” camera with a color CCD device. I worried about how the variation of the responses from the red, green and blue pixels would play out with variations of the starlight placement on the pixels and variations in color focus with the “seeing”. So I used a Baader 500nM green band pass filter on the CCD camera. With only green light of significance, I would extract and use only the green color planes from the raw image data.
Image calibrations were carefully done with 30 bias, 30 dark and 30 flat field images.

There are two separate green pixels in each “Bayer matrix”. Each would have its own random thermal noise. Averaging these two pixels would reduce the noise at each “composite” pixel.

Noise can be treated statistically if it is actually random. I needed lots of images if measurement averaging techniques were to be used. The camera took a 15 second exposure about twice a minute for over five hours! 525 images were captured.

MaximDL 5.08 software was used to crunch the images. Batch processing helped with some of the work, but there still was about 6 hours of mouse exercise to combine the two green planes and perform the photometry. The photometry data is exported into Excel where it can be plotted and time series averaged.

So, now that you’ve read down this far you are wondering if it all worked. I think it did!

HD80606b transit (2nd half) on night of January 13th-14th (Harken Observatory, Pewaukee WI)

The observations started at about mid-transit and ended at 2:10 AM. I could not stay awake for another hour or so to be sure that the transit was complete (I had to work in a few hours and needed some sleep!) There is about a 0.01 magnitude increase in the brightness of HD80606. This is what was expected.

This is the preliminary light curve plot. Preliminary in that maybe others will have suggestions of how to refine the data reduction process.

I hope to present more about this and other exoplanets (now over 400 known) at the February 13th astronomy talk at the Harken Observatory (Pewaukee Library). Hope to see you there!

Your comments and questions are welcomed!

Clear Skies!

Randy Buchwald

About rbuchwald

Electrical engineer by day, astronomer at night!
This entry was posted in General, Imaging, Stars. Bookmark the permalink.

4 Responses to Observations of an Exoplanet!

  1. twalkowski says:

    Randy,

    First, let me say thank you for such a fantastic effort! It is amazing to realize that you can detect a planet crossing in front of a star 190 light years away!! The change in brightness matched your prediction — even though the change is so small, the light curve clearly shows it.

    Second, are you able to plot the light curve of the other star as a reference? I don’t know how difficult that would be, though.

    Lastly, which star is has the planet — right or left in the image?

    Well Done!

    Tim

    • rbuchwald says:

      Hello all! Sorry for the delayed reply. The planet is going around the star on the left, which is the slightly brighter member of the pair. This was the ideal situation to observe because the right hand star was used as the reference star (it is known to not vary with time and is of the same type-sequence as the measurement object). By definition, the reference star was “constant” at a magnitude 9.1. On Feb 13th I will talk more about the “noise” in the measurements and how statistical analysis is useful in the analysis to this level of accuracy.
      Clear Skies!

  2. mpaquette says:

    Randy,
    That is some amazing work you did! I really didn’t follow all the technical issues you brought out in your analysis, but I got the gist of what you were doing. Besides all the planning for this, you were very persistent and patient throughout the entire process, and I’m so glad you had optimum viewing conditions as a reward for all your preparations. I am looking forward to your upcoming talk on Feb 13th about more of how you did it all. Many, many kudos to you!

  3. twalkowski says:

    I think there is another planet transit candidate for the summer. I’ll have to look it up, but there is one near the Dumbbell Nebula that we should be able to detect. The latest Sky & Telescope magazine carries an article about the Kepler spacecraft and its serach for traniting planets. Worth a look-see!

    Tim