Exoplanet transit of TrES-3 attempted tonight

That data looks pretty bloody good! Can't wait to see the final improved versions!

Bart.

Thats an impressive data set guys. How are you accounting for error due to seeing conditions? What sort of confidence limits do you think you will end up with?

Good questions Kieran - unfortunately the answers are pretty long winded but I'll give it a lash:

1. "Seeing" errors are minimised by using as many stable comparison stars as possible. There's an element of trial and error at the moment - I'm not using "standard" stars as such (e.g. Landolt or Henden stars) but with the setup I have (as does Eamonn) the FOV is large enough to give us quite a few decent comparisons. The data above uses 4 comparison stars that were selected on the basis of their stability. So that's seeing errors minimised...

2. As regards confidence limits are concerned, I've no idea really - that's for more detailed analysis. However, all other things being equal, I was getting error bars for magnitude difference between the target and comparisons of the order of 0.007 magnitudes. Which is pretty small. Timing errors are a little more tricky to characterise - times on the chart above are heliocentric julian date - but of course that all depends on the accuracy of the PC clock recording the image, latency between the image being taken and the time-stamp applied to the FITS header etc. I've no idea what those timing errors are at the moment.

Overall, the time of ingress, egress and transit durance compare very well with predicted times - within a few minutes at least on a 2 hour long transit. The curve certainly shows a dip - and the mag dip itself is in reasonable agreement with previous dips (around 30 millimags).

I think there's a lot more (and complex) data analysis required though I personally don;t have the skills or the tools to do that right now. I'm hoping Mike and Eamonn will deal with that

Overall, its pretty cool to see the lightcurve dip of a planet not in this solar system.

Cheers
Dave

By the way Keiren (et al) - you'll notice that the curve itself slopes from left to right - ie its not a nice equal "U" shape - that's because no correction has been made for changing airmass over time. It seems there are those who correct, and those who don't. It matters little however with such a larg (30 millimag) drop. For smaller drops it may be more crucial and corrections should be applied. Of course, if you were to do this to the book, we'd be using Landolt fields, doing corrections (transforms) for each of our systems and a whole bunch more - but that's for the future no doubt.

What is clear though, that for this reasonably large dip, I can certainly produce data with my system thats well into millimag (thousandths of a magnitude) range.

Cheers
Dave

Impressive stuff guys. How was the curve recorded?

Dave et al,

By shear luck, I sat down last night as well to do some planning for exoplanet transit work.

In relation to the above work, great stuff! I wanted to ask you about the curve itself - it is a regression-type curve based on least-squares or is it a different beast (ex. moving average etc)? Also, do you have any reference material surrounding corrections for air mass since this is something I have encountered elsewhere but I have yet to find a source describing its physical computation and correction.

I recently downloaded Bruce Gary's ebook on exoplanet work and I will check and see if he describes the process behind correcting for air mass.

Anyway, it is very exciting seeing such work and results. Keep up the excellent work!