"Backyard astronomers" looking for transiting extrasolar planets

[kukulcan]

I have been interested in extrasolar planets ever since 51 Pegasi was discovered when I was almost a teenager. I have always wondered if it would be possible for me to find any myself using a decent sized telescope and a CCD camera. Well, it looks like some people are doing just that as found in this article. http://www.space.com/scienceastronomy/exoplanet_transit_040824.html I am having a problem finding out what kind of telescope would be best. Obviosly one with auto tracking abilities is a must, so I am narrowing my decision down...most likely to the Meade LX90GPS model. Can someone please give any help or advice on if I am going in the right direction for extrasolar research? Also, what software would you recommend for detecting and measuring the light reduction for a transiting object in front of its star? Thanks!

[Jim]

Welcome to BAUT, kukulcan!

I think this forum will get you the answers you want. Thread moved.

[RickJ]

What you are considering is very difficult. You will need to start simple and hone your skills on other photometry before you go after planets. You'll need to start with variable stars. To see a planet transit you are looking for a very tiny change in intensity. Lots of statistical work will be needed to coax this out of the noise.

Basically you are measuring the intensity of one star as compared to several comparison stars. It is the difference that is important here. In a perfect world you'd see the same difference if the star is constant. Unfortunately we have to look through our atmosphere. This adds noise. All CCD's add noise both during the exposure (most of which can be subtracted out with very careful use of dark frames) and when reading the data -- nothing can be done about this. So you need to get a CCD with as low of readout noise as you can afford. But it must be matched to the scope. Under sampling reduces accuracy as the star image is small and much can be lost to the area not sensitive to light. As the star moves it will vary because more or less of it is hitting insensitive areas. Over sampling isn't as bad but does increase read out noise as you are reading out more pixels. I find an star 2.5 pixels across is about optimum. Consider your seeing when calculating the needed scope CCD combination. Also be sure it is a NABG camera. If it has a blooming gate up to 25% of each pixel is blocked increasing the dead areas that mean you'd need 3 to 3.5 pixels across a star's disk and deal with the increased read noise. The other problem is the gates are not constant. Each gate reacts differently. That means there's a small difference in when a gate starts to drain charge from its pixel and how fast it does this. So unless the flat was taken at exactly the same intensity as the stars you are measuring the reading will be off. Move the star to another set of pixels and you get a very slightly different reading. One no flat can compensate for. So only use a NABG chip. Also it must be a mono CCD of course. Flats must be very very precise as well. Far more so than needed for good to excellent astro photography. I use 40 to 50 stacked to get the needed accuracy and redo them constantly as chips can change surprisingly rapidly with time. The difference is immaterial for the eye and photo work but it can really screw your photometry up! I use an old ST-7 for my photometry work.

The scope is rather unimportant, even a 4" refractor will be sufficient. Its the match to the CCD and your seeing conditions that is highly important. The mount is highly important. The more the star drifts the less accurate your readings will be. To reduce read noise compared to signal you'll want rather long exposure times putting more stress on the mount. Seeing moves a star so good seeing is important. My seeing is too lousy to ever get my readings down to that level of accuracy. I doubt the LX90 mount will have the needed accuracy without guiding with a second CCD. Even then it may not be sufficient. Serious imagers avoid them and what you are wanting to to is far more demanding on the mount.

One good book, old but good, is sold by Willmann-Bell; ASTRONOMICAL PHOTOMETRY A Text and Handbook for the Advanced Amateur and Professional Astronomer, Henden and Kaitchuck.

As to software, most CCD packages contain the required software. I do some photometry of AGNs for an astronomer at the University of Texas and use the software in CCDSoft as well as that in AIM4Windows version 1.0. Most others also contain what you need to get the raw data. But then you'll need very sophisticated statistical routines to look at this data and find the planet. Since these fellows worked with professional astronomers I assume that was provided by them. I know of nothing off the shelf that would work but if you are an expert on statistical analysis you could probably get a spread sheet to do the work. In my case I just provide the raw data as output by AIM4Windows and he does the rest. While I've been doing this for a couple years now I don't have the required seeing to try for planets so I stick to AGNs for now.
--
Correct domain name is arvig and it is net not com. Prefix is correct. Third character is a zero rather than a capital "Oh".

Rick

[StupendousMan]

To see a planet transit you are looking for a very tiny change in intensity. Lots of statistical work will be needed to coax this out of the noise.

...
As to software, most CCD packages contain the required software. I do some photometry of AGNs for an astronomer at the University of Texas and use the software in CCDSoft as well as that in AIM4Windows version 1.0. Most others also contain what you need to get the raw data. But then you'll need very sophisticated statistical routines to look at this data and find the planet.
...
Rick

Well, it's not quite that bad. Some of the known transiting planets create a dip of almost 1 percent in the light of the host star over a period of several hours. By taking many exposures -- say, one per minute -- and averaging together the individual results over periods of 5 or 10 minutes, one can see the event without TOO much effort.

Many of the standard, commercially available software packages will perform simple aperture photometry well enough for this job. Of course, you must take care to do a good job with the dark subtraction and flatfielding, so it's not trivial.

You can see an example of some really crummy-looking data which clearly shows a transit, taken with a 12-inch Meade SCT, at

http://spiff.rit.edu/richmond/ritobs/sep19_2004/sep19_2004.html

If you want to see a much cleaner example of a transit by the same planet (TrES-1), but taken with a larger telescope at a dark site and better image quality, look at

http://spiff.rit.edu/richmond/sdss/transit/pt_54380/pt_54380.html