My second exoplanet observation

[dokeeffe]

My second exoplanet observation. Much clearer than my first one :-)

Also put a video together of the time of capture

<iframe width="560" height="315" src="https://www.youtube.com/embed/CYZBLVuM4nU" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>

 

[JohnSadlerAstro]

Hi,

I don't want to sound ignorant, (which I am) but, did you observe this with your own kit or by 'renting' a pro observatory? I mean, eclipsing exoplanets would be invisible to anything except the largest of instruments?

John

[Chris]

Superb light curve capture and a matching time lapse! Seriously cool 

20 minutes ago, JohnSadlerAstro said:

Hi,

I don't want to sound ignorant, (which I am) but, did you observe this with your own kit or by 'renting' a pro observatory? I mean, eclipsing exoplanets would be invisible to anything except the largest of instruments?

John

Having taken light curves for variable stars and just the one exo planet HD20958b back in 2004. I'd say it's more about the camera and photometry software package than the scope. I'd wager you could probably detect a transiting exoplanet with you 130pds with the right extras....and maybe me with my Celestron 5SE even. 

Also it helps if you take lots of exposures/data points on the graph, so you can average out the noise to clearly see the light curve.

 

[dokeeffe]

Yea you dont need a big scope at all. I used my C11.

Actually this particular exoplanet was discovered by a 200mm telephoto lens :-)

https://en.wikipedia.org/wiki/XO_Telescope

 

[des anderson]

Wow, just goes to show you don`t need a massive telescope to get results like that, well done. Des

[RobertI]

Very well done and a great timelapse video. Can you say a little more about how you did this? 

[dokeeffe]

Yes I can explain a bit more on how I did it. I guess there are 2 main parts, data collection and data reduction

# Data collection

I use INDI and EKOS to control a small observatory. It has a Celestron CPC1100, Atik 383L+ and filter wheel and OAG. I created a schedule to take 200 60sec images of this star through a photometric V filter (probably should have used a different filter). INDI managed everything, startup, opening roof, shutdown, parking, closing roof. I was not in the observatory when the capture happened. I used this website to decide the target and timing http://var2.astro.cz/ETD/predictions.php?delka=-8&submit=submit&sirka=52

# Data reduction

On shutdown, a shell script gets called to do basic reduction of the collected images (bias,dark,flat correction/calibration). It also plate solves the images and copies to dropbox. This reduction is done using astropy ccdproc. https://github.com/astropy/ccdproc. This stage is really important. The quality of data really depends on this calibration. All the flats, bias, darks are recent and updated regularly. Once I grab the calibrated images from dropbox the manual work starts. I loaded the images into astroimagej and used that to do the analysis. Its pretty easy to do this once you know where to click (will make another video of this process soon). Astroimagej is good for exoplanet analysis but for standard single measurement photometry I use astropy photutils.

The time lapse video was assembled using a python script from the images taken with the all sky camera and the plot from astroimagej. If anyone wants this let me know. Its very much hacked together...

Derek

[JohnSadlerAstro]

11 hours ago, Lockie said:

with the right extras....

Hmmm.... I wonder how much they would cost  

I never realized it was possible to detect exoplanets with ordinary-size telescopes, I'm guessing the only reason why people didn't find them sooner is because they didn't have good digital imaging equipment?

Although the results aren't as eye-catching as a deep-sky image, that is one impressive observation! Are there any websites where you can fin the times for transits, and their affect on the parent star's magnitude?

John

[dokeeffe]

1 minute ago, JohnSadlerAstro said:

I'm guessing the only reason why people didn't find them sooner is because they didn't have good digital imaging equipment?

Yes thats true and also the software to detect a signal in the millions of point sources captured by a survey each night needs to be fairly slick. Its easy to detect a transit when you know the star and time of the transit :-) Its tricky to discover one, like finding a needle in a haystack. 

[Chris]

4 minutes ago, dokeeffe said:

Yes I can explain a bit more on how I did it. I guess there are 2 main parts, data collection and data reduction

# Data collection

I use INDI and EKOS to control a small observatory. It has a Celestron CPC1100, Atik 383L+ and filter wheel and OAG. I created a schedule to take 200 60sec images of this star through a photometric V filter (probably should have used a different filter). INDI managed everything, startup, opening roof, shutdown, parking, closing roof. I was not in the observatory when the capture happened. I used this website to decide the target and timing http://var2.astro.cz/ETD/predictions.php?delka=-8&submit=submit&sirka=52

# Data reduction

On shutdown, a shell script gets called to do basic reduction of the collected images (bias,dark,flat correction/calibration). It also plate solves the images and copies to dropbox. This reduction is done using astropy ccdproc. https://github.com/astropy/ccdproc. This stage is really important. The quality of data really depends on this calibration. All the flats, bias, darks are recent and updated regularly. Once I grab the calibrated images from dropbox the manual work starts. I loaded the images into astroimagej and used that to do the analysis. Its pretty easy to do this once you know where to click (will make another video of this process soon). Astroimagej is good for exoplanet analysis but for standard single measurement photometry I use astropy photutils.

The time lapse video was assembled using a python script from the images taken with the all sky camera and the plot from astroimagej. If anyone wants this let me know. Its very much hacked together...

Derek

Thanks for sharing, I'll be taking a look at some of these links as I've been meaning to get back on the horse with photometry at some point  

[Chris]

7 minutes ago, JohnSadlerAstro said:

Hmmm.... I wonder how much they would cost

Probably as much or as little as one is willing to invest  I bet there are some transits your 1000D could manage, and a lot of software is free. Take as many images as possible just before, during, and after the transit in order to see the light curve through the noise. And as said take darks, bias too. 

If you fancy a warm up, variable stars are easier and fun to capture curves for. I've not done any photometry for 14 years, but I remember really enjoying it, it's quite a buzz   

[JohnSadlerAstro]

That sounds good, I have a science project coming up with my A-level equivalents, and I'm (obviously) going to do something with astronomy, but this looks like it just might be one of the best areas to do that in. The main issue I can see is, getting the scope on target. 

The mount I'm using is an EQ5 with dual axis motors (non GoTo, I believe its advertised as the EQ5 Deluxe on FLO). I have yet to use/tackle the coordinate setting circles, as most of the objects Ive been imaging have been easy to star hop to, but I'm assuming they will be accurate enough to bring a star into the (quite large) FOV. The problem is...how would I know which star was which! Guiding and Imaging present no issue, though whether the 1000d would be able to detect the brightness changes, I don't know, the sensor is quite a grandfather of DSLRs.  

Are there any books you can recommend on this subject? I am a complete noob to this area!  

John

[dokeeffe]

1 hour ago, JohnSadlerAstro said:

problem is...how would I know which star was which!

You need to save/convert your image files as fits files and plate solve them to add metadata called WCS. Then if you use any fits viewer application it will show you RA-DEC as you hover the mouse over the image. You can upload to http://nova.astrometry.net/upload to do this. However it would take ages for hundreds of images. I have installed it locally (http://astrometry.net/doc/build.html) which is not exactly trivial but I can platesolve hundreds of images in a few minutes.

Maybe start out with basic photometry. There is a good DSLR guide on the aavso site https://www.aavso.org/observing-main

This book is excellent even its 10years old > http://brucegary.net/book_EOA/EOA.pdf

Derek

[RobertI]

13 hours ago, dokeeffe said:

Yes I can explain a bit more on how I did it. I guess there are 2 main parts, data collection and data reduction

# Data collection

I use INDI and EKOS to control a small observatory. It has a Celestron CPC1100, Atik 383L+ and filter wheel and OAG. I created a schedule to take 200 60sec images of this star through a photometric V filter (probably should have used a different filter). INDI managed everything, startup, opening roof, shutdown, parking, closing roof. I was not in the observatory when the capture happened. I used this website to decide the target and timing http://var2.astro.cz/ETD/predictions.php?delka=-8&submit=submit&sirka=52

# Data reduction

On shutdown, a shell script gets called to do basic reduction of the collected images (bias,dark,flat correction/calibration). It also plate solves the images and copies to dropbox. This reduction is done using astropy ccdproc. https://github.com/astropy/ccdproc. This stage is really important. The quality of data really depends on this calibration. All the flats, bias, darks are recent and updated regularly. Once I grab the calibrated images from dropbox the manual work starts. I loaded the images into astroimagej and used that to do the analysis. Its pretty easy to do this once you know where to click (will make another video of this process soon). Astroimagej is good for exoplanet analysis but for standard single measurement photometry I use astropy photutils.

The time lapse video was assembled using a python script from the images taken with the all sky camera and the plot from astroimagej. If anyone wants this let me know. Its very much hacked together...

Derek

Fascinating, thankyou for posting. My interest has been tweaked, I shall have a read. Is a high resolution camera needed for this kind of work or would a low res but highly sensitive camera like a Lodestar work?

 

 

[dokeeffe]

On 1/18/2018 at 23:02, RobertI said:

Fascinating, thankyou for posting. My interest has been tweaked, I shall have a read. Is a high resolution camera needed for this kind of work or would a low res but highly sensitive camera like a Lodestar work?

 

 

I would try the lodestar first. No real need for high res. Better to match pixel size to scope so you get a FWHM of 2/3 pixels. Lodestar has quite large pixels so would suit a C8 maybe...

[Helen]

This is great fun

@JohnSadlerAstro

Here are some good educational resources:

As a starter.... https://lco.global/agentexoplanet/

 

http://resources.faulkes-telescope.com/course/view.php?id=131

http://resources.faulkes-telescope.com/course/view.php?id=105&topic=0

You could also sign up to the Faulkes project and get direct access to large scopes for your pupils to gather prfessional level data (particularly useful when you can't rely on British weather!!)

Helen

[robin_astro]

Yep the bigger dips are certainly possible to see using very modest equipment. Here is an observation of mine back in 2004 using just a modified webcam when the technique was in its infancy even for professionals.

http://www.threehillsobservatory.co.uk/astro/TrES_1.htm

The  Sky at Night magazine even got me to write up a tutorial on how I did it, published back in 2005 and reprinted here

http://www.skyatnightmagazine.com/feature/how-guide/how-find-exoplanet

Cheers

Robin