Detecting exoplanet

[Petar b]

Could anyone tell me what exposure length I should use and what filter I should use when detecting an exoplanet. I was planning to take pictures with T17(http://support.itelescope.net/support/solutions/articles/231915-telescope-17) or T21 (http://support.itelescope.net/support/solutions/articles/231906-telescope-21)using the iTelescope service. And which one is better?

[michael8554]

"A Practical Guide to Exoplanet Imaging"

http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwj8zemZ7vreAhXFGuwKHQqIDkwQFjAAegQIChAC&url=http%3A%2F%2Fastrodennis.com%2FGuide.pdf&usg=AOvVaw3OqHDfwoRUtuH7qISywI98

"Exoplanet Observing for Amateurs"

http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwigkJzr7vreAhWCCOwKHSekAygQFjAAegQIDxAC&url=http%3A%2F%2Fbrucegary.net%2Fbook_EOA%2FEOA.pdf&usg=AOvVaw2I6kNV7dcqqqe4CM1VkrcR

From what I've read you need to know the linearity of the CCD, and then expose to just below that value, using a CBB Clear Blue Block Filter.

Michael

 

[robin_astro]

Interesting.  Can you rent these robotic telescopes for long enough to follow a transit? (You would need to have sole use of the telescope for several hours at a particular date and time)

Robin

[robin_astro]

By my reckoning a single transit using 3 hours on the T17 would cost about $600 (or $300  during full moon) 

https://go.itelescope.net/info.aspx

Robin

[Petar b]

I could image around 2-3h max.

[michael8554]

Again, from what I've read, you image 30 mins either side of the transit to get a good reference, which leaves you a 2 hour transit - one of the WASP ones perhaps ?

Michael 

[Whirlwind]

I would suggest that if you've never undertaken transit photometry before then I would stay well away from these telescopes to start with otherwise you might find it's an expensive waste of money.  Even someone that knew what they were doing from a transit photometry perspective (bur didn't know the telescopes) would likely need 30 minutes just to find the settings that they would need.  If you have your own equipment then try using this first. WASP exoplanets were all found using camera lenses. If you have even some modest equipment like a C8 and a DSLR then you should be able to capture these from your own home.  Once you are comfortable with this methodology and you want to advance then you can start to think of using the semi-professional set ups.  You can even practice with some variable stars first (there are examples on this forum).

Oh and there is a database here that will give you transit times:-

http://var2.astro.cz/ETD/index.php

[Petar b]

I already paid so I have no choice but to use them. The gear I have is an 80mm ed that is currently being repaired and a DSLR (Canon EOS 700D). I think I am going to use the T21, and I will try to detect WASP-52 b.

[Whirlwind]

OK, well you could use the T5 instead.  That should get you some more time.  The transit appears to be just short of two hours which means you need 3 hrs minimum really (30 mins before and after the transit).  

Here is an example using a C8.  The T5 is a 10" (likely in better conditions).

http://var2.astro.cz/EN/tresca/transit-detail.php?id=1382798470

[robin_astro]

Can you change exposure on  the fly with these systems ie fine tune on the run into the transit or are you committed to  preprogrammed sequence of exposures ?

I would take a test shot or two first before committing to a full 3 hour run to check exposure keeping in mind that the star will get brighter as it gets higher.  Also check the image field with the telescope and camera you choose to make sure you have some useful comparison stars in the field and these are correctly exposed. 

Robin

[Petar b]

5 minutes ago, robin_astro said:

Can you change exposure on  the fly with these systems ie fine tune on the run into the transit or are you committed to  preprogrammed sequence of exposures ?

I would take a test shot or two first before committing to a full 3 hour run to check exposure keeping in mind that the star will get brighter as it gets higher.  Also check the image field with the telescope and camera you choose to make sure you have some useful comparison stars in the field and these are correctly exposed. 

Robin

6

Technically you can but it is a little bit complicated. To see my field of view I would use this website: https://dso-browser.com/deep-sky/object/859/pleiades/m-45/bright-nebula?itelescope_mpc_code=H06&date=2018-12-01. You can change the camera angle and what telescope you are using.

 

22 minutes ago, Whirlwind said:

OK, well you could use the T5 instead.

I've used it before and half of the time the guider wasn't working.

[robin_astro]

I would say if this is to be successful, preparation is going to be everything (even down to keeping an eye on the weather forecast. I know these are good sites but you don't want to be clouded out half way through a run. ) To maximise the probability of success, I would probably do a short, say 10-15 min trial run outside transit first and analyse these  before committing to a full run to understand better what the potential issues are. 

Robin

[Petar b]

That's what I thought. First I would do a 15min trail run and test what would be the best exposure. Maybe even 30min run to try capturing the beginning or the end of the transit.

[robin_astro]

It is tempting to catch ingress or egress but I would say you would be better doing the test run outside transit.  Then you will know for certain if your processed results are good enough to detect the transit, ie your light curve should be a horizontal line with variability less than the transit depth

[Whirlwind]

On ‎01‎/‎12‎/‎2018 at 18:24, Petar b said:

I've used it before and half of the time the guider wasn't working.

You might not need to guide as long as the tracking can keep the object on the roughly the same pixels over the three hours and there isn't a drift.  It is a paramount so I would be hopeful.  It also has a large pixel scale (4.4") so any residual PE should be less important.  To put it into context this setup used a 115mm refractor:-

http://var2.astro.cz/EN/tresca/transit-detail.php?id=1514312910

It is a much less sensitive CCD as well and still observed the transit using 60 seconds exposures.  So on the T5 with approx 3 times the aperture (after taking into account the central obstruction) and a CCD that is close to twice as sensitive I would 

The full well depth is greater but you still only want to reach that to about 50% in any one particular exposure.  Otherwise sudden improvements in seeing, high cloud can result in pixels reaching non-pixel regimes and blooming.  My best guess would be that you won't need anything more than 30s and even that might be too much....

At these length exposures you shouldn't really need guiding as long as the mount is working properly.

[Petar b]

The T5 sounds tempting because with that telescope I could maybe detect 2 transits in multiple nights because it is less expensive. Could I measure the full well depth of a star in a picture? So that I know what exposure to use. And how accurate are these predictions: https://astro.swarthmore.edu/print_transits.cgi?observatory_string=Specified_Lat_Long&use_utc=0&observatory_latitude=32.5413&observatory_longitude=-105&timezone=MST7MDT&start_date=today&days_to_print=28&days_in_past=0&minimum_start_elevation=+&and_vs_or=or&minimum_end_elevation=+&minimum_depth=25&target_string=&single_object=0&ra=&dec=&epoch=&period=&duration=&target=&show_ephemeris=0&print_html=1&twilight=-12&max_airmass=2.4

[Whirlwind]

I haven't used Itelescope so am not sure what software you are given access to.  I am assuming a lot is hidden behind a web interface to stop people messing with settings.  As such it really depends on whether you get an active display or not that you can put an aperture over. It will still be 16 bit converter though although the maximum well depth will be greater than 100,000 I'm assuming it will report back only 65536 maximum levels (2^16).  So you want roughly half of this value in any one pixel on average, it will swing below this and above this during you observations and it will change steadily because of the airmass change (pointing altitude).  Looking at the spectral type (K2V) you'll probably want to use the V filter (not the B or I).

If you can't see the output flux live then you should email itelescope and see whether they have any test exposures of varying exposures for the telescope using this filter.  This way you'll get a good idea of what exposure you might need.  If you are unsure on the night it is best to err on shorter exposures.  You can always bin short exposures to increase your signal to noise.  If you loose flux due non-linearity/bleeding then you can't recover this.

[Petar b]

I took some test images with the T21 and these are the averages of the star are: for 15s: 3000-4000 median value, for 30s: 5000 median value and for 60s: 7000 median value.

 

[Whirlwind]

Is that the median value across the aperture or is it the highest pixel value?  It is the highest pixel value that is important and it is this that needs to be less than about 35000

[Petar b]

It is the highest pixel value. For 30 seconds it is 6357

[Whirlwind]

You should be able to increase the exposure time then (I'm assuming it was perfect conditions, no high cloud etc).  However, it will reduce the sampling of the transit (especially the ingress and egress).  You would be able to bin data points if you left it at 30s as well if you needed to so I'd probably leave it at this.

[robin_astro]

Have you tried measuring the brightness in your check images?  What uncertainty does the software report? (You will need to enter the gain of the camera in e-/ADU)

This will tell you how many sub exposures you would need to combine to detect the transit. 

Robin

[robin_astro]

Using Michael Richmond's photometry calculator

http://spiff.rit.edu/richmond/signal.shtml

you should be able to get a signal/noise ratio  ~ 1000 (ie a measurement uncertainty of 0.1%) on a mag 12 star with a 0.5m aperture and 30s exposure so easily good enough to detect a 2.7% transit.  A median count in the star image of 6000 seems rather low to me for a telescope this size though this will depend on the FWHM of the star image (focal length seeing focus pixel size etc) .  What was the integrated count and what is the camera gain?

Robin

[Whirlwind]

There's not enough information really.  Poor seeing, high cloud, using the 'wrong' filter (e.g. a B for a cool star etc), what the airmass was would all have an impact. The pixel scale seems to be about 0.98"/pixel so you could always bin as well which might help slightly.  The telescope might even be out of focus? Difficult to say without seeing an image.

[Petar b]

Here is some of the information. I used the T21. Filter: V, for the clouds and the seeing I don't know I forgot to check the forecast, but it was clear if that means anything. I can't change the gain of the camera. And the image was a little bit out of focus. In the files, you will find 15sec (6 images), 30sec (5 images) and 60sec (5 images) fits files.

Archive.zip