Detecting an Exoplanet from your backyard!
In 1995, Michel Mayor and Didier Queloz announced the discovery of the first exoplanet orbiting a main sequence star, 51 Pegasi b. They used the “radial velocity” method to detect small changes in the movement of the star, 51 Pegasi towards and away from the earth as a result of the gravitational pull of the orbiting planet.
The planet they detected had a mass around half that of Jupiter. What was surprising was how close it orbited to 51 Pegasi – at a distance of 0.05 AU (1/8th the distance between Mercury and the sun) with an orbital period of just 4.2 earth days! Clearly this was nothing like the gas giants we see in our own Solar System. At the time, the idea of a gas giant orbiting so close to its star seemed very strange.
However, their discovery was shortly confirmed by a number of other teams, opening the door to a new era of exoplanet discovery. Over the following 25 years over 4000 exoplanets have been discovered, many of which are classified as these “Hot Jupiters” – large gas giants orbiting close to their host star.
Mayor and Queloz received the Nobel Prize in 2019 for their contribution to this area of astronomy. Around 800 other exoplanets have been detected using this radial velocity method. Whilst this is probably out of reach for back garden astronomers in the near term, the good news is over 3000 exoplanets have been detected using a different method which is much more achievable using amateur telescopes.
The "Transit Method"
This method of detecting exoplanets involves measuring changes in brightness of a star as an exoplanet passes between the star and the earth.
In the same way as during a solar eclipse when the moon momentarily blocks out the sun’s light, if an exoplanet transits in front of its host star, it will cause a dip in the star’s brightness.
So, by measuring the change in brightness of a star during one of these transits, it is possible to detect the dimming caused by an exoplanet passing in front of it.
Process for Exoplanet Photometry
In a nutshell, we are going to use a camera sensor as a photon detector to measure the varying brightness of a star.
By taking a series of photos of a star, say every 2 minutes: before, during and after the exoplanet's transit and analysing how the light registered on the camera sensor changes over this time period relative to other stars in the frame, we can extrapolate the brightness level of the star during a transit. The four steps for photometry therefore are:
1. Planning: to identify a suitable target
2. Setup: similar to DSO astrophotography
3. Imaging: to take a series of scientific images frames
4. Photometry: to analyse image frames and produce a light curve
In the next section we will cover these four steps in more detail.
1. Planning
The two key considerations to picking a target are:
1. Is it detectable with your equipment?
2. Will it have optimal visibility from your location?
There are a number of resources you can use to identify a target that should be detectable with your equipment.
The Exoplanet Transit Database (see http://var2.astro.cz/ETD/predictions.php; Poddany et al., 2010) is superb and provides up to date ephemerides of several hundred exoplanets. You simply plug in your location and it shows the expected timings of transits at your location in addition to the transit depth to help identify if it is likely to be detectable with your equipment.
In addition to ensuring the transit begins and ends while there is darkness in your region, ideally you should look for a target that will be visible under dark sky for a full hour before and after the transit to detect the ingress and egress more clearly.
Another important consideration is whether you will need to perform a meridian flip during the transit. If the meridian flip coincides with the start or end of the transit, that is clearly not ideal and even a meridian flip during the middle of the transit adds the usual complications of lost imaging time, refocusing, identifying a new guide star and inverted images for subsequent alignment and processing.
Stellarium includes a database of most exoplanets (press ctrl+alt+e to show the exoplanets visible in your sky).
Stellarium can also be used to calculate the position of possible targets throughout the night.
It is best to perform photometry on a target that is >30 degrees above the horizon for the entire imaging session. And the closer it is to the zenith the better.
Weather.
Yes the bane of all northern hemisphere astronomers. Since our goal is to take a time series of photos over a few hours, you can imagine that any significant atmospheric changes can generate a lot of error. High cloud cover, wind or moon reflection makes photometry very difficult.
So I would recommend only attempting exoplanet photometry if you have a clear 3 hour window of limited wind, no cloud and less than 30% moon.
2. Setup
The setup for photometry is very similar to deep sky astrophotography.
You will need a telescope with an aperture of at least 4” but ideally with 8”+.
A reasonably long focal length is also helpful so that your target star crosses a number of pixels to prevent saturation of the star light. However, there is a balancing act here: if the focal length is too long you may require too long an exposure to obtain a good enough SNR for the image.
Even relatively small refractors have been used to detect larger exoplanets.
As with Astrophotography, it is very important to correctly balance, polar align, and have good guiding for photometry. Click below to see my guide on setting up a telescope for astrophotography
Almost any camera will work for photometry. However, the bayer matrix of a colour camera will complicate things so it is best performed with a monochrome sensor and R,G,B filters. Traditionally people have used CCD sensors but modern monochrome CMOS camera have sufficient sensitivity to be successful too. You will need to experiment with different filters and exposure times to get a strong enough SNR without risking saturation of pixels. Photometric filters are commercially available on the internet and are a must if you are attempting to calibrate precise magnitudes. But if you are simply looking to generate a relative flux light curve, standard RGB filters will work fine.
It is very important to take Calibration Frames along with your science frames to mitigate possible sources of error. For more information on how to do this, see the Calibration Frames page by clicking on the below button.
3. Imaging
Once you are ready to begin photometry, the first challenge is actually finding the target star!
Some of the most interesting and detectable transits occur on relatively obscure stars of magnitude 10 or more. So if you are not an expert with a star finder chart, a go-to mount and possibly even pate-solving are extremely helpful. Take a few practice frames to make sure your target is as close to the centre of frame as possible to minimise the impact of vignetting.
The process of photometry is then pretty straight forward, once you have the target star in the centre of the frame experiment with different exposure times, binning and filters to get a strong SNR.
Once you have an optimal image, setup an imaging run using your normal image capture software. I use ASIair but any astrophotography image capture software will work. Try to save your images in FITS format so that your computer populates the FITS header information with the location of the target and the time of each observation. This will make processing easier.
You should aim to take your first image 30-60 mins before the expected start time of the transit and aim to take your last image 30-60 mins after the expected end of the transit.
4. Image Processing and Photometry
There are a number of photometry applications available on the internet include AstroImageJ (AIJ) and Aperture Photometry Tool (APT). They generally work in a relatively similar way and the choice probably comes down to personal preference and how well they are suited to the project you are working on.
My personal favorite is HOPS which is a Python based application developed by Angelos Tsiaras at UCL as part of the Exoclock team who are observing potential exoplanet targets ahead of the Ariel Space Mission.
This software can be downloaded on the ExoWorlds webpage:
The first thing you will need to do is download the Anaconda Python toolkit from:
https://www.anaconda.com/products/individual#Downloads
Then download the HOPS application and open it. You will see straight away that it is very intuitive involving 6 very clear steps to upload, analyse and produce a light curve for your images.
The first step is to put all of your science frames and calibration frames in the same folder/. Click on "select data & target" and identify this folder containing all of your image frames.
Once you have selected the folder that your image frames are located in, tell HOPS how to identify the different types of files. A good practice is to label your science frames with "science" or "light" and label your calibration frames with the name of the type of calibration used.
HOPS will automatically identify your science and calibration frames based on this file name identifier. Assuming you have saved your files in FITS format, HOPS should be able to identify the date and time of each observation from the FITS header data.
HOPS does not recommend binning the images. But depending on the speed of your computer or the size of your files, it may speed up the process if you reduce the data by using 2x2 or even 3x3 binning.
Once all of the information has been provided click on "Save Options and Proceed"
HOPS will now attempt to reduce and calibrate all of your frames. This step may take a bit of time.
On the next window you can review the quality estimate of your data.
The top graph provides an indication of the background noise. A high value indicates background sky glow. A low value indicates dark sky. In this example, my target was moving progressively higher in the sky throughout the observing session. Because I am in a relatively light polluted area, the sky quality was progressively improving as the imaging session went on.
The second graph provides an estimate of the Half Width at Half Maximum, an estimate of how "spread out" the stars are. A high HWHM indicates the stars are further away from being point sources and may have suffered from focus or mount tracking issues.
You may want to consider excluding lower quality frames. Once you are happy with your data set, click "Save Options & Proceed"
HOPS will now try to align all of your frames based on the star locations. This step may take some time. Potentially longer if you have had to adjust your camera or perform a meridian flip during the imaging schedule.
Now it is time to perform aperture photometry. Select the target star and HOPS will select a number of suitable comparison stars with similar size, brightness and relative camera sensor location.
HOPS will now perform Photometry on the target star. By measuring how the brightness of the target star changes compared to 2 (constant brightness) comparison stars.
If the comparison stars show variability you will need to select different comparisons and re run the photometry. Similarly if the light curve is not very clear, you may need to tweak the aperture settings. Once you are happy that you have a detectable light curve, you can proceed to Fitting.
This final window provides key information around the data you have collected. Most importantly, how the transit timing varies from prior observed literature timings.
Hopefully you now have a light curve showing the dip of your exoplanet!
Congratulations! You can now submit your data to Exoclock or one of the other citizen science projects.
If you have not managed to detect a light curve, don't despair. Remember you are trying to detect a 1% change in brightness from a planet thousands of light years away. From your back garden. This is supposed to be difficult.
It took me three attempts to produce any meaningful data. But as with all aspects of science, try to enjoy the process and remember that the more challenging the task, the greater your satisfaction with the final product.