AstronomyOnline.org
home observation science solar system stars our galaxy cosmology astrobiology exoplanets astrophotography
Exoplanets - The Search for Other Worlds - by Ricky Leon Murphy:

Back to Exoplanets

Welcome to part one of my two part series in Exoplanet Detection and Research. Part one, Exoplanets - The Search for Other Worlds, was initially a project for one of my courses with the Swinburne Astronomy Online Master's Degree program. While this paper was written to serve the purpose of competing an assignment, I wrote this to introduce exoplanet studies from the perspective of an amateur astronomer. Part two, On the Stability of the Gliese876 System of Planets and the Importance of the Inner Planet, was a major project paper designed to meet the requirement for the Master's degree program. Part two looks at how a professional astronomer uses data collected in exoplanet detection to determine orbital characteristics.

To read part two, click here.

Introduction
The Formation of a Solar System
Evidence of possible planetary formation
Detecting Exoplanets
Radial Velocity
The Transit Method
Radial Velocity – Tau Bootis
Issues to overcome
Telescope aperture
Spectral Resolution
Signal to Noise
Light Loss
Spectroscope stability
Transit Method – HD209458
Other Methods
A Brief Window into the Future
Conclusion
Recommended Internet Resources
References
Image Credits

Back to Exoplanets

Introduction:

As of September 2004, there are 136 known planets outside our solar system (http://exoplanets.org). These extra-solar planets, or exoplanets, are one of the most current and highly studied subjects in Astronomy today and it is one of the very few subjects that involve both amateur and professional astronomers. The huge telescopes perched atop Mauna Kea in Hawaii are pointed at these objects, as are 8” telescopes purchased from the local shopping mall – and many others around the world. Why are we finding these planets now if only 8” telescopes can detect them? Simple; we know what we are looking for and we have better tools to get the job done. While telescope size does not seem to matter with the search and study of exoplanets, it’s what you do not hear about that really matters: improved sensitivity in CCD cameras, improved resolution in spectroscopy, and fast computers to perform the mathematics. The goal of gathering repeatable data is very important when studying exoplanets. The rewards of such study carry implications across the board in Astronomy: we can learn about our own solar system and test the theories of solar system formation and evolution, improve the sensitivity to detect small Earth-like planets, and possibly provide targets for the spaced based telescopes and SETI projects; however, the most important implication is that perhaps for the first time in history, amateurs and professionals from around the world are engaged in this subject and working together to share the data. The greatest reason for this collaboration is telescope time: there are only a finite number of professional telescopes with tightly guarded schedules that limit prolonged data collection. Amateurs have all the time they can spend and the equipment to help. There are several methods of detecting these exoplanets; however, amateur astronomers are only capable of performing only two – the measurement of radial velocity and the transit method, both will be discussed in detail.

Back to Top | Back to Exoplanets

The Formation of a Solar System

The foundation is the base in which ideas are built upon. In this case, the foundation of this project will be to briefly visit the current theory of how a solar system is formed. This is important because this gives clues for us to know what we are looking for when we study other star systems.  The formation of our Solar System can be traced all the way back to the Big Bang. With hydrogen being the most abundant element in the Universe, clouds of hydrogen began to form under its own gravity. By gravity and rotation, these clouds compressed to form the first stars (and galaxies) of the Universe – Population II stars. Population II is the designation for stars that do not contain heavy elements – that is heavier than helium. The natural cycle of these stars resulted in supernova explosions that introduced the heavier elements into the interstellar spaces due to intense heat and pressure of the explosion. While hydrogen is still the most abundant element, other heavier elements are now present – elements like iron, carbon, silicone and many others. Supernova stimulates nearby hydrogen clouds and introduces the heavy elements. As the stimulated clouds collapse, they form something a little different, a proto-planetary disk, as well as the central proto-star (figure 1). When these metal-rich stars – called Population I – are formed, they may host a ring of molecular material that begins to collide with one another resulting in the sweeping up of material. 
 

 

Figure 1: The molecular cloud begins to collapse, eventually forming a dense area at the heart of the cloud. The continual contraction raises the temperature of the cloud causing rotation. Eventually, there will be enough heat and density that a T Tauri star will form at the heart of this cloud with the remaining disk material possibly forming planets.

A: the slowly rotating molecular cloud

B: Faster rotation with denser and hotter central region

C: Faster rotation with T Tauri star at the center of rotation

By retaining orbital momentum, the shape of the disk is formed. The consequence of this is the formation of larger objects that also contain gravity and also spin as a result of the orbiting momentum and the further collection of material as it “sweeps” through the ring. These objects are called planetesimals, and continue to collect more material as they continue to orbit the host star. Proof to support this Solar Nebula theory has been discovered by the study of ancient meteorites - called chondrites - found on Earth (Beatty et al, 1999). The material makeup of the chondrites, which originate from space, is found to contain the same elements found on Earth – one of which is an isotope of hydrogen called deuterium. Additional proof to this theory comes from remarkable images taken by the Hubble Space Telescope. Looking deep into the heart of the Great Orion Nebula (figure 2), the Hubble Telescope was able to spy tiny solar systems in the making (figure 3). These images show the T Tauri star in the center of a protoplanetary disks, or proplyds. This offers us a remarkable look into the very earliest history into the formation of a stellar system (Figures 4, 5, and 6).
 

Figure 2: This beautiful image of the

Orion nebula was captured using

special filters by amateur astronomer

Russell Crowman using a 14.5 inch telescope

and a Santa Barbara Instrument Group

ST-11000 CCD camera.

 

 

Figure 3: This close-up image of the center of Orion nebula – care of

the Hubble Space

Telescope – shows several knotted looking objects.

These are the proplyds.

While these proplyds look very impressive, the host T Tauri stars are stars that have not initiated hydrogen fusing – and must be observed in the near infrared. Much of the material that surrounds the proto-star will be blown away by the shockwave resulting from the initiation of hydrogen fusion at the heart of the star (Ostlie and Carroll, 1996).
 

 

Figures 4, 5, and 6 (clockwise): These are Hubble 

Telescope close-up images of three of 

these knots. They show disks of dusty material surrounding

their host T-Tauri stars. 

Because of the density of the 

dust, these images are photographed in

 the near-infrared so the host star is visible.

 

Back to Top | Back to Exoplanets

Evidence of possible planetary formation:

Have you ever looked into a telescope at a star, only to find that the star does not appear any larger than with the unaided eye? This same phenomenon is familiar with even the largest telescopes. As a matter of fact, the only star that has been able to be resolved into a disk on a consistent basis is Betelgeuse - the red giant star forming the upper left shoulder of the constellation of Orion (Burnham, 1978). This red giant star is only 520 light years away, and with a diameter between 550 to 920 times our Sun (Betelgeuse is a variable star, so its size fluctuates) it can be easily resolved by our largest optical telescopes (figure 7). With this exception, the majority of stars cannot be resolved as a disk.
 

 

Figure 7: This image of Betelgeuse was taken by the 50cm COAST telescope. It shows the surface of Betelgeuse, which from Earth is only 0.1 arc-seconds across. For comparison, the planet Pluto is also only 0.1 arc-seconds across. On average, the planet Mars is around 3 arc-seconds across. 

 

With this fact alone, it seems impossible to detect a small planet orbiting a star – after all, if we cannot even resolve the star, how can we possibly detect a planet which is much smaller and does not give off any light? There is one method of imaging that will allow us a view of the end stage of the protoplanetary disk. By masking out the bright central star, it is possible to image the residual disk material called the circumstellar disk (figure 8). 
 

Figure 8: By masking out this central star (star designation HD 141569A), imaging of the circumstellar disk is possible. All the stars in the field are overexposed, so look much larger than they would be under normal imaging circumstances.

Masking of the image is important as the host object is a main-sequence star – that is, a star that has already initiated hydrogen fusing. While much of the protoplanetary nebula can be swept away during the initiation of hydrogen fusion, these images of the circumstellar disks around a main-sequence star tells us that planetary formation is even more likely than with evidence of the protoplanetary disks.

While the proplyds and circumstellar disks offer evidence to material that can result in planetary formation, another type of object can also be used to look for evidence of circumstellar disk formation. Herbig-Haro objects (figure 9) are T Tauri stars with an active circumstellar accretion disk (Ostlie and Carroll, 1996). The rotation of this disk is shown by massive lobes of gas that appear perpendicular to the rotating disk. These tell us that the protoplanetary material does in fact rotate about their host star, which can result in planet formation.

Figure 9: These are four Herbig-Haro objects photographed by the Hubble Space Telescope. The green jets are the expulsion of gas from the perpendicular circumstellar disk. These jets are present as a result of circumstellar disk rotation.

Perhaps the most remarkable image of a circumstellar disk is that of Beta Pictoris. This Hubble Space Telescope image (figure 10) is using the masking technique to block the light of the over-exposed star to reveal what looks like several distinct orbits of dense material. While the orbits appear to be very elliptical, it is remarkable that such orbits can be resolved at all.

 

Figure 10: The first image of what appears to be four distinct rings orbiting the star Beta Pictoris. While no planet has been detected in either of these orbits, this supports our theory that as the disk material continues to rotate; they begin to form individual planetesimals. This image may show the early stages of such evolution.

Back to Top | Back to Exoplanets

Detecting Exoplanets:

Radial Velocity -  

Now that we have identified the presence of planet making debris around other stars, let’s focus on how formed planets are detected around stars. Until direct detections are made, we must rely on indirect methods. The two main techniques in detecting these planets are radial velocity measurements and the transit method. Both are used by professional and amateur astronomers. 

The first exoplanets were discovered by what is called the “wobble.” This sounds low tech, but this is very significant. When two objects orbiting each other contain any mass, they will have an affect on each other resulting in an inertial center - called the barycenter (Mayor and Frei, 2003). Using our own solar system as an example, Saturn and Jupiter provide enough mass that the effect is a wobble of the Sun (Marcy and Butler, 1997). The net affect of both planets produce a wobble of around 13 m/s (and if Jupiter were the only planet, a wobble of around 12 m/s would be present). Studies of pulsars and binary stars show that both stars rotate about a common point and not around each other (Mayor and Frei, 2003). The implication is the smaller the companion, the less dramatic this rotation will be. Because of such small variances in stellar wobble, detection is only possible via measurement of the Doppler shift (Butler et al, 1996). Either way, you can think of this effect being analogous to a very heavy person and a very light person on a see-saw with a moveable focus. 

The Doppler shift is a method used in almost every area of Astronomy, from mapping out the rotation of our galaxy to study of the expanding Universe. When an object like a star (or anything that give off or reflects light) is in a stationary position, it can be viewed with a spectrograph to reveal a color spectrum with a specific footprint. Because various chemicals exist, these result in missing spaces within a spectrum called absorption lines. An example would be a star that contains all hydrogen1. Light given off by this star would reveal the familiar color spectrum, but because it contains hydrogen, a certain frequency will be absorbed because of interactions with the hydrogen atom. The electron in the atom absorbs some of the light energy and moves to a higher orbit around the nucleus. This absorbed energy - called a Balmer line when referring to the hydrogen atom - results in a missing portion of the spectrum as shown by example in figure 11  (Freedman and Kaufmann, 2001).
 

Figure 11: This shows an example of a single absorption line. The image at the top shows the object (a galaxy in this case) that is not moving, with the hash marks indicating where that line should be. The middle image shows the object moving away, causing the line to be shifted towards the red – called the redshift. The bottom image shows the object moving towards us causing the line to shift towards the blue – called the blue shift.

This shift can be measured. By using simple mathematics, comparing a reference object that contains identical spectra to the shifted object, it is possible to determine the velocity of the object.  

Equation 1

Determining Doppler redshift and blue shift:

z = =  

z = redshift

∆λ = shift in wavelength

λ = wavelength of stationary object

λο = stationary wavelength – reference spectra

v = velocity

c = speed of light (300,000 km/s)

It is important to state that:

Equation 1 is the basis of determining the orbital velocity of the object orbiting the affected star or determining the radial velocity of the affected star. Figure 12  shows how this works. Once the orbital velocity is determined, simple usage of Kepler’s Third Law will determine the distance the planet is to the host star.

Kepler’s Third Law:

P2 = a3

P = object’s rotation period in years

a = object’s distance to star in Astronomical Units

Figure 12: The unseen orbiting planet creates an inertial center of this system resulting in a wobble that can be detected by the shifting of the stellar spectra. Using equation 1, it is possible to determine the speed of this wobble thereby determining the orbital speed of the unseen planet. This is also called the radial velocity.

The tool used to measure the spectra of an object is called a spectrograph. There are many different flavors of spectrographs, but all of them work using the same principle: separate the visible light into their fundamental wavelengths. Figure 13  shows a very basic spectrograph that is using a special diffracting prism, which is a prism with at least one 60 degree angle. In general, prisms cannot be rotated to adjust resolution.

Figure 13: This is an example of a standard spectrograph. A slit is used to block unwanted light while the lens magnifies the image onto the screen – or our in our case, the CCD sensor.

The sensitivity of the spectrograph is a very important consideration and there are several factors that determine the sensitivity of the spectrograph: how much light can it see, the angle in which the spectra is being observed, quality of design, design of light beam travel within the device, as well as slit dimensions. The most important factor in spectrograph sensitivity is the use of diffraction gratings2 versus a diffracting prism for one reason only: the diffraction grating can be adjusted by the turning of a knob to improve resolution (Tonkin, 2004). Discussing the theory and various flavors of spectrographs is beyond the scope of this paper, so we will focus on the spectrographs used by both amateur and professional planet hunters. I highly recommend Stephen Tonkin’s book Practical Amateur Spectroscopy  for additional study.

Additional precision of the Doppler shift measurements are possible by extending the focal length of the light within the spectrograph. This is performed by using mirrors to fold the light while collimators (lenses) are used to keep the light focused. An example of this type of spectrograph is the Echelle (figure 14), which happens to be the design used by the California & Carnegie Planet Search group (http://exoplanets.org/ ).

Figure 14: The Echelle spectrograph extends the focal length of the light by using mirrors and collimators. The collimators keep the light focused. The diffraction grating is at the bottom of the image – labeled Echelle.

The spectrum of a star can contain a large number of absorption lines as a result of the many elements present in the atmospheres of stars. Because of this, the spectra must be extended to allow viewing of all the absorption lines. This is the main reason why the light path within the spectrograph must be extended (Kitchin, 1998). The spacing between the absorption lines allows the Doppler shift to be determined (equation 1). 

While the Echelle is capable of Doppler precision measurements of around 15 m/s, a method of introducing iodine gas near the slit entrance has allowed for precision measurements of up to 3 m/s (Butler et al, 1996).  The iodine is use to create a composite spectra to overlay the analyzed star that enhances our view of the absorption lines while acting like a ruler. By eliminating any uncertainty between stellar absorption lines with a laboratory standard, precision measurements are attained. 

While the use of iodine has enhanced our abilities to take accurate measurements, the choice of targets also play an important role. The majority of exoplanets discovered have been around metal rich main sequence stars (with a few exceptions) – specifically stars with a spectral class of F, G and H (Butler et al, 2000). The reasons are twofold:

            1. F, G, and K type stars are “normal” sized stars like our Sun, and will more than likely exist long enough for planets to form. Larger, hotter burning stars end their lives much sooner so the possibility of mature planets to form is highly unlikely; although planets have been found to orbit stellar remnants such as pulsars.

            2. Metal rich stars contain heavier elements in their atmospheres as a result of enriched molecular clouds from which they have formed. This results in more absorption lines to be examined.  

There is a disadvantage to using the Doppler shift to measure radial velocities: the star must be as close to the host star are possible. For example, one of the first exoplanets discovered – the companion to 51 Pegasi – is only 0.05 AU’s3 (Mayor and Frei, 2003). That means the planet, which is 0.5 times the mass of Jupiter, is much closer than the orbit of Mercury is about our Sun. For planets that orbit at a larger distance from the star, more precise astrometry measurements are desired – mostly because larger orbits require many years to study versus days of a closer orbiting planet. Both radial velocity and astrometric observations by professionals have revealed a number of exoplanets; however, astrometry will not be covered here as this is beyond the current capabilities of the amateur astronomer (for now!).

Back to Top | Back to Exoplanets

The Transit Method: -

Another important method of detection is by measuring the transit of a planet over the face of the host star. By performing careful photometric plots of the host star, drops in stellar brightness as a planet moves across the face of the star can be measured (figure 15). The first ever measurement of a stellar transit was made by the Elodie group (discoverers of the companion to 51 Pegasi) and the David Latham group (Mayor and Frei, 2003). The results were shared (as with almost all things in Astronomy) to David Charbonneau and Timothy Brown who are the directors of project STARE (STellar Astrophysics & Research on Exoplanets). What is remarkable about project STARE is their equipment: A 12” Schmidt telescope and a 2K by 2K CCC camera mounted on a Meade LX200 computer controlled mount (http://www.hao.ucar.edu/public/research/stare/stare.html ).

Figure 15: As a planet moves across the face of a star, the brightness curve of the star drops and can be measured using sensitive CCD camera and computer software.

The transit method can only be used for planet systems that face us edge on (as the orientation of figure 15  indicates). An additional limitation is the size of the planet. As we will see later, a planet 0.64 times the mass of Jupiter decreased the brightness of its host star by only 0.0011 magnitudes.

Determining the transit of an exoplanet is not as difficult as performing radial velocity measurements. While determining the radial velocity requires carefully calibrated equipment, specialized spectroscopes, and lots of patience, determining the transit only requires the skill of photometry and a personal computer. Quite simply, photometry is the study of stellar brightness. Stars with a particular brightness (called luminosity by astronomers) have associated relations to size and spectral class. Using online databases and star charts, we can determine the accepted value of brightness for any given star. However, there are a group of stars which can fluctuate in brightness called variable stars. As a star leaves the main sequence and begins to burn up what little hydrogen is left near the core, the outer layers of the star expand and contract. This is very convenient because this “breathing” of the star can be studied using same iodine infused high resolution spectroscopy to determine the speed and duration of this breathing. By gathering a large sample of variable stars that inhabit the Cepheid variable strip (the numerous and most common type of variable star) and evaluating them over a long period of time,  it has been concluded that such stars are photometrically stable, and demonstrate peaks of radial pulsation anywhere between 50 to 80 days (Butler, 1998). While variable stars have a particular sequence in their variations in brightness, a handful of normal main sequence stars have been shown to oscillate. Once again, high resolution iodine induced spectrography has revealed a definite pattern of oscillation that can also induce slight variations in stellar brightness (Bedding et al, 2001) when measured photometrically. While this seems to carry serious implications to the accuracy of transit measurements, variations in luminosity as a result of a transit fall in between oscillations – which can occur rapidly, and variations due to the expanding shell of a older star, the stability of which is shown by a much greater delay (table 1).
 

Table 1: Approximate time variations as a result of competing causes.

Stellar Oscillations

Several times a second

Transit of an exoplanet

A few days

Photometrically stable variable star

50 to 80 days

(Butler, 1998)(Bedding et al, 2001)

Back to Top | Back to Exoplanets

Radial Velocity – Tau Bootis:

Star Name

Tau Bootis

Distance

15.6 parsecs

Apparent Magnitude

4.5

Spectral Class

F7

Metallicity

0.28

Planet Mass

4.13 time the mass of Jupiter

Orbital Distance

0.05 AU

Orbital duration

3.313 days

Measured Radial Velocity

15 km/s

(Fischer et al, 2001)

A group of dedicated amateur astronomers at the Spectrashift.com  extrasolar planet search project have successfully measure the radial velocity of a known extrasolar planet around Tau Bootis.
 

To capture the Doppler shift using store bought equipment of such a small target is no easy task, however many of the hurdles that would have blocked any attempt to get accurate readings have been successfully avoided by designing a device specific to the measure of radial velocity. 

Issues to overcome:

            1. Telescope aperture

            2. Spectral resolution

            3. Signal to noise

            4. Light loss

            5. Spectroscope stability  

Telescope aperture:

Since project STARE has in its program a 12” telescope, telescope aperture does not play a vital role in photometric evaluations; however, the use of a spectrometer will require a greater aperture as the light entering the spectrograph is already reduced by the entrance slit. The Spectrashift.com  group has selected a Meade 16” SCT telescope (figure 16) for spectroscopy; however a 1.1 meter telescope is currently under construction.

Figure 16: While their 1.1 meter telescope is still under construction, the Spectrashift.com team successfully recorded the radial velocity values of Tau Bootis with this Meade 16” SCT.

Spectral Resolution:

This can be a major setback for amateurs as the design and stability of the spectrograph play an important role in resolution. This group designed and constructed their own spectrograph, so many of the inherent design limitations from commercial models have been eliminated – mostly because commercial models are designed for the study of a wide variety of spectra and compromise resolution as a result4. None the less, the best resolution of this particular piece of equipment can only measure radial velocities of 200 m/s (Tonkin, 2003). As a result, only large exoplanets can be studies. The spectrograph uses a Czerny-Turner design that uses internal beam folding similar to the Echelle (figure 17).
 

Figure 17: The Czerny-Turner design of the spectrograph allows for good resolution similar to the Echelle (figure 14) while at the same time allows for the swapping of internal parts to adjust resolution.

Signal to Noise:

The sources of noise when using a spectrograph are the equipment itself as well as the CCD camera use to capture the images. By using a stable table for the equipment, noise introduced by vibration can be eliminated. In addition, internal light reflections are eliminated by blackening the entire inner structures of the spectrograph. To eliminate noise from the CCD camera, additional cooling is required5. This group used a novel approach and used an office water cooler to feed cooled water through the cooling tubes of the CCD camera.

Light Loss:

Although directly related to telescope aperture and slit dimensions, the light loss of concern is in the fabrication of the fiber. Because of the large size of the spectrometer (as well as the desire to be stable), it resides on a table some distance from the telescope, and a fiber optic cable is fed from an eye piece adapter in the telescope to the entrance slit of the spectrograph. To best illustrate the design requirements of the fiber optic cable, I used Adobe Illustrator to show the fiber orientation:

Using small strands of fiber instead of a solid glass fiber is ideal. The fiber exits the eyepiece in its traditional round format, while the other end is terminated fiber over fiber so all the light can enter the spectrograph slit.

In addition to fiber design, careful polishing of the fiber ends is very important to maximize light transmission. Careful alignment on both ends is possible by using a reference light source as a guide. It will be necessary to have at least two of these fibers: one for the eyepiece and one for the reference light source – in spectroscopy, a reference light source (Argon was chosen by this group) is provided to calibrate the spectrometer (Tonkin, 2003). While fiber optics can be used on just about any telescope, telescopes with a large focal ration are desirable. The longer the focal length of the image, the more narrow the image cone. As a result, the fiber will be able to use all available light (Kitchin, 1998).

Back to Top | Back to Exoplanets

Spectroscope stability:

The stability of the spectroscope is just as important as maximizing available light. Any vibrations induced on the instrument will prevent accurate imaging of spectra, and may prevent imaging altogether as the already small target can shift as a result of cable movement. This issue is the reason telescope mounted spectroscopes – especially in the amateur world – are avoided. A very sturdy work bench, preferably isolated from any walkways – is highly desired to eliminate any induced vibrations.

Looking at the design requirements, the Spectrashift.com  group managed to construct a very nice setup using a Meade 16” SCT telescope on a computerized mount, a table-mounted hand made Czerny-Turner spectroscope, an Argon reference lamp, hand-made fiber optic cables, and an Apogee CCD (www.ccd.com) with a 512 x 512 pixel array with 24 micron square pixels6. The software of choice is MaxImDL (http://www.cyanogen.com/) for camera control and intermediate image processing, IRAF (http://iraf.noao.edu/) which is the standard for astronomical image analysis, and Microsoft Excel to create the plots. 

The process of gathering data, image reduction, and analysis is a very time consuming endeavor and will not be covered here. Instead, here is the sequence of events to serve as an overview.
 

          1. Inventory equipment and decide what will be used

          2. Test the equipment to ensure working condition

          3. Turn on computers, CCD cameras, spectroscopes, reference light sources, and any other pieces of equipment to ensure temperature equilibrium

          4. Connect the fiber optic cable from the Argon reference source to the spectroscope

          5. Ensure the argon spectrum will overlay the target spectrum – try this with a test star

          6. Using computer control, slew the telescope to the desired object (in this case, Tau Bootis)

          7. Confirm target in the eyepiece

          8. Remove the eyepiece and place the fiber optic cable

          9. Begin image capture – 45 minutes per exposure is typical, possible via computer tracking

          10. Periodically capture other stars in the same field of view for reference

          11. Once enough images are captured – the more the better – image calibration can commence. Please see the attached Appendix: Image Reduction Step by Step 

          12. Use of the IRAF software is to be used at this point (http://iraf.noao.edu/) which will create any data points required

          13. Data points are entered into an Excel spreadsheet. Scatter plots are preferred.   

The results of all this hard work is an Excel scatter plot that graphs out the radial velocities of the orbiting object (figure 18 ). The positive numbers on the y-axis indicate radial velocity toward us, and the negative numbers on the y-axis indicate radial velocity away from us. The x-axis is time. Notice the orbital period: measured 3.41 days here which is in good agreement with the published results.
 

Once the data has been reduced and analyzed, it is always a good idea to have an independent group evaluate the data. This data was sent to NOAO (National Optical Astronomy Observatory: http://www.noao.edu/) for analysis, with the results equal to the Spectrashift.com  group (figure 19). 
 

 

 

Figure 18: The result of numerous images of spectra from Tau Bootis.

Figure 19: This graph, the result of independent analysis from NOAO, shows almost identical results.

Back to Top | Back to Exoplanets

Final Test:

The final test as to the accuracy attained by an amateur group is to compare the results with the published data (figure 20):

The California & Carnegie Planet Search Team:

The Spectrashift.com Team:

 

 

Figure 20   Orbital Period: 3.312 days; Amplitude: 471 km/s

Orbital Period: 3.41 days; Amplitude: 420 km/s

While these results are not exact, it shows that a dedicated group of amateurs can yield results very similar to the professionals. So why are the numbers not exact? According to the Spectrashift.com  website, the initial series of data was out of phase 180%, and the mathematics was not tested accurate 99 times out of 100. Basically the errors were in the image processing and not in the techniques used to capture the data.

Back to Top | Back to Exoplanets

Transit Method – HD209458:

Star Name

HD 209458 or SAO 107623

Distance

47 parsecs

Apparent Magnitude

7.65

Spectral Class

G0

Metallicity

0.04

Planet Mass

0.62 time the mass of Jupiter

Orbital Distance

0.046 AU

Orbital duration

3.5239 days

Differential Magnitude7

0.0011

(Henry et al, 2000)

Greg Laughlin and Tim Castellano – founders of http://transitsearch.org  - have demonstrated that photometry to measure a stellar transit can be obtained with a 8” telescope, an entry-level CCD camera, and over the counter Astronomy software (not to mention of course clear skies). Specifically, an Meade 8” LX200 telescope (www.meade.com) armed with a Santa Barbara Instrument Group (www.sbig.com) ST-7 CCD camera and CCDSoft (www.bisque.com) software will yield very impressive results (figure 21).

Figure 21: This Meade 8” telescope and SBIG ST-7 camera can be purchased for around $7500.00. With it, it is possible to obtain professional quality light curves of a transiting exoplanet.

CCD cameras are very sensitive to changes in brightness of a star. With a properly reduced image, changes in luminosity of as little as 0.011 magnitudes are possible. The procedure of gathering photometric data is simple, but requires patience and skill with a telescope and CCD imaging software:

            1. Gather a target list. In this case, the target is HD 209458

            2. The Meade LX200 is computer controlled. Software Bisque makes a wonderful software package called TheSky – which is a computer planetarium and offers telescope control. Click on the desired star, and tell the software to move the telescope in position.

            3. Use CCDSoft – the included CCD camera control software when purchasing an SBIG camera– to begin capturing a series of 30 second images with the ST-7.

            4. Perform image reduction of all the images within CCDSoft.

            5. Use the imbedded photometry tools within CCDSoft to gather photometric points of the target star, as well as a few other stars that are in the same field.

            6. Input the target points into an Excel spreadsheet – a scatter plot is preferred.  

Since the interest is to evaluate changes in brightness, photometric calibration of accurate stellar magnitudes is not required. However, it is a good idea to make sure the surrounding stars do not exhibit the same changes in brightness.

When compared to a plot generated by professional equipment (figure 23), it is clear the amateur has much to offer (figure 22).

Figure 22: This data plot is the result of a Meade 8” telescope, an SBIG ST-7 CCD camera, and it’s provided CCD imaging software – CCDSoft. The scatter plot was created in Microsoft Excel.

Figure 23: This plot, released in the Astrophysical Journal Letters by a professional Astronomer shows an identical photometric curve of star HD 209458-b (Charbonneau et al, 1999).

 

Both professional and amateur plots reveal the orbital period of HD 209458 to be 3.52 days.  

Back to Top | Back to Exoplanets

Other methods:

The radial velocity and transit methods of planetary detection are the most common used techniques in the search of exoplanets, and the only methods used by amateurs; however, there are many more techniques available to the professional, as well as future space missions designed by NASA and the ESA for the sole purpose of improving our resolution capabilities.

            Astrometry – this method is used for long term accurate measure of the star apparent motion in the sky, and is used to detect planets greater than around 3 AU’s from the host star. This method does not seem very popular and is passed over in favor of the accurate measurements of the host stars radial velocity.

            Microlensing – this technique is used to in attempt to locate Dark Matter and black holes, but has been very successful in looking for orbiting planets. During a microlensing effect on a star, a brief but noticeable deviation of the light curve, as shown by figure 24  can be imaged. 

            Optical and Infrared Interferometry – an interferometer is a device that is used to combine the wave sources from two or more instruments and combing them to produce an image of much higher resolution. The Keck 1 and 2 in Hawaii, and the VLT Interferometer (VLTI) in Chile are the two current interferometers used to help detect exoplanets. The VLTI is in operation, but is also a work in progress. The resolution capabilities of this system hope to reach 10 micro-arcseconds (Mayor and Frei, 2003).

 

Figure 24: This photometric microlens record shows an Earth sized planet orbiting pulsar PSR B1257+12. The light source for this lens is a distant galactic bulge, and the deviation is a result of the orbiting planet being on either the front or backside of the pulsar resulting in a net increase of mass thereby producing a more powerful lens. A lens effect is the result of the deviation of light as a result of a massive object placed in between the source of light and the observer.

Here is a nice summary of detection methods – both present methods and proposed methods:

Back to Top | Back to Exoplanets

A Brief Window into the Future:

Telescopes in space offer tremendous benefits: there is no atmosphere to affect the quality of the images, and the already low temperature will ensure better noise control and infrared images require temperatures as low as possible. Three major space-based projects are in the design stage: NASA’s Kepler project, Terrestrial Planet Finder – or TPF and ESO’s DARWIN project (also known as the IRSI). The Kepler (figure 25) mission will utilize a very sensitive photometer to examine the transits of Earth-sized planets. With a planned launch date of 2007, 100,000 stars will be evaluated with the goal of obtaining a list of targets for the following Terrestrial Planet Finder mission.

 

Figure 25: The Kepler is still in the design stage. With a proposed launch date around 2007, its 37” photometry lens will study 100,000 stars to look for transits of Earth-sized planets.

The TPF will use two space-based, infrared sensitive telescopes in concert to create an infrared interferometer. The use of spectroscopy in the infrared will allow the study of cooler objects that orbit the stars instead of the star itself. The sensitivity of the TPF has a goal to view Earth-like atmospheres around Earth sized planets that orbit within the habitable zone. While this zone will be difficult to determine due to differences in stellar mass and temperature, the idea is this zone is approximately the Earth-Sun distance. While obtaining spectroscopic data on the various minor gaseous elements in an atmosphere will prove to be difficult, the goal is to at least obtain spectra of water, ozone, and carbon dioxide. Such elements in an atmosphere would be considered Earth-like (Mayor and Frei, 2003). The European Space Agency has their own project to look for exoplanets as well, called DARWIN. The goal of DARWIN is very much the same as the TPF: to look for Earth-like atmospheres around planets orbiting within the habitable zone. DARWIN has a target launch date around 2015, so we have some waiting to do. In the meantime, have a look at the proposed design of the DAWRIN interferometer (figure 26):

 

Figure 26: The DARWIN spaced-based interferometer will hopefully launch by around 2015. Its mission is to study the characteristics of nearby exoplanets, search for Earth-like atmospheres, and perform some “general” Astronomy. NASA’s TPF will use a very similar design.

More information on Kepler can be found here: http://discovery.nasa.gov/kepler.html  

More information on the TPF can be found here: http://planetquest.jpl.nasa.gov/TPF/tpf_index.html

More information on DARWIN can be found here: http://ast.star.rl.ac.uk/darwin/

Back to Top | Back to Exoplanets

Conclusion:

The study of exoplanets is ongoing. Continued advances in professional astronomy are allowing for increased sensitivity. The two most common and successful tools used by both amateurs and professionals are the measure of transit brightness and radial velocity. By collaborating with amateur astronomers, professional telescope time is preserved. The gathering of orbital data is vital to ensure repeatability with software analysis, and the amateur is poised to provide this important data. Most of all, we have shown that these two methods of planetary detection is possible, and that amateurs can also join the hunt. Greg Laughlin and Tim Castellano of http://transitsearch.org  are actively recruiting amateur astronomers that wish to participate. This website contains wonderful information on how to put together a telescope ready to capture transit data, and also coordinates target stars with participating members to avoid any overlap or missing data. With careful planning, it is possible to duplicate the methods used by the Spectrashift.com  group to gather data on radial velocity of exoplanets. While no professional group is seeking amateurs to provide spectroscopic data, this is sure to change as the contribution of the amateur have proven valuable for those involved in transit searches.

The hunt is on………

Back to Top | Back to Exoplanets

Recommended Internet Resources:

California & Carnegie Planet Search: http://exoplanets.org

Anglo-Australian Planet Search: http://www.aao.gov.au/local/www/cgt/planet/aat.html  

The European Southern Observatory VLTI: http://www.eso.org/projects/vlti/

NASA Origins of Solar Systems amateur project: http://origins.jpl.nasa.gov/index1.html  

The Geneva Extrasolar Planet Search: http://obswww.unige.ch/~udry/planet/planet.html

NASA Terrestrial Planet Finder: http://planetquest.jpl.nasa.gov/TPF/tpf_index.html

Advanced Fiber Optic Echelle Program: http://cfa-www.harvard.edu/afoe/index.html

Project STARE: http://www.hao.ucar.edu/public/research/stare/overview.html

Planet Homepage - Microlensing: http://planet.iap.fr/

Back to Top | Back to Exoplanets

References

Beatty, J. Kelly. Carolyn C. Petersen and Andrew Chaikin. The New Solar System 4th Edition. Cambridge University Press, 1999.

Bedding, Thomas et al. “Evidence for Solar-Like Oscillations in ß Hydri.” The Astrophysical Journal, 549: L105-L108, March 1, 2001.

Burnham, Robert Jr. Burnham’s Celestial Handbook – Volume Two. Dover Publications, Inc., New York, 1978.

Butler, Paul. “A precision Velocity Study of Photometrically Stable Stars in the Cepheid Instability Strip.” The Astrophysical Journal, 494: 342-365, February 10, 1998.

Butler, Paul, et al. “Attaining Doppler Precision of 3 m s -1.” Publications of the Astronomical Society of the Pacific, v 108: 500-509, June 1996.

Butler, Paul et al. “Planetary Companions to the Metal-Rich Stars BD -10o 3166 and HD 52265.” The Astrophysical Journal, 545: 504 – 511, December 10, 2000.

Charbonneau, David et al. “Detection of Planetary Transits Across a Sun-like Star.” Astrophysical Journal Letters, 23 November, 1999.

Fischer, Debra et al. “Planetary Companions to HD 12661, HD92788, and HD 38529 And Variations in Keplerian Residuals of Extrasolar Planets.” The Astrophysical Journal, 551: 1107 – 1118, April 20, 2001.

Freedman, Rodger and William Kaufmann. Universe 6th Edition. W.H. Freeman and Company, New York 2001.

Henry, Gregory et al. “A Transiting “51 Peg-Like” Planet.” The Astrophysical Journal, 529: L41 – L44, January 20, 2000.

Kitchin, C. R. Astrophysical Techniques 3rd Edition. Institute of Physics Publishing. Bristol, 1998.

Marcy, G. W. and Paul Butler. “Characteristics of Observed Extrasolar Planets.” The Tenth Cambridge Workshop on Cool Stars, Stellar Systems and the Sun. Cambridge, Massachusetts. July 16-20, 1997.

Mayor, Michael and Pierre-Yves Frei. New Worlds in the Cosmos. The Discovery of Exoplanets. Cambridge University Press, 2003.

Ostlie, Dale. Bradley Carroll. An Introduction to Modern Stellar Astrophysics. Addison-Wesley Publishing Company, Inc. Reading, Massachusetts, 1996.

Tonkin, Stephen. Practical Amateur Spectroscopy. Springer. London, 2003.

Back to Top | Back to Exoplanets

Image Credits:

Figure 1: http://www.star.ucl.ac.uk/~rhdt/diploma/lecture_1/contraction.jpg  

Figure 2: http://www.rc-astro.com/nebulae/m42_2004-01-27.htm

Figure 3: http://hubblesite.org/newscenter/newsdesk/archive/releases/1995/49/image/b

Figure 4: http://hubblesite.org/newscenter/newsdesk/archive/releases/1995/45/image/b  

Figure 5: http://hubblesite.org/newscenter/newsdesk/archive/releases/1995/45/image/b  

Figure 6: http://hubblesite.org/newscenter/newsdesk/archive/releases/1995/45/image/b  

Figure 7: http://www.mrao.cam.ac.uk/telescopes/coast/betel.html

Figure 8: http://hubblesite.org/newscenter/newsdesk/archive/releases/2003/02/image/b  

Figure 9: http://hubblesite.org/newscenter/newsdesk/archive/releases/1999/05/image/c  

Figure 10: http://hubblesite.org/newscenter/newsdesk/archive/releases/2000/02/  

Figure 11: http://www.gb.nrao.edu/~rmaddale/Education/Wvsta'98/200c.gif  

Figure 12: http://cfa-www.harvard.edu/afoe/doppler-shift.gif  

Figure 13: http://www.astrophys-assist.com/educate/solarobs/ses01p16.htm  

Figure 14: http://msowww.anu.edu.au/observing/74in/Echelle/ech_get_go_quick.html  

Figure 15: http://www.hao.ucar.edu/public/research/stare/overview.html

Figure 16: http://www.spectrashift.com/meade.jpg

Figure 17: http://www.spectrashift.com/spectro.html

Figure 18: http://www.spectrashift.com/tauboo.html

Figure 19: http://www.spectrashift.com/tauboo.html

Figure 20: http://exoplanets.org  

Figure 21: http://transitsearch.org

Figure 22: http://transitsearch.org

Figure 23: Charbonneau et al, 1999

Figure 24: http://www.nd.edu/~srhie/MPS/  

Figure 25: http://discovery.nasa.gov/kepler.html

Figure 26: http://ast.star.rl.ac.uk/darwin/pics/alcatel_ff_jul99.jpg


1 For analogy purpose only: Stars produce helium through hydrogen fusion, and can contain many other elements is the atmosphere.

2 A diffraction grating is a piece of glass that contains hundreds of evenly spaced grooves cut at an angle. These provide the same affect of viewing the spectra as a prism, but can be rotated to increase the viewing angle for improved resolution. The larger surface area also improves resolution over a prism.

3 1 AU is the Earth-Sun distance, or 93 million miles.

4 Not all objects in the Universe benefit from accurate measure of radial velocity. For example, such specific designs cannot be used to identify all of the variety of elements in a stars atmosphere.

5 Noise is the result of the CCD interpreting heat as a signal. While image reduction can filter out the noise, it’s best to start with an already cooled CCD camera.

6 The size of the pixels in analogous to film speed; 24 microns is very sensitive, but a lower number means higher resolution. In this case, we want the most light possible so a more sensitive CCD is required.

7 The differential magnitude is the level of magnitude change when the planet is at maximum transit


Back to Top | Back to Exoplanets

 

Search | Site Map | Appendix
©2004 - 2024 Astronomy Online. All rights reserved. Contact Us. Legal. Creative Commons License
The works within is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.