IS THAT A PLANET ON MY IMAGE?!?! ------------------------------ John Pazmino NYSkies Astronomy Inc firstname.lastname@example.org 2007 February 8
Introduction ---------- In fall of 2006 the star epsilon Eridani came into higher attention from two news items. One was observation by the Hubble Space Telescope of its 'hot Jupiter' planet. This planet was announced in 2000 and studied since by many observatories. Hubble obtained additional astrometric data thru its fine guidance system, confirming many of the planet's orbital parameters. The other is the program for optically imaging this planet with Hubble and other telescopes in late 2007. The planet is approaching its periastron, the 'perihelion' of another star, in April 2007. Orbital geometry as seen from Earth favors a little later date for a prime chance to capture the planet by optical means. Epsilon Eridani has been a favorite star for home astronomers, and for space travel and space fiction enthusiasts, since the 1960s because of its proximity to the Sun, only 10-1/2 light years away, and its long suspicion of hosting planets and potentially human-level culture. The two current news items heightened this interest.
Epsilon Eridani ------------- Epsilon Eridani is a so-so star in Eridanus, a ways west of Rigel in Orion. It is near the threshold of easy vision under the average night sky of New York City. It comes into plain view when the sky darkens to reveal the other dimmer stars of Eridanus, Taurus and Cetus. This happens typically in winter, with Eridanus high in the sky. The name is routinely mispronounced, as are many other astronomy words of foreign origin. The constellation is 'eh-RIH-da-nuss' and the star itself is 'EPP-sih-lonn eh-RIH-da-nee'. In spite of the antiquity of the constellation, it’s still unclear which Eridanus it stands for. There are several in classical mythology, plus one as the earlier name for the Padua, or Po, river. Epsilon has no proper name, there being only a few named stars in all of Eridanus. Although the constellation is an overall dim one, its stars make patterns of curves that help to distinguish them from the random peppering of dim stars elsewhere. Epsilon is on the northern crest of one of these curves. Epsilon is the closest known planetary star, 10.5 light years away. There are nearer planetary star candidates that haven't yielded up any planets. Tau Ceti is a candidate a bit farther from us than epsilon, yet not showing any planets. If it does in the future, it, with epsilon, will be the closest pair of planetary stars, only 5.5 light years apart. With the discovery of its first planet in 2000, epsilon became the brightest of the planetary stars, at magnitude +3.7. Since then it was passed over by gamma Cephei in 2002 at magnitude +3.2. In 2006 epsilon dropped to third place with the confirmation of a planet at beta Geminorum, Pollux, at magnitude +1.2. Parameters of epsilon are presented under 'tau Ceti' in comparison with that other star.
Diurnal and annual motion ----------------------- Epsilon Eridani is only 9 degrees south of the celestial equator, making it observable for most of the year from both hemispheres. Its diurnal arc is not quite 11 hours from New York City latitude. The table here gives the rise/transit/set times, in EST, throughout the year. ------------------------------- | Rise | Trans | Set Date |Azm 102| Alt 40|Azm 258 -------+-------+-------+------- 31 Dec | 15:18 | 20:47 | 02:20 10 Jan | 14:38 | 20:08 | 01:41 20 Jan | 13:59 | 19:28 | 01:02 30 Jan | 13:20 | 18:49 | 00:23 09 Feb | 12:40 | 18:10 | 23:39 19 Feb | 12:01 | 17:30 | 23:00 01 Mar | 11:22 | 16:51 | 22:21 11 Mar | 10:42 | 16:12 | 21:41 21 Mar | 10:03 | 15:32 | 21:02 31 Mar | 09:24 | 14:53 | 20:23 10 Apr | 08:45 | 14:14 | 19:43 20 Apr | 08:05 | 13:35 | 19:04 30 Apr | 07:26 | 12:55 | 18:25 10 May | 06:47 | 12:16 | 17:45 20 May | 06:07 | 11:36 | 17:06 30 May | 05:28 | 10:57 | 16:27 09 Jun | 04:49 | 10:17 | 15:47 19 Jun | 04:09 | 09:39 | 15:08 29 Jun | 03:30 | 08:59 | 14:29 09 Jul | 02:51 | 08:20 | 13:50 19 Jul | 02:11 | 07:41 | 13:10 29 Jul | 01:32 | 07:01 | 12:31 08 Aug | 00:53 | 06:22 | 11:52 18 Aug | 00:13 | 05:43 | 11:12 28 Aug | 23:30 | 05:03 | 10:33 07 Sep | 22:51 | 04:24 | 09:54 17 Sep | 22:12 | 03:44 | 09:14 27 Sep | 21:32 | 03:05 | 08:35 07 Oct | 20:53 | 02:26 | 07:56 17 Oct | 20:13 | 01:47 | 07:16 27 Oct | 19:34 | 01:07 | 06:37 06 Nov | 18:55 | 00:28 | 05:58 16 Nov | 18:15 | 23:45 | 05:18 26 Nov | 17:36 | 23:06 | 04:39 06 Dec | 16:57 | 22:26 | 04:00 16 Dec | 16:17 | 21:46 | 03:20 26 Dec | 15:38 | 21:07 | 02:40 05 Jan | 14:58 | 20:27 | 02:00 -----------------------------------------
Project Ozma ---------- It is usually presumed that a culture capable of communicating by radio lives on a planet. This is not at all a requirement but is based on the lone example of intelligent life that we’ve got on Earth. Hence, SETI, as the search is abbreviated, is associated with the search for extrasolar planets. This latter search, in these or other words, has no abbreviation! The two searches remain separate in that planet hunters are happy to find ANY cold (compared to a star), round, orbiting body at another star with no regard to life residing on it. With a planet in hand, the SETI folk postulate the conditions of what life it can host. Epsilon Eridani broke into the news when Project Ozma, the first serious SETI, ran in 1960. Project Ozma, headed by Otto Struve and Frank Drake, used the new 25-meter radio dish at National Radio Astronomy Observatory. Struve and Drake figured that stars similar to our Sun would be the most likely ones to host planets and intelligent life. They also worked out that if there was a radio transmitter, like those in common use on Earth, within about 15 light years, we could hear them with existing radio telescopes. So, they listened to the two closest of the sun-like stars, epsilon Eridani and tau Ceti, both well within 15 light years. Ozma lasted only three months, found no extraterrestrial signals and triggered several false alarms. Never the less, epsilon Eridani, with tau Ceti, were on the space faring map for ever after.
Tau Ceti ------ It is hard to speak of epsilon Eridani without mentioning its stellar neighbor, tau Ceti. This star is also commonly mispronounced. It's 'taow SEH-tee' and the constellation is 'SEH-tuss' or 'SEE-tuss'. Tau Ceti and epsilon Eridani are neighbors, being only 5.5 light years apart. Other than chance proximity, the two stars have no relation to each other. Tau Ceti has a dust disc, like epsilon, found in 2004. It is from 10AU to ~60AU from the star with a central open zone. This implies a planetary mass to clear out this zone but none is evident so far. If, by some chance, there are civilizations at both stars, they would be the closest pair of interstellar peoples, just 5.5 light years apart. They could, using only methods now available on Earth, talk with each other by radio. The travel time for radio is 5.5 years, the same as for light, short enough for the peoples to interact on a timescale of decades. Both stars are surrounded by lots of red dwarfs more densely clumped together than in the solar vicinity. These could be way stations for physical travel between the stars The two stars are compared in this table
------------------------------------- parameter | eps Eri | tau Cet ---------------+----------+---------- Flamsteed | 18 Eri | 52 Ceti ) Henry Draper | 22049 | 10700 ) HIPPARCOS | 16537 | 8102 )-- names in various BSC, HR | 1084 | 509 ) catalogs of stars Smithsonian | 130564 | 147986 ) Bonner Durch'g | -09:697 | -16:295 ) spectrum | K2-V | G8-V eff temp | 5,100K | 5,670K luminosity | 0.28Sun | 0.59Sun age | ~700My | ~9,000My mass | 0.85Sun | 0.81Sun radius | 0.84Sun | 0.82Sun metallicity | 0.79Sun | 0.32Sun habitabkw zone | 0.53AU | 0.72AU rotation | 11day | 34day app magn | +3.73 | +3.49 abs magn | +6.19 | +5.68 Sun app magn | +2.32 | +2.58 other star mag | +1.81 | +2.32 RA (2000) | 03h 33m | 01h 44m dec (2000) | -09d 27m | -15d 56m distance | 10.5ly | 11.9ly radial vel | +16Km/s | -17Km/s ---------------+----------+----------
The 'metallicity' is the ratio of iron/hydrogen abundance in the star compared to that in the Sun. Because tau is a prior generation star, its initial material is reasonably laced with much less heavy elements than the Sun. The 'habitable zone' is the distance from the star where an Earth- like planet can sustain liquid water on its surface solely as a function of heating by the star. The zone is a broad one, the figure being a middle point within it. The 'Sun apparent magnitude' is the brightness of the Sun as seen from the star. Because both stars are quite less luminous than the Sun, the Sun is quite brighter in their sky than they are in ours. The 'other star magnitude' is the apparent magnitude of the one star as seen from the other. Because the two are only about 1/2 as far apart as they are from us, they appear much brighter to each other than they do to us.
Science fiction -------------- Epsilon Eridani is a favorite place name in science fiction. It is a locale in the Babylon 5 and Star Trek series. It is misnamed 'epsilon erandi’ [!] in the Space Precinct show. Novelists Isaac Asimov, Alistair Reynolds, Harry Turtledove and David Weber all have scenes related to epsilon Eridani. Games such as Battletech and Halo include epsilon Eridani as a locale. Tau Ceti enjoys a parallel favor in science fiction, once in a while with epsilon. One computer space adventure game, Tau Ceti, is named for the star. In many instances, the star is mispronounced as 'EPP-sih-lonn eh-rih-DAY-nee'. This is excusable being that Eridani, and Eridanus, do not comply with English rules of sounds. They are Latin words. A better pronunciation would come from continental values to the letters, with the 'a', a short, not long, vowel. The earliest reliable reference to the use of these two stars in fiction dates to the 1960s. I suspect that they came to the attention of authors through the Project Ozma hunt for extraterrestrial life, which started in 1960. However, I have no indication that the authors appreciate that epsilon now has a planet, or two, and that both stars have dust discs. There are other stars in fictional works, like Aldebaran, Sirius, and alpha Centauri. They seem to be chosen for the euphony of their names rather than awareness of the presence or absence of planets around them.
Dust disc ------- In 1998, a dust disc was found at epsilon, resembling an Edgeworth-Kuiper belt for the star. Its spectrum in the far infrared to sub-millimeter bands indicates it is made of cold icy dust. The disc is more of a torus tilted 25 degrees from our line of sight. Its inner diameter is ~35AU; outer is ~75AU; densest zone is ~60AU. The texture is lumpy, with slow orbital motion around epsilon. There is a vacant hole in the center out to about 35AU cleared of dense material. The first of epsilon's planets seems too close to the star to remove this inner region by itself. More planets are postulated to account for this cleared outer zone. The disc could be an example of what the outer solar system looked like when the Sun was younger. Its inclination is roughly the same as the orbit of the first planet. This supports the connection between dust discs and planet creation.
First planet ---------- Epsilon's first planet, epsilon Eridani b, was announced in 2000 August. The planet was first suspected by Peter van deKamp in the 1970s and Bruce Campbell in the 1980s but not confirmed. In 2006 October, the Hubble Space Telescope issued further proof of the planet by astrometry. The bulk of the data accumulated for confirming the planet is for radial velocity. This was complicated by the star's active chromosphere and corona, superimposing Doppler shifts from convection and explosive motions into its spectrum. As is typical for exoplanets, the 'discovery' is actually a realization after many years, about twenty for epsilon Eridani, of data collection. Such data can come from earlier unsuccessful planet hunts. Often the information was gathered for other purposes and then applied to a hunt for planets. Epsilon Eridani b is a 'hot Jupiter', the first kind of extrasolar planet found because it is massive enough to register its presence. A hot-Jupiter is a large, Jupiter-order or more, massive body in a close orbit. It is strongly heated by the star, or more so than Jupiter in the solar system. If the planet is anything like Jupiter in composition, it could have a boiled off atmosphere like Mercury, or only one of heavy molecules like Venus. The planet never crosses the star, so there is no hope of seeing atmospheric absorption lines in epsilon's spectrum. The orbit carries the planet in 2007 April thru periastron. Nominally it is brightest then, reflecting the most of its star's illumination to us. It is also closest to the star then, deep within the glare of the star's radiation. It so happens that by waiting a few months, to November and December of 2007 the star is still bright from reflection but is removed to a wider angular separation from the star. It is then that the push for the optical imaging will occur.
Second planet ----------- A second planet, epsilon Eridani c, was announced in October 2002. This was the result of studying the motion of the clumps in the dust disc and modeling a planet that could cause them. This planet is one possible means of creating the vacant zone of the torus, in conjunction with the positively known first planet. Planet c is not yet positively observed. Its location and motion are inferred from its effect on the dust disc. Besides planet c, other planets are hypothesized, up to five or six, to fill out the epsilon Eridani system. This table gives parameters for the two planets based on information as at fall of 2006.
------------------------------------- parameter | eps Eri b | eps Eri c -------------+-----------+------------ Announced | 2000 Aug | 2002 Oct M*sin(i) | 0.87Mj | 0.1Mj Inclination | 30.1deg | -- ) Asc'g Node | 254deg | -- ) Arg Periastr | 47deg | -- )-- not yet determined Periastron | 2007 Apr | -- ) Mass | 1.55Mj | -- ) Period | 6.85y | 280y Seminaj Axis | 3.39AU | 40AU Excentricity | 0.702 | 0.3 ---------------------------------
In the usual planetary star, the inclination of the orbit is not known. The calculated mass is a dilution of the true mass, leading to some mistakes in the popular literature. The 'mass', unless specifically qualified, is the (true mass)*sin(inclination). One way to caution about the cited mass is to note that it is a lower limit or minimum value. For planet b, the inclination was obtained from combining the proper motion and radial velocity data to map the orbit in 3D space. Hence, a true mass is available for it. Planet c still is uncertain, with no inclination to hand. Only the M*sin(i) can be cited.
Extrasolar planets ---------------- The notion of planets at other stars is hardly new. It started in the early 1600s with the realization that stars must be so far away they can not shine by reflected sunlight. They must be whole other suns. Plausibly, these other suns have their own sets of planets? This was not a pretty idea at the time, resulting in some nasty penalties to those who entertained it. In the mid 1700s the variation of light sent out by certain stars, notably beta Persei and beta Lyrae, were explained by planets at them. The planet, large and dark, obscured the star when it passed in front to eclipse it. In the 19th century, the companion was proved to be a real star, only much dimmer than the main star. This was the discovery of eclipsing binaries, which later allowed the first determination of the sizes of stars. In the 19th century the subtle periodic wander of proper motion in Sirius and Procyon proclaimed the existence of planets. Within a decade or so, these planets, too, revealed themselves as stars, although of a tinier, weaker kind. They were the first white dwarf stars. In spite of the centuries-old concept of planets at other stars, the name of such bodies is still not settled. They are called 'exoplanets' or 'extrasolar planets' or 'stellar planets' with no particular prejudice. As yet, there is no generally used name for a set of planets at another star, analogous to 'solar system'. 'Stellar system' is too vague; 'extrasolar system' and 'exoplanetary system' haven't caught on. By the 20th century, the extrasolar planet idea waned; leaving the solar system as a rare, if not unique, feature of the universe. After World War II, new understanding about star creation allowed for the casual formation of planets. In the protostellar nebula, there could be tiny pieces, way too small to become stars, which collect near stars and take up orbit around them. There could be, hence, a hell of a lot of planets out there. But, where are they? Otto Struve in 1952 made the first modern proposal to look for exoplanets. He figured that a Jupiter-mass planet at a solar-mass star would cause a Doppler shift of about 200 meter/second. This was at the threshold of spectrometry in the 1950s. He allowed that as skills and arts of spectrometry advanced, such small Doppler shifts could be confidently measured. In this same era, Peter van de Kamp claimed to discern planetary disturbance in the proper motion of Barnard's star. The amount of wiggle was within the noise of atmospheric and chemophotographic distortion of the images. Barnard's star is still a candidate planetary star with no planets found to date.
Obstacles ------- The real impediment against finding planets at other stars was instrumental, not theoretical. The central star is many orders more brilliant than the planet. The planet would be completely swamped out in the optical noise spewed onto it by the star. This was the situation until the 1970s when astronomy was still tied to passive optics, turbulent atmosphere, crude computers, and chemophotography. There could never be an image pure and ideal enough to let the minuscule pinpoint of a planet resolve from the dazzling dot of its star. Since the 1970s new tools and techniques were perfected to the point where in 2000 it was within sight that some day real soon a regular photograph would be captured of an extrasolar planet. These tools include digital and high-efficiency imaging, adaptive telescope optics, space observatories like the Hubble Space Telescope and super strength computers. The discussion here of planet finding methods is merely an overview. Space agencies, mainly NASA and ESA, are preparing probes to hunt for planets. These include Kepler, Terrestrial Planet Finder, Space Interferometry Mission, and Darwin. One probe, COROT from France, is the first planetary star explorer actually commissioned. It was launched in December 2006 and cleared for operation in January 2007. They enjoy absence of atmospheric distortion, long-duration monitoring, full-spectrum access, and optically perfect imaging. In spite of the widely accepted theme of planetary stars, there is no common acronym for planet hunting, comparable to SETI. In the next sections are brief summaries of the various methods, now or planned, for seeking exoplanets.
Planet transits ------------- An opaque planet, a body an order or two smaller than the star, would obstruct only a minute portion of the star's disc. Ordinary photometry would miss the vanishingly tiny change in brightness. If you saw the transit of Venus in 2004 or the transit of Mercury in 2003 or 2006, you saw just how SMALL the planet is compared to the Sun. This method, also called photometry, requires a planet with an edge on orbit to maximize the dip in brightness. Today, photometry, especially from space and away from the scintillation of the air, can capture such minute alteration in brightness. A few planets have been found by this method so far. Instruments and computers once in the realm of only large observatories are now on hand for home astronomers. With these, home observations of transiting planets are feasible. Some almanacs, computer programs, and websites have timetables of exoplanet transits exacta mente for home astronomers to watch.
Proper motion ----------- The proper motion method looks for the wiggle of a star as it proceeds across the celestial sphere in its proper motion. The wander is due to the orbiting planet. This method, also called astrometry, is severely limited by air distortion. The star image is a blob, not a point, or irregular shape on the photographic plate. Any displacement of motion from a planet is lost within this blob. Another obstacle was the motion, often unknown, of the surrounding stars used as benchmarks for assaying the motion of the target star. It often is a surprise to many new astronomers that until the late 20th century we had only crappy knowledge of the 3D position and motion of the stars as a whole. Space observatories and ground ones with adaptive optics can track the swing in the proper motion of stars for planet hunting. So far, no planet was discovered by astrometry. Astrometry is used to study planets already found by other means.
Pulsar frequency -------------- The frequency shift of pulses from a pulsar as it wiggles to and from us can reveal the orbital motion of a planet. The shift is in both the frequency of the individual pulses and the arrival rate of the pulses. This allows for a very sensitive double-check on the observations. Pulsar observations are mostly done in the radio part of the spectrum, where electronics traditionally was far more matured than the handy light waves. The Doppler shift is easily detected in the meter/second range with high-dispersion electronic spectrometers and atomic clocks. In fact, the first exoplanets were found at pulsars. The discovery of planets at pulsars forced a rethink of the star lifecycle. A pulsar is the product of a supernova. Wouldn't such a violent explosion destroy a planet? Apparently not. Were the planets formed after the supernova died down? Maybe.
Radial velocity ------------- A Doppler, or radial velocity, oscillation of a star could mean it hosts a planet. Unlike for a spectrometric binary, there is only one spectrum and the shift of line frequency is in the meter/second range. Both features indicate a dark low-mass orb at the star. Planetary Doppler measurements are best taken on stars of spectral types F, G and K. They have clean spectra with sharp easily-measured lines. These stars are similar to the Sun and probably went through a planet-creation period like the Sun. Hence, these stars are the most studied by the Doppler method as better candidates for planets. M-type stars have complex spectra with lines and bands from molecules. They also can have convection in the photosphere to scramble the line frequencies. O, B, and A stars seem to be poor candidates for planets because their intense heat, ultraviolet and X- ray radiation could have melted them away. Since the first confirmed exoplanet in 1995, the spectrometric method yielded some 85% of discoveries.
Magnetic field ------------ Many of the exoplanets are close to the star, well within one Earth orbit radius away. It is possible that, if the star has a magnetic field, the planet interacts with the field in a rhythmic manner. A few stars, like epsilon Eridani, have indirect evidence of a magnetic field, but until 2006 no planetary star showed one directly. In 2006 tau Bootis was seen to have a cyclic change in magnetism in step with the period of revolution of its planet. This magnetic field method is not yet a prime way to find planets. For tau Bootis, it merely gathered additional information about a known extrasolar planet. It could be an emerging technique in the near future.
Biological activity ----------------- This is a very long shot. If the planet has life similar to Earth forests, the star's spectrum could have lines from the biological processes of that life. Forests are the suggested life form because they cover an extensive portion of the Earth, are a relatively homogeneous life structure, and carry on pretty much the same biochemical processes. Since forests can exist only in moderate temperatures, their spectral features can not come from the very star. They are produced in a cool body near it, a planet. Given the tiny size of any reasonable planet, and the low chances of it having a substantial cover of Earthlike forest, the biological spectral lines will be vanishingly weak and likely smothered by regular stellar spectral features. Yet, some work is under way to develop spectrometers and select targets to try this method.
Atmospheric gases ---------------- A potential new trick, still under study, is to catch in the star spectrum the absorption likes of a transiting planet's atmosphere. Its gases will consist of molecules that can survive only in a low temperature compared to the star. They must reside on a cool body, a planet. The hope is that the atmosphere, especially for a close in, well-heated planet, will be large and dense enough to produce its lines in the star spectrum. In a couple instances, such atmosphere has been detected after the planet was already known, but as yet no planet was first found by this method.
Blackbody emission ---------------- This technique hasn't yet been put into routine use but is considered for projects under construction or in planning. The idea is that a planet has a blackbody temperature thousands of degrees cooler than the star. Its blackbody emission curve peaks in the far infrared with only a minor portion in the visual band. The star peaks in the middle of the visual band with a far lesser portion in the infrared. By imaging the star through narrow-pass filters centered on the expected peak wavelength for the planet, the star's immensity of visual radiation is blocked. We increase the planet/star illumination ratio, hopefully to let the planet register in the image.
Diffraction --------- Optical systems are routinely designed to minimize diffraction to concentrate as much of the photons into the central Airy disc of the image. A reverse think exploits diffraction to render the star much weaker yet leave the planet image only modestly harmed. In this idea the optics are purposefully made to diffract the star's light to remove its Airy disc! It's as if there was no star to interfere with the planet. This effect is biased along the optical axis, where the star is placed during the hunt for planets. The planet, off axis, suffers far less diffraction, leaving hopefully enough of a firm image to detect. Experiments using the diffraction method are under construction and test.
Photography --------- The holy grail of planet hunters is a genuine ordinary 'picture' of an exoplanet; one that will be the TIME magazine 'picture of the year'. So far, efforts to image an exoplanet, even one well known, have failed. It's that immense disparity of brilliance between the planet and star that killed off all attempts to date. Progress is underway to perfect the instruments and processing of image data. In any month real soon, we will have in hand that once and forever view of a world beyond the solar system. Epsilon Eridani's inner planet in 2007 April rounds its periastron for the first time since discovery. Near periastron the planet receives, and then reflects, the most starlight and should be brightest against the star. It is also closest to the star, so there is the antagonism of angular separation and brightness. The orbit geometry favors waiting until November or December of 2007, well after periastron, to make the shot. Only the best equipped observatories can even try for the shot. Hubble Space Telescope is the runaway favored scope for its optically perfect imaging and superior image processing. Others on the ground include Keck, South African Large Telescope, Very Large Telescope, Gemini and Subaru. The trick is to suppress the starlight enough to let the feeble planet shine through and register on the detector. Some astronomers argue that an electronic image is not really a 'photograph' in the traditional sense. It isn't because it is a synthetic composition made from numerical data; not any more than any other astronomy 'photograph' that has been captured in the last decade or so. Observatories by the late 1990s have migrated out of chemophotography to digiphotography. It gets worse. Look at the TIME magazine issue with the 'man of the year' picture. Is it a ‘photograph’? Or a reconstituted set of numbers?