Saturday, August 31, 2013

Detecting Exoplanets Using Nanosatellites

ExoplanetSat is a proposed miniature space telescope that is designed to seek out Earth-like planets orbiting nearby bright Sun-like stars using the transit method. When a planet happens to pass in front of its host star, it dims its host star by a small amount which depends on the planet-to-star area ratio. ExoplanetSat is comprised of 3 CubeSat units joined together with total dimensions of just 10 cm × 10 cm × 34 cm. CubeSats are a class of nanosatellites developed at Stanford University and California Polytechnic State University to facilitate low-cost access to space. A single CubeSat unit measures 10 cm on a side and has a mass on the order of 1 kg. By adopting the standardized CubeSat form factor, ExoplanetSat can be built at a much lower cost than a larger or non-standardized configuration. The compactness of ExoplanetSat allows it to piggyback the launch of large spacecraft to orbit which further reduces cost.

 Figure 1: Artist’s impression of an Earth-like planet. Credit: Bryan Kolb

ExoplanetSat will monitor a single Sun-like star for the presence of transiting planets. If a transiting planet is detected, the star will become a high-priory target for follow-on observations. A transiting planet offers a unique opportunity for its atmospheric constituents to be characterized. During a transit event, some of the starlight passes through the planet’s atmosphere and picks up spectral features to create a planetary transmission spectrum. By studying the planetary transmission spectrum, the planet’s atmospheric constituents can be determined. Although this technique allows the atmospheres of Earth-like planets to be studied, it only works for stars that are bright enough. Hence, ExoplanetSat will target the nearest and brightest Sun-like stars. Once a single ExoplanetSat prototype has been flown successfully, multiple copies can be made and launched to allow for the monitoring of multiple target stars simultaneously for transiting planets.

The brightest Sun-like stars are too widely separated in the sky for a single telescope’s field-of-view to continuously monitor. Because of that, the ultimate goal is a fleet of ExoplanetSats, with each ExoplanetSat monitoring a single Sun-like star. ExoplanetSat will be place in a low-inclination low Earth orbit (LEO) at approximately 650 km altitude. It will observe its target star only during orbital night (when in Earth’s shadow) due to thermal, power and lighting constraints during orbital day. Since an Earth-like planet takes over 10 hours to transit a Sun-like star, the observational cadence from LEO (~ 30 minutes of orbital night followed by ~ 60 minutes of orbital day) is sufficient to detect a transit event. During orbital dawn, ExoplanetSat will reorient to point its solar panels toward the Sun. Likewise, during orbital dusk (just before orbital night); ExoplanetSat will slew to re-acquire its target star.

Figure 2: Baseline configuration of the ExoplanetSat.

ExoplanetSat uses an optical payload which consists of a commercially available single lens reflex (SLR) camera lens and a composite focal plane array made up of a single CCD detector surrounded by multiple CMOS detectors. The CCD performs the photometric function of accurately monitoring the brightness of its target star for transiting planets. The CMOS detectors track the centroids of surrounding “guide” stars and keep the telescope precisely pointed at the target star. To detect the tiny dimming caused by a transiting Earth-size planet, ExoplanetSat will need a photometric precision of better than 10 ppm. This will be a challenging goal to reach within the constraints of a CubeSat. If the 10 ppm photometric requirement cannot be met, then the science objective may be adjusted to focus on detecting larger planets or Earth-size planets around dimmer stars.

Smith, Matthew W. et al., “ExoplanetSat: detecting transiting exoplanets using a low-cost CubeSat platform”, Space Telescopes and Instrumentation 2010: Optical, Infrared, and Millimeter Wave

Friday, August 30, 2013

Uncovering Nearby Brown Dwarfs

The shift in all-sky surveys to longer wavelengths, from the optical, to the near-infrared and mid-infrared, have led to significant progress in the discovery of brown dwarfs with ever cooler temperatures. Brown dwarfs fall in four spectral classes (M, L, T and Y) with spectral class Y being the coolest. Since brown dwarfs cool as they age, the majority of brown dwarfs in the Sun’s neighbourhood with typical ages of several billion years are observed to fall under the T-type and Y-type spectral classes.

Artist’s rendition of a T6.5-type spectral class brown dwarf called 2MASSJ22282889-431026. Credit: NASA/JPL-Caltech

Nevertheless, a complete census of brown dwarfs in the Sun’s neighbourhood is far from complete. Kirkpatrick et al. (2012) found that there are currently about six times more stars than brown dwarfs within 8 pc of the Sun. The known proportion of brown dwarfs to stars is expected to increase with time as new discoveries are catalogued. In fact, the discovery of a pair of brown dwarfs located at just 2 pc from the Sun by Luhman (2013) shows that such an expectation is justifiable.

A recent search by Bihain et al. (2013) using data collected by NASA’s Wide-field Infrared Survey Explorer (WISE) discovered 3 new brown dwarfs in the Sun’s neighbourhood. One brown dwarf in particular, designated as WISE J0521+1025, is a T7.5-type spectral class brown dwarf located at a distance of just 5 pc away. The other 2 brown dwarfs are located slightly beyond 10 pc away. These nearby brown dwarfs were found by looking for high proper motion objects in the WISE dataset. This is because objects nearer to the Sun tend to have higher proper motions, just like how trees along the side of a road appear to move faster than distant mountains.

1. Kirkpatrick et al. (2012), “Further Defining Spectral Type “Y” and Exploring the Low-mass End of the Field Brown Dwarf Mass Function”, arXiv:1205.2122 [astro-ph.SR]
2. Luhman (2013), “Discovery of a Binary Brown Dwarf at 2 Parsecs from the Sun”, arXiv:1303.2401 [astro-ph.GA]
3. Bihain et al. (2013), “An overlooked brown dwarf neighbour (T7.5 at d~5pc) of the Sun and two additional T dwarfs at about 10pc”, arXiv:1307.2722 [astro-ph.SR]

Wednesday, August 28, 2013

Convection in Enceladus’ Ice Shell

Observations of Saturn’s moon Enceladus by NASA’s Cassini spacecraft have revealed the presence of jets of water vapour and ice particles emanating from warm tectonic ridges at the south pole of the moon. The heat required to power these jets is much larger than what may be produced by the decay of radioactive elements in the moon’s rocky core. Instead, the heating power is most likely produced by tidal heating resulting from the damping of Enceladus’ orbital eccentricity.

Convective processes occurring within Enceladus’ ice shell is believed to be causing the activity currently observed at the south pole. A study done by M. Behounkova et al. (2013) investigates the effect of tidal heating on the onset of convection in Enceladus’ ice shell. Convection in Enceladus’ ice shell can only occur if there is sufficient tidal heating and if the ice grains are smaller than a critical size. In the study, the amount of tidal heating depends on the orbital eccentricity of Enceladus and on the width of the internal liquid water reservoir at the boundary between Enceladus’ ice shell and rocky core. For this study, the internal liquid water reservoir is assumed to be centred under the south pole and reservoir widths of 120°, 180° and 360° are considered. Also, a variety of ice grain sizes are also considered.

The study indicates that for low tidal heating rates, convection only occurs for ice grain sizes smaller than 0.5 mm. However, it is unlikely that ice grain sizes smaller than 0.5 mm are present within Enceladus’ ice shell. Ice grains of somewhat larger sizes are more realistic. Since the minimum ice grain size needed to drive convection increases with the tidal heating rate, larger ice grain sizes will require a higher amount of tidal heating to drive convection.

In the presence of tidal heating, the width of the internal liquid water reservoir has a strong effect on the onset of convection. A considerable larger orbital eccentricity is required to generate sufficient tidal heating to drive convection if Enceladus has a small internal liquid water reservoir. For example, if the ice grain size is 1 mm, the eccentricity needed to trigger convection is 4 times current eccentricity for a reservoir width of < 120°, while it is only twice current eccentricity for a reservoir width of > 180°.

The present-day value of Enceladus’ orbital eccentricity is too small to generate enough tidal heating to account for the heat loss by thermal convection. This has led to the suggestion that Enceladus may have experienced a recent period of enhanced orbital eccentricity followed by a rapid damping of orbital eccentricity through tidal heating. Hence, Enceladus is expected to switch between conductive (low activity) and convective (high activity) states depending on its orbital configuration. The present-day Enceladus may be near the end of its convective state.

During periods of low activity, the orbital eccentricity of Enceladus is expected to gradually increase over timescales on the order of 100 million years. When the orbital eccentricity of Enceladus reaches 2 to 5 times the present-day value, tidal heating then becomes large enough to drive convection even for a maximum ice grain size of 1.5 mm. This period of high activity is expected to last for less than 10 million years as the orbital eccentricity of Enceladus damps out. As a result, high activity periods associated with convection and internal melting should be brief (~ 1 to 10 million years), followed by relatively long periods of low activity (~ 100 million years) during which convection is likely to cease.

M. Behounkova et al., “Impact of tidal heating on the onset of convection in Enceladus’s ice shell”, Icarus 226 (2013) 898-904

Saturday, August 24, 2013

Ice-Albedo Feedback on Terrestrial Planets

The ice-albedo feedback is expected to play an important role in determining the climates of many terrestrial planets. It is based on the positive feedback between decreasing surface temperatures, an increase of snow and ice cover, and an associated increase in the planet’s overall reflectivity, which then further decreases surface temperature. Basically, the ice-albedo feedback describes the possible runaway cooling of a planet’s surface. This can cause a terrestrial planet to be locked in a snowball state where ice and snow completely covers the planet from Pole to Pole.

Previous studies have shown that the strength of the ice-albedo feedback is reduced for terrestrial planets around M-dwarf stars. This is because M-dwarf stars emit mainly in the near-infrared where the snow/ice reflectivity is low. In comparison, our Sun emits primarily in the visible where the snow/ice reflectivity is high. A recent study by Paris et al. (2013) investigates the influence of a planet’s atmosphere on the ice-albedo feedback.

Artist’s impression of a frozen terrestrial planet with ice covering much of the planet. Credit: Scott Richard

The study shows that for a planet with a dense carbon dioxide atmosphere, the ice-albedo feedback is suppressed whereby the difference in the planet’s reflectivity between the ice and ice-free cases is strongly reduced. A terrestrial planet located towards the outer edge of the habitable zone around its host star can be expected to have a very dense atmosphere comprising up to several bars of carbon dioxide. As a result, a suppressed ice-albedo feedback can allow such a planet to remain habitable by keeping it from entering a snowball state via runaway cooling.

The type of star around which a terrestrial planet orbits also has a large effect on the planet’s overall reflectivity. For a planet with a dense carbon dioxide atmosphere, its overall reflectivity is found to be 2 to 3 times higher if it were orbiting around a Sun-like star instead of an M-dwarf star. This implies that terrestrial planets around M-dwarf stars can be 10 to 20 percent further away and still receive the same net stellar energy input into their atmospheres. It could mean that the habitable zone around an M-dwarf star is widened with respect to the habitable zone around a Sun-like star.

Trace gases such as water vapour, methane and ozone in a terrestrial planet’s atmosphere can affect the strength of the ice-albedo feedback. The presence of small amounts of water vapour and methane can weaken the ice-albedo feedback by several percent for planets around both Sun-like and M-dwarf stars. For planets with dense carbon dioxide atmospheres around Sun-like stars, the presence of significant amounts of ozone can strongly suppress the ice-albedo feedback.

Paris et al. (2013), “The dependence of the ice-albedo feedback on atmospheric properties”, arXiv:1308.0899 [astro-ph.EP]

Thursday, August 22, 2013

Star-Hugging Planets

A preliminary search for very short period transiting planets in the publicly available Kepler dataset has revealed 13 planet candidates with orbital periods ranging from 3.3 to 10 hours. Confirming whether these planet candidates are indeed planets will require additional follow-up observations and data analysis. If revealed to be true planets, they can only be small rocky planets since gas giant planets cannot survive at such small distances to their host stars.

Even with just a few Earth masses, these planet candidates can induce stellar radial velocity signatures (~ 10 m/s) that are large enough to be detectable by current ground-based instruments. The reason is because these planet candidates orbit incredibly close to their host stars and exert much larger gravitational tuggings on their host stars as compared to similar planets orbiting further out. For example, a 10 Earth mass planet in a 5 hour orbit around a solar mass star induces a radial velocity of 10 m/s.

The upcoming Transiting Exoplanet Survey Satellite (TESS) mission has a short observational cadence length of one minute. This makes TESS ideal for finding transiting planets with very short orbital periods that may be missed by Kepler’s 30 minutes observational cadence. If very short period planets are common, they should represent a significant fraction of the planets TESS finds.

Very short period planets are unlikely to form where they are currently observed to be. This is because at such close distances from their host stars, the temperatures are simply too high for rock material to condense. One possible way to get a very short period planet is by the inward migration of a gas giant planet. In this scenario, a planet that is located closer to its host star gets captured into resonance by an inward migrating gas giant planet and the end result is a very short period planet with a gas giant planet further out. This process can be tested by looking for outer, more massive planets that accompany very short period planets.

Another possible origin for very short period planets is by the tidal disruption of gas giant planets that come too close to their host stars. In this process, what is left of the tidally disrupted gas giant planet is its small rocky core in a very close-in orbit around its host star. The detection of these 13 very short period planet candidates adds yet another fascinating and unanticipated species to the growing menagerie of planetary systems.

Jackson et al. (2013), “A Survey for Very Short-Period Planets in the Kepler Data”, arXiv:1308.1379 [astro-ph.EP]

Sunday, August 18, 2013

Capturing Terrestrial-Sized Moons

Terrestrial-sized moons can exist around gas giant planets. This is especially interesting for gas giant planets that occupy the habitable zone around their host stars because terrestrial-sized moons around such planets can support Earth-like conditions. A terrestrial-sized moon requires a minimum mass of at least 0.1 to 0.2 Earth mass in order to hold on to an atmosphere for billions of years. This is about 4 to 5 times more massive than the largest moons in our Solar System and is comparable to the mass of Mars.

The largest moons in our Solar System presumably formed through accretion of material in the circumplanetary disks around Jupiter and Saturn. This poses a problem for forming a terrestrial-sized moon because there is not enough material and/or accretion efficiency in the circumplanetary disk around Jupiter or Saturn to form anything more massive than ~ 0.025 Earth mass. The moons Callisto and Ganymede around Jupiter, and Titan around Saturn show that a formation process via accretion in a circumplanetary disk is unlikely to form a moon larger than 1/10,000th the mass of its host planet.

Williams (2013) proposes that terrestrial-sized moons can exist around gas giant planets through a formation process known as binary-exchange capture. This process occurs when a binary comprising of two terrestrial-sized objects is tidally disrupted during a close encounter with a gas giant planet. One member of the binary gets ejected (escaping mass) while the other remains behind as a moon (captured mass) around the gas giant planet. Binary-exchange capture can only take place if the ratio of captured mass to escaping mass is not too large.

An example of a binary exchange capture involves a Jupiter-mass gas giant planet in orbit around a Sun-like star at a distance of 1 AU and a binary comprising of one object with the mass of Mars and one object with the mass of Mercury. If the binary approaches as close as ~ 5 planetary radii from the gas giant planet, the object with the mass of Mars remains behind as a moon (captured mass) while the object with the mass of Mercury gets ejected (escaping mass). In this scenario, the ratio of captured mass to escaping mass is ~ 2:1.

However, if the binary is comprised of one object with the mass of Earth and one object with the mass of Mars, then capture of the Earth-mass object as moon around the gas giant planet is not possible because the binary’s mass ratio of ~ 10:1 is simply too large. Nevertheless, if the distance of the Jupiter-mass gas giant planet is doubled to 2 AU, binary exchange capture of the Earth-mass object as moon around the gas giant planet becomes possible.

In our Solar System, Neptune’s moon Triton is believed to have been placed into orbit around Neptune via binary exchange capture. This is because Triton’s orbit around Neptune is retrograde, which means it orbits in the opposite direction to the planet’s spin and could not have accreted out from a circumplanetary disk around Neptune. Although Triton in nowhere as massive as a terrestrial-sized moon, its existence does supports binary exchange capture as a viable mechanism.

Williams (April 2013), “Capture of Terrestrial-Sized Moons by Gas Giant Planets”, Astrobiology, vol. 13, issue 4, pp. 315-323

Saturday, August 17, 2013

A Brown Dwarf’s Planet

Unlike normal stars, brown dwarfs are objects that are not massive enough to sustain hydrogen fusion in their cores. Disks of material are known to exists around young brown dwarfs and this have led to speculation that planetary systems similar to those around normal stars can also form around brown dwarfs. In fact, there have been several detections of planetary-mass companions around brown dwarfs. However, these brown dwarfs and their planetary-mass companions have large mass ratios (companion comparable in mass to the brown dwarf) and large companion-brown dwarf separations. For example, the planetary-mass companions around brown dwarfs 2MASS 1207-3932 and 2MASS 0441-2301 have large mass ratios of ~ 0.16 for 2MASS 1207-3932 and ~ 0.25 - 0.5 for 2MASS 0441-2301. In addition, they also have large separations of ~ 15 AU for 2MASS 0441-2301 and ~ 45 AU for 2MASS 1207-3932. This suggests a formation scenario similar to a binary system rather than a star-planet system.

A recent paper by Han et al. (2013) reports on the discovery of a low mass ratio planetary-mass object in a close orbit around a brown dwarf. This planetary-mass object was discovered via a gravitational microlensing event in 2012. Microlensing is the astronomical phenomenon wherein the brightness of a background star is magnified due to the bending of light by the gravity of a passing foreground object. In this case, the foreground object was the brown dwarf and its planetary-mass companion. This brown dwarf is designated OGLE-2012-BLG-0358L and it is estimated to be ~ 24 Jupiter masses while its planetary-mass companion is ~ 2 Jupiter masses. OGLE-2012-BLG-0358L and its planetary-mass companion have a small mass ratio of ~ 0.08 and a small separation of ~ 0.87 AU.

The mass ratio of OGLE-2012-BLG-0358L and its planetary-mass companion is a factor of 2 to 3 times smaller the mass ratios of 2MASS 1207-3932 and 2MASS 0441-2301. In particular, the separation of OGLE-2012-BLG-0358L and its planetary-mass companion is ~ 15 and ~ 40 times smaller than those of 2MASS 1207-3932 and 2MASS 0441-2301 respectively. As a result, the small mass ratio and small separation suggests that the planetary-mass companion around OGLE-2012-BLG-0358L may have formed from a disk of material around the brown dwarf, in a manner similar to the formation of planets around normal stars.

Han et al. (2013), “Microlensing Planet Around Brown-Dwarf”, arXiv:1307.6335 [astro-ph.EP]