Thursday, July 31, 2014

Powering an Ultra-Luminous Galaxy Cluster

In some galaxy clusters, the central intracluster gas can become dense enough to cool radiatively within the cluster’s lifetime. This can drive a continuous flow of cooling gas plunging towards the cluster’s centre. The paucity of such cooling flows suggests that in galaxy clusters where such a process might occur, some form of astrophysical feedback kicks in to prevent the development of a runaway cooling flow. A study by M. McDonald et al. (2012) of the Phoenix Cluster (SPT-CLJ2344-4243) reveals it to be one of the most massive galaxy clusters known. The Phoenix Cluster lies at a distance of almost 6 billion light years, and its mass is estimated to be ~2500 trillion times the Sun’s mass. For comparison, the Milky Way, with its few hundred billion stars, is estimated to be only ~1.5 trillion times the Sun’s mass. The Phoenix Cluster also has ~2 trillion times the Sun’s luminosity.

Artist’s illustration of the Phoenix Cluster, showing the strong flow of cooling gas sinking towards the central galaxy. Image credit: Chandra X-Ray Observatory.

Observations of the Phoenix Cluster by a number of telescopes show the presence of an extremely strong cooling flow bound for the cluster’s centre. The flow rate is estimated to be ~4000 solar masses per year. With such a tremendous inflow of material, the central galaxy of the Phoenix Cluster is undergoing a massive starburst episode, churning out new stars at a remarkable rate of ~750 solar masses per year. For comparison, the present-day star formation rate in the Milky Way is roughly one solar mass per year. The high inflow rate and the high star formation rate suggest that the astrophysical feedback to prevent the formation of a runaway cooling flow has yet to be fully established in the Phoenix Cluster.

At the heat of the central galaxy in the Phoenix Cluster lies a supermassive black hole that is estimated to have a mass of about 20 billion times the mass of the Sun. The supermassive black hole is also accreting material at a prodigious rate of ~60 solar masses per year, or roughly one Earth mass every two seconds. The strong cooling flow in the Phoenix Cluster is believed to be short-lived in contrast to long periods of strong astrophysical feedback; else both the central galaxy and its supermassive black hole would become too massive.

M. McDonald et al. (2012), “A Massive, Cooling-Flow-Induced Starburst in the Core of a Highly Luminous Galaxy Cluster”, arXiv:1208.2962 [astro-ph.CO]

Wednesday, July 30, 2014

Largest Mass of Water in the Known Universe

In July 2011, a team of astronomers reported on the discovery of the largest reservoir of water ever detected in the universe in a distant quasar identified as APM 08279+5255. A quasar is an extremely luminous object that is powered by a supermassive black hole accreting material at a stupendous rate. APM 08279+5255 sits in the core of a giant elliptical galaxy that is located at a distance of 12 billion light years, near the “edge” of the known universe. What this means is that light observed from APM 08279+5255 left it 12 billion years ago, at a time when the universe is still relatively young since the universe itself is only 13.8 billion years old.

Figure 1: Artist’s impression of a quasar, similar to APM 08279+5255. Image credit: NASA/ESA.

APM 08279+5255 is incredibly luminous. Its radiant power is estimated to be roughly equivalent to a thousand trillion Suns. At its heart is a supermassive black hole whose mass is estimated to be a whopping ~20 billion times the Sun’s mass. Observations of APM 08279+5255 were performed in the millimetre waveband using the Z-Spec instrument at the Caltech Submillimeter Observatory (CSO) and the Combined Array for Research Millimetre-Wave Astronomy (CARMA). The observations revealed an enormous mass of water vapour swirling around this cosmic monstrosity. The mass of water vapour is distributed around the supermassive black hole in a gaseous region spanning a few hundred light years across. Measurements show that the total mass of water vapour around the quasar is at least ~100,000 times the mass of the Sun. That is over ~100 trillion times the amount of water in Earth’s oceans.

In a typical galaxy like the Milky Way, most of the water is frozen as ice. In the case of APM 08279+5255, the immense amount of energy being put out by the quasar keeps the water in its gaseous phase. Water-ice sublimates into water vapour when it is heated above ~100 K. In fact, the water vapour in APM 08279+5255 is observed to have temperatures ranging from 100 to 650 K. The energised mass of water vapour is continuously cooling; generating a total observed cooling luminosity of at least ~6.5 billion times the Sun’s luminosity. This discovery shows that water is common throughout the universe and it can occur in ginormous masses even in the early universe.

Figure 2: Comparison of the water spectrum in Mrk 231 with that in APM 08279+5255, as measured with Z-Spec. Bradford et al. (2011).

- Bradford et al. (2011), “The Water Vapor Spectrum of APM 08279+5255: X-Ray Heating and Infrared Pumping over Hundreds of Parsecs”, arXiv:1106.4301 [astro-ph.CO]
- Riechers et al. (2008), “Imaging the Molecular Gas in a z=3.9 Quasar Host Galaxy at 0.3" Resolution: A Central, Sub-Kiloparsec Scale Star Formation Reservoir in APM 08279+5255”, arXiv:0809.0754 [astro-ph]

Tuesday, July 29, 2014

Observations of a Low-Gravity Brown Dwarf

Brown dwarfs occupy the mass range between the most massive planets and the least massive stars. They are not massive enough to sustain hydrogen fusion in their cores and so they cool gradually with time. As a brown dwarf cools, it contracts, evolving from a low to a high “surface” gravity. The word “surface” is shown in quotation because a brown dwarf does not have a solid surface. Instead, a brown dwarf’s “surface” simply refers to its observable photosphere. Brown dwarfs are gaseous throughout. They are primarily composed of hydrogen and helium, with trace amounts of heavier elements.

Figure 1: Artist impression of a brown dwarf that is still glowing red-hot from heat acquired during its formation.

For a young brown dwarf that has yet to cool and contract to its final radius, it can display low-gravity features that can be identified from observations in the near-infrared waveband. This is because the low-gravity atmosphere of a young brown dwarf drives the formation of thicker than normal clouds in the brown dwarf’s photosphere. The thicker clouds give rise to a redder near-infrared spectrum because shorter wavelength light (i.e. bluer light) is attenuated and scattered by clouds more than longer wavelength light (i.e. redder light). As a result, a brown dwarf with a redder near-infrared spectrum can signify its youthfulness.

Gagné et al. (2014) present the discovery of SIMP J2154-1055 - an L4 spectral type brown dwarf displaying signs of low-gravity in its near-infrared spectrum. SIMP J2154-1055 has a redder near-infrared spectrum compared with other known L4 brown dwarfs. This is consistent with the presence of thicker than normal clouds in a low-gravity photosphere. SIMP J2154-1055 has a good probability of being part of the Argus Association, a loose group of stars with similar ages. If it is indeed a member of the Argus Association, SIMP J2154-1055 would be roughly 30 to 50 million years old and its mass would be ~10 times the mass of Jupiter, indicating that SIMP J2154-1055 is a relatively young and low-mass brown dwarf.

Figure 2: Near-infrared spectrum of SIMP J2154-1055 compared with other known L4 brown dwarfs. SIMP J2154-1055 is the reddest L4 brown dwarf yet identified. All spectra are normalized to their median in the 1.27 to 1.33 μm range. Gagné et al. (2014).

Gagné et al. (2014), “SIMP J2154-1055: A New Low-Gravity L4β Brown Dwarf Candidate Member of the Argus Association”, arXiv:1407.5344 [astro-ph.SR]

Monday, July 28, 2014

Discovery of a Stellar Behemoth in Westerhout 49

Very massive stars with masses exceeding ~100 times the Sun’s mass are incredibly rare. Nevertheless, these stars have strong influence on their environment through their powerful winds and prodigious amounts of ionizing radiation. The existence of very massive stars with reported masses of up to ~300 times the Sun’s mass indicate that there may not be a ‘real’ upper mass limit for very massive stars. These very massive stars are found in massive star-forming clusters such as NGC 3603, the Arches cluster and R136 in the Large Magellanic Cloud.

Artist’s impression of the Sun in comparison to R136a1 - a very massive star with an estimated ~300 times the Sun’s mass at birth. Image credit: ESO/M. Kornmesser.

A study by Shiwei Wu et al. (2014) presents the spectroscopic identification of a very massive star in the heart of the star-forming region Westerhout 49, hereafter, referred to as W49. The central cluster of W49 consists of dozens of massive OB-type stars. Intervening interstellar dust greatly obscures the massive stars in W49. As a result, these stars can only be observed in the near-infrared waveband. Referred to as W49 nr1, this very massive star resides in the heart of W49. It was spectroscopically identified using instruments on three telescopes - the Very Large Telescope (VLT) at Paranal in Chile, the New Technology Telescope (NTT) at La Silla in Chile, and the Large Binocular Telescope (LBT) at Mount Graham in Arizona.

W49 nr1 is the brightest star in the central cluster of W49 and it is classified as an O2-3.5If* star. Its estimated effective temperature is between 40,000 and 50,000 K, and its estimated luminosity is between 1.7 million and 3.1 million times the Sun’s luminosity. W49 nr1is more luminous than tens of billions of the coolest red dwarf stars put together. Stellar evolutionary models of W49nr1 suggest an initial mass of 100 to 180 times the Sun’s mass. If more variations are included in the models, the initial mass range is 90 to 250 times the Sun’s mass. Very massive stars live fast and die young. W49 nr1 is believed to be no more than 3 million years old.

Shiwei Wu et al. (2014), “The Discovery of a Very Massive Star in W49”, arXiv:1407.4804 [astro-ph.SR]

Sunday, July 27, 2014

An LBV Masquerading as a Cool Hypergiant

The search for supernovae (plural for supernova) has led to the discovery of a population of “supernova impostors”. These outbursts appear like supernovae, but exhibit much lower luminosities and ejecta velocities. A study by Mauerhan et al. (2014) presents observations of a supernova impostor identified as SN Hunt 248. The outburst associated with SN Hunt 248 was observed in May to June 2014 and it occurred in two stages. Between May 21and June 3, the source brightened slowly. On June 4, it began to brighten rapidly, reaching a peak on June 16. The source then plateaued for ~10 days at peak brightness before fading away.

Analysis of the photometric and spectroscopic data indicates that SN Hunt 248 is consistent with an outburst from a massive star. Archival images from the Hubble Space Telescope between 1997 and 2005 reveal that the precursor star is a cool hypergiant with ~400,000 times the Sun’s luminosity and ~32 times the Sun’s mass. SN Hunt 248 is believed to be the first outburst observed from a cool hypergiant that is similar to the giant eruptions typical for luminous blue variable stars (LBVs).

Figure 1: Artist’s impression of a hypergiant.

Figure 2: Light curve of SN Hunt 248 compared with other supernova impostors, including SN 1997bs, SN 2006jc, SN 2002bu, SN 2009ip and SN 2008S. The vertical axis indicates the absolute magnitude and the horizontal axis indicates the number of days relative to peak brightness. Mauerhan et al. (2014).

The slow rise in brightness followed by an episode of rapid brightening indicates that something sudden took place. LBVs and cool hypergiants are known to have strong winds that launch stellar material off them, causing them to be surrounded by circumstellar material of their own. In the case for SN Hunt 248, the initial phase of slow brightening is from the initial outburst. Subsequently, ejecta from the initial outburst collide into the circumstellar material around the cool hypergiant, resulting in the sudden conversion of kinetic energy into radiation. This explains the rapid brightening following the initial phase of slow brightening. Rough estimates suggest an ejecta mass of ~1 times the Sun’s mass and an ejecta velocity of ~1000 km/s.

In fact, the precursor star of SN Hunt 248 is probably a LBV masquerading as a cool hypergiant. This is because there seems to be an absence of LBVs with effective temperatures in the range between 15,000 K to 21,000 K (Figure 3). It has been proposed that at these temperatures, the stellar winds emanating from LBVs become considerably denser. The density can be high enough to produce an opaque pseudo-photosphere around the LBV, allowing the LBV to masquerade as a cool hypergiant with an effective temperature of less than ~8,500 K. If such a connection between LBVs and cool hypergiants is true, then one might expect some cool hypergiants to also exhibit giant outbursts like LBVs do.

Figure 3: Hertzsprung-Russell (HR) diagram for LBVs and related stars, including SN Hunt 248 (purple square). The diagonal and vertical grey strips illustrate the regions of the S Doradus instability strip and the minimum temperature strip for classical LBVs near visual maximum. Mauerhan et al. (2014).

Mauerhan et al. (2014), “SN Hunt 248: a super-Eddington outburst from a massive cool hypergiant”, arXiv:1407.4681 [astro-ph.SR]

Saturday, July 26, 2014

Signature of a Giant Planet’s Rocky Core

Stars in binary systems generally have identical chemical compositions since they formed from the same natal cloud of material. Nevertheless, small differences in chemical composition can exist between a pair of stars in a binary system and one explanation is the process of planet formation. When planets form around a star, it can cause the star to be slightly depleted in heavy elements (i.e. elements heavier than hydrogen and helium) compared to its companion star.

Observations of 16 Cygni, a binary system comprised of two stars 16 Cygni A and 16 Cygni B (hereafter components A and B), reveal that component B has a giant planet with at least 1.5 times Jupiter’s mass. The giant planet, identified as 16 Cygni Bb, orbits its host star in a highly-eccentric 800-day orbit. At its minimum and maximum distances from its host star, the giant planet receives, respectively, 4.4 and 0.16 times the amount of insolation Earth gets from the Sun. Being a giant planet, 16 Cygni Bb is composed primarily of hydrogen and helium, much like Jupiter.

Figure 1: Artist’s impression of a giant planet with a system of moons around it.

Figure 2: Differences in heavy element abundance between components A and B versus condensation temperature. The dashed line is the average of the volatiles and the solid line is the average of the refractories. The dot dashed line is the mean trend for 11 Sun-like stars compared to the Sun. Maia et al. (2014).

A study by Maia et al. (2014) show a small difference in the abundance of heavy elements (i.e. metallicity) between components A and B of 16 Cygni. Component A has an overall metallicity (i.e. abundance of heavy elements) that is 0.047 ± 0.005 dex higher than the metallicity of component B. The abundance differences range from 0.03 dex for volatiles (i.e. elements such as carbon and oxygen), and up to 0.06 dex for refractories (i.e. elements such as iron, vanadium and magnesium). The lower abundance of heavy elements in component B is likely due to the formation of the giant planet that is presently in orbit around it, where the “missing” heavy elements were used to form the giant planet.

The higher deficiency in refractories compared to volatiles in component B means that the giant planet, 16 Cygni Bb, has a corresponding excess of refractories. This suggests that 16 Cygni Bb formed by the core accretion mechanism where an initial rocky core, comprised primarily of refractories, becomes massive enough to start accreting hydrogen, helium and other volatiles to form a giant planet. Estimates of the initial rocky core of 16 Cygni Bb places it at around 1.5 to 6 times the mass of Earth, consistent with Jupiter’s core mass. These findings validate the core accretion model for the formation of 16 Cygni Bb and offer yet another means to examine the relationship between a star and its planet.

Maia et al. (2014), “High precision abundances in the 16 Cyg binary system: a signature of the rocky core in the giant planet”, arXiv:1407.4132 [astro-ph.SR]

Friday, July 25, 2014

Globular Clusters and Dark Satellite Galaxies

Globular clusters are dense spherical collections of stars. Every large galaxy, such as the Milky Way, contains a system of globular clusters. Observations of globular clusters show that they do not contain gravitationally bound dark matter. Most of the matter in the universe is in the form of dark matter. Dark matter neither emits nor absorbs light, and its presence can only be inferred from its gravitational effects on normal matter and radiation. Nevertheless, the existence of dark matter is important because it provides the gravitational framework for normal matter to come together to form galaxies and clusters of galaxies. As a result, it remains a challenge to explain how normal matter could gravitate so tightly together to form globular clusters.

The globular cluster NGC 1806 located within the Large Magellanic Cloud as observed by the Hubble Space Telescope. Image credit: ESA/Hubble & NASA.

A study by Noaz & Narayan (2014) suggests that globular clusters can form naturally whenever there is some relative velocity between normal matter and dark matter. In this scenario, the formation of a globular cluster begins with a collapsing clump of normal matter in a dark matter halo which is itself also collapsing. The gravity that is driving the collapse comes mostly from dark matter. However, the collapsing clump of normal matter eventually finds itself outside the dark matter halo due to the relative velocity between the normal matter and dark matter components. If the relative velocity is small, then the clump of normal matter remains in the dark matter halo and forms a typical dwarf galaxy with somewhat comparable proportions of normal matter and dark matter, albeit more dark matter.

As a consequence of the relative velocity between the normal matter and dark matter components, the collapsing clump of normal matter becomes a long-lived dark matter-free gravitationally self-bound object (i.e. a globular cluster). Such a clump of normal matter can have a mass ranging from roughly a hundred thousand to a few million times the Sun’s mass, consistent with the masses of present-day globular clusters. On the contrary, the corresponding dark matter halo, depleted of normal matter, could become a dark satellite galaxy or an ultra-faint satellite galaxy. Such a galaxy would be comprised almost entirely of dark matter and would contain extremely few stars, possibly none at all, since stars are made of normal matter.

Noaz & Narayan (2014), “Globular Clusters and Dark Satellite Galaxies through the Stream Velocity”, arXiv:1407.3795 [astro-ph.GA]

Thursday, July 24, 2014

Ultra-Dense Ocean on a Neutron Star

A neutron star is an ultra-dense remnant core leftover from the violent demise of a massive star. It packs roughly as much mass as the Sun in an incredibly tiny volume measuring just several kilometres across. A spoonful of its material would contain a mass of roughly a billion tons. If the neutron star has a sufficiently close stellar companion, it can strip material from the companion in a process known as accretion. The accreted material can lead to the formation of an ocean on the neutron star. This ultra-dense and exotic ocean is comprised of elements with atomic number Z = 6 and larger. Most of these elements are formed from nuclear burning of the accreted hydrogen and helium from the companion star. Here, the ions behave like a liquid, hence the term “ocean”. Nonetheless, it is in no way like the oceans on Earth. The densities, pressures and temperatures are so extreme that they are only comprehensible numerically.

Figure 1: Artist’s impression of an accreting neutron star. Material stripped from the companion star forms an accretion disk around the neutron star. Image credit: NASA / Goddard Space Flight Centre / Dana Berry.

The ability to observe the sky in X-rays using space-based instruments has led to the discovery of superbursts. These energetic outbursts recur on timescales of years and are believed to be driven by the unstable ignition of a carbon-enriched layer on a neutron star. To ignite a superburst, a carbon-enriched layer needs to contain a carbon mass fraction of roughly 20 percent. However, such a carbon-enriched layer is difficult to produce in most theoretical models. Besides requiring enough carbon, models for superbursts also require large ocean temperatures of roughly 600 million K. Such high temperatures are difficult to attain from standing heating models of neutron stars.

A study by Medin & Cumming (2011) suggests that the preferential freezing of heavier elements at the base of the ocean on an accreting neutron star can substantially enrich the ocean with lighter elements such as oxygen and carbon. At the base of the ocean, the increasing pressure from the continuous accretion of material onto the neutron star forces the preferential freezing of heavier elements. The separation of lighter elements from heavier elements releases energy and provides an additional source of heating for the ocean. After the preferential freeze-out of heavier elements, the remaining fluid becomes lighter than the fluid immediately above it and acts as a source of buoyancy which drives convective mixing of the ocean. Convection distributes the heat throughout the ocean in the form of a convective flux. The extra heat input can raise the temperature of the ocean up to the required ignition temperature of around 500 to 600 million K to produce a superburst.

In the study, a 300 million K ocean consisting of a mixture of iron (Z = 26) and selenium (Z = 34), and a mixture of oxygen (Z = 8) and selenium (Z = 34) is examined. At the base of the ocean, the preferential freezing of heavier elements enhances the abundances of lighter elements in the ocean. For example, a mixture of oxygen and selenium with initial 2 percent oxygen by mass can be enriched to almost 40 percent oxygen by mass. Although oxygen was chosen as the light element in this study, models with carbon (Z = 6) were also investigated and shown to yield similar enrichment results. The carbon mass fraction can be brought up by enrichment to the required ~20 percent for superburst ignition.

Figure 2: Phase diagram for crystallization of an iron/selenium mixture (top panel) and an oxygen/selenium mixture (bottom panel) in a 300 million K ocean on a neutron star. The stable liquid region of each phase diagram is labelled as “L”, the stable solid region(s) are labelled as “S” or “S1” and “S2”, and the unstable region is filled with plus symbols. Additionally, in each panel the composition at the top of the ocean is marked by a vertical dashed line, the ocean-crust boundary is marked by a horizontal dotted line, the composition of the liquid at the base of the ocean is marked by a filled square, and the composition of the solid(s) in the outer crust are marked by filled circles. Medin & Cumming (2011).

Figure 3: Thermal profile of an ocean on an accreting neutron star. The ocean is composed of a mixture of oxygen and selenium. The solid line represents the thermal profile when the convective flux (i.e. energy released at the base of the ocean from the separation of lighter elements from heavier elements) is included in the total heat flux. The dashed line represents the thermal profile when the convective flux is ignored (i.e. the total heat flux is due only to the heat emanating from the neutron star’s interior). Medin & Cumming (2011).

Medin & Cumming, “Compositionally Driven Convection in the Oceans of Accreting Neutron Stars”, ApJ 730:97 (10pp), 2011 April 1.

Wednesday, July 23, 2014

Could it be a “Q-Star” instead of a Black Hole?

Compact objects fall under two categories - neutron stars or black holes. Neutron stars are the ultra-dense, compact remnant cores of massive stars. They are made almost entire of neutrons and have densities comparable to the density of an atomic nucleus. These neutrons are held together and kept from transmuting back into normal matter by the neutron star’s intense gravity which arises from its extraordinary compactness. A teaspoon of neutron star material would contain a mass of roughly a billion tons. The minimum and maximum mass possible for any neutron star is between ~0.1 and ~3 times the Sun’s mass. Below the minimum mass, the neutron star’s gravity is too weak to hold the star together and the star “decompresses” into normal matter. Above the maximum mass, the neutron star’s gravity becomes sufficiently strong to crush it into a black hole.

Figure 1: Artist’s impression of a neutron star whose intense gravity is lensing light from the background.

Nevertheless, the physics of matter at ultra-high densities remains poorly understood. Bahcall, Lynn & Selipsky (1990) propose that the same type of matter found in a neutron star could be stably confined by an alternative means other than gravity. Such a form of matter, though still considered ultra-dense, would have densities far below what is found in a neutron star. The outcome is that a compact object made of such a form of matter could exceed 3 times the Sun’s mass and would not collapse into a black hole under its own gravity since it is not as compact as a neutron star. These objects are termed “Q-stars”.

Theoretical models by Miller, Shahbaz & Nolan (1997) show Q-stars can be up to several times the Sun’s mass, far above the maximum mass for neutron stars. Furthermore, Q-stars that are several times the Sun’s mass can have radii less than 1.5 times the event horizon radius of a black hole of corresponding mass. Basically, a black hole’s event horizon is a non-physical boundary around a black hole, and within it, gravity is strong enough to keep even light from escaping. Since a black hole does not have a true surface, its event horizon could be regarded as its “surface”.

Figure 2: Radius of a Q-star plotted as a function of its mass. Miller, Shahbaz & Nolan (1997).

A non-rotating Q-star with 12 times the Sun’s mass can have a radius as small as ~52 km. In comparison, a black hole of the same mass would have an event horizon radius of 36 km. This difference is less than a factor of 1.5 and shows that a Q-star can be comparable in size to the event horizon of a black hole of corresponding mass. As a consequence, it may be difficult to observationally determine whether a high-mass compact object with several times the Sun’s mass is a black hole or a Q-star.

One possible method to distinguish a black hole from a Q-star would be to observe the accretion of material by the high-mass compact object. If the object were a Q-star, the accretion flow would eventually intersect the surface. If the accretion flow extends further inwards, closer than what would otherwise be the surface of the Q-star, it would be good evidence that the high-mass compact object is a black hole rather than a Q-star. An example of a known high-mass compact object that could turn out to be a Q-star is V404 Cygni - an object currently thought to be a black hole with ~12 times the Sun’s mass. Even so, one should be mindful that Q-stars are purely theoretical constructs and they may not exist at all.

- Bahcall, Lynn & Selipsky, “New Models for Neutron Stars”, ApJ (1990) 362, 251.
- Miller, Shahbaz & Nolan, “Are Q-stars a serious threat for stellar-mass black hole candidates”, MNRAS (1990) 294: L25-L29.

Tuesday, July 22, 2014

Formation of Binary Giant Planets

Giant planets seem to be ubiquitous around Sun-like stars. Our Solar System has two giant planets - Jupiter and Saturn. Both planets are primarily composed of hydrogen and helium. Jupiter and Saturn have 318 and 95 times the mass of Earth, respectively. Beyond Saturn, the planets Uranus and Neptune are generally classified as “ice giants” because they have much smaller masses and differ considerably in composition compared to Jupiter and Saturn. The orbits of Jupiter and Saturn form a 5:2 orbital resonance. For every five times Jupiter circles the Sun, Saturn would circle the Sun twice. On the whole, the orbits of Jupiter and Saturn are stable over the entire age of our Solar System.

In a planetary system with two giant planets, such as our Solar System, energy and angular momentum are conserved between the two giant planets, and the planetary system is stable. Instability only occurs if the orbits of the two giant planets bring them very close to one another. Exoplanet discoveries over the years have revealed a remarkable diversity of planetary systems. A number of studies have shown that planetary systems with three or more giant planets tend to be unstable. For such a planetary system, perturbations by the additional giant planet(s) tend to destabilise the system.

Figure 1: Artist’s impression of a pair of binary giant planets.

Figure 2: Artist’s impression of a giant planet.

When a planetary system consisting of three or more giant planets is destabilised, it can lead to a number of interesting outcomes. Ochiai et al. (2014) show that gravitationally bounded pairs of giant planets (i.e. binary giant planets) can form via planet-planet scattering during the destabilisation of a planetary system with three giant planets. In their study, N-body simulations of planetary systems with three Jupiter-mass giant planets were performed. The N-body simulations show that as much as ~10 percent of the planetary systems result in the formation of binary giant planets.

During the destabilization of a planetary system with three giant planets, the possible outcomes are - ejection of a planet, planet-planet collision, planet-star collision, formation of a hot-Jupiter and formation of a pair of binary giant planets. A hot-Jupiter forms when a giant planet is thrown inwards to its star whereby planet-star tidal interactions can circularise the orbit of the giant planet into a close-in orbit around the star, leading to the formation of a hot-Jupiter. As for binary giant planets, such a pair could form when two giant planets pass sufficiently close to one another that enough tidal dissipation occurs between them to form a gravitationally bound pair.

In their N-body simulations of planetary systems with three giant planets, Ochiai et al. (2014) used four sets of 100 simulation runs corresponding to the four different initial stellarcentric semimajor axes - 1, 3, 5 and 10 AU for the innermost giant planet. In the nomenclature, “stellarcentric semimajor axis” refers to the average distance of the giant planet from its host star and 1 AU is a unit of measurement equal to the average Earth-Sun separation distance. For the two outer giant plants, their semimajor axes are, respectively, factors of 1.45 and 1.9 times the semimajor axis of the innermost giant planet. The four sets of 100 runs follow the evolution of the planetary system over a period of 10 million years.

The results from the 400 simulation runs show that the formation rate of binary giant planets is ~10 percent and nearly independent of the stellarcentric semimajor axis. Binary giant planets generally form near their initial orbits because the period when they form is normally during the early stages of orbital instability. Regardless of the initial stellarcentric semimajor axes, the distribution of the semimajor axes of the binary giant planets (i.e. average distance between the two giant planets in the binary) show a peak at 2 to 4 times the combined planetary radii of the two giant planets in the binary. Also, the 400 simulation runs show that ejection rates increase and collision rates decrease as stellarcentric semimajor axis increases.

Figure 3: Distribution of the semimajor axes of the binary giant planets obtained from the 400 simulation runs. For each pair of binary giant planets, the semimajor axis is expressed as a ratio to the combined planetary radii of the two giant planets in the binary. Ochiai et al. (2014).

Figure 4: Results obtained from the 400 simulation runs for the four different initial stellarcentric semimajor axes - 1, 3, 5 and 10 AU. The colours represent binary giant planets (red), planet-planet or planet-star collisions (light green), hot-Jupiters (blue), ejections (magenta), and three giant planets still remaining after 10 million years (light blue). Ochiai et al. (2014).

Binary giant planets are expected to be stable over the long-term. If the stellarcentric semimajor axis of a pair of binary giant planets is larger than ~0.3 AU, the system is stable for ~10 billion years, which is similar in duration to the main-sequence lifespan of a Sun-like star. Interestingly, binary giant planets can have moons with wide orbits that circumscribe both planets. A loosely bound moon around one of the two giant planets has a roughly 20 percent chance of surviving the formation process leading to a pair of binary giant planets. Additionally, binary giant planets can also capture large moons into orbit around them, much like how Neptune captured its large moon Triton. Current planet detection methods might be able to detect binary giant planets.

Ochiai et al., “Extrasolar Binary Planets. I. Formation by Tidal Capture during Planet-Planet Scattering”, ApJ 790:92 (10pp), 2014 August 1

Monday, July 21, 2014

Kepler-421b: A Uranus-Sized Planet near the Snow-Line

“In future, children won’t perceive the stars as mere twinkling points of light: they’ll learn that each is a ‘Sun’, orbited by planets fully as interesting as those in our Solar System.”
- Martin Rees

A protoplanetary disk is a circumstellar disk of material around a young star in which the formation of planets occurs. The snow-line marks the distance from the central star where the protoplanetary disk becomes cool enough for volatiles such as water to condense into solid ice grains. By analysing publicly available data from NASA’s Kepler space telescope, Kipping et al. (2014) present the discovery of a cold transiting planet near the snow-line. This planet, identified as Kepler-421b, is the first of its kind to be discovered. It is similar in size to Uranus and it circles a star that is slightly cooler than the Sun in a nearly-circular orbit with an orbital period of 704.2 days. Kepler-421b is the longest period transiting planet discovered to date.

Figure 1: Artist’s impression of a Uranus-like planet with a large moon in orbit around it.

Figure 2: Transit light curve of Kepler-421b. Based on how much light it blocks when it passes in front its parent star, Kepler-421b is estimated to be ~4 times the Earth’s diameter, roughly the size of Uranus. Kipping et al. (2014).

“Finding Kepler-421b was a stroke of luck,” says lead author David Kipping of the Harvard-Smithsonian Center for Astrophysics (CfA). “The farther a planet is from its star, the less likely it is to transit the star from Earth’s point of view. It has to line up just right.” Kepler-421b is ~1.2 AU from its parent star. At that distance, the planet is closer to its parent star than Mars is from the Sun. Since its parents star is only ~40 percent as luminous as the Sun, Kepler-421b receives only ~64 percent of the insolation Mars gets from the Sun, or ~28 percent of the insolation Earth gets from the Sun. Kepler-421b receives the same amount of insolation as an object at ~2 AU from the Sun. If Kepler421b has a Uranus-like albedo, the planet’s effective temperature would be ~180 K. For comparison, Earth has a mean surface temperature of 288 K, or 15°C.

Assuming Kepler-421b has a Uranus-like composition (i.e. an ice giant), the planet probably formed at its current distance from its parent star (i.e. in situ formation). At that distance, it is cool enough for icy planetesimals to form in the protoplanetary disk, eventually leading to the creation of an ice giant. The large orbital period of 704.2 days means that transits of Kepler-421b are relatively infrequent. In fact, only two transits have been observed so far and those were sufficient to result in its initial detection. Unfortunately, the 3rd transit occurred in March 2014, after Kepler’s primary mission. Nevertheless, the 4th transit opportunity is in February 2016. Kepler-412b is the first known transiting Uranus-sized planet in a long-period orbit. Determining its mass and finding more of its kind would be the next logical steps. Finally, the large distance of Kepler421b from its parent star makes it an appealing target in the search for exomoons.

Kipping et al. (2014), “Discovery of a Transiting Planet Near the Snow-Line”, arXiv:1407.4807 [astro-ph.EP]

Hot-Jupiters around Red Giant Stars

Figure 1: Artist’s impression of a gas giant planet.

Figure 2: Artist’s impression of a gas giant planet.

After ~6 billion years or so, the Sun will start running out of hydrogen in its core and being to enter its post-main-sequence phase of evolution characterised by a large increase in its luminosity. All the planets circling the Sun will receive much greater insolation than they do now. Presently, Jupiter orbits the Sun at a distance of roughly 5 AU, where 1 AU is the average Earth-Sun separation distance. When the Sun enters post-main-sequence evolution, Jupiter might become so intensely irradiated that it becomes a “hot-Jupiter”.

This occurs because the Sun’s luminosity will increase by a factor of several thousand during two stages in its post-main-sequence evolution - the red giant branch (RGB) stage, followed by the asymptotic giant branch (AGB) stage. During the RGB stage, the Sun’s interior is characterised by an inert helium core surrounded by a hydrogen-burning shell (i.e. hydrogen fusing into helium). For the subsequent AGB stage, the Sun’s interior is characterised by an inert carbon core surrounded by a helium-burning shell (i.e. helium fusing into carbon), and a hydrogen-burning shell.

A study by Spiegel & Madhusudhan (2012) show that Jupiter’s atmosphere can be transiently heated to temperatures of up to ~1000 K or more when the Sun goes through its RGB and AGB stages. Many of the currently known Jupiter-mass planets in wide, several-AU orbits around Sun-like stars (i.e. stars between 1 to 3 times the Sun’s mass) will also experience such a temperature increase when their host stars evolve off the main-sequence. The authors term such planets “red giant hot-Jupiters” (RGHJs) to distinguish them from typical hot-Jupiters that circle in short-period, close-in orbits around main-sequence stars.

Figure 3: Artist’s impression of a gas giant planet.

Figure 4: Orbital separations where RGHJs can be found around a Sun-like star. The first rise in temperature corresponds to the RGB phase and the second rise corresponds to the AGB phase. Spiegel & Madhusudhan (2012).

Gas giant planets like Jupiter start out warm and gradually cool over time. However, the intense heating a RGHJ receives from its post-main-sequence host star could “reset” its evolutionary clock. As a result, RGHJs or post-RGHJs could appear younger than they actually are. Nevertheless, there is a major difference between RGHJs and typical hot-Jupiters around main-sequence stars. Typical hot-Jupiters trap some fraction of incident stellar irradiation to produce bulk heating (i.e. internal heating) on timescales spanning tens of millions of years, allowing these objects to settle into a quasi-steady thermal state. By comparison, RGHJs do not have the luxury of time since the RGB and AGB phases last nowhere as long. A RGHJ would not be intensely irradiated for a long enough time to produce any significant bulk heating, even though their outer layers might appear strongly heated.

When a star enters the RGB stage, and subsequently, the AGB stage, it will undergo a huge increase in mass loss. The star loses mass in the form of a continuous stellar wind streaming away from the star. Since stellar wind speeds are typically a few times a star’s surface escape velocity, the stellar wind from a post-main-sequence star would travel much slower compared to the stellar wind from a main-sequence star. This is because a post-main-sequence star has puffed up so greatly in size that its surface escape velocity is small.

In fact, measurements of the stellar wind speeds of AGB stars by Zuckerman & Dyck (1989) show velocities less than 40 km/s, with a clustering around 5 to 25 km/s. For comparison, stellar winds from main-sequence stars, such as the present-day Sun, have speeds of a few 100 km/s. A Jupiter-mass planet around a post-main-sequence star would have sufficient gravity to capture and accrete the slow stellar wind flowing pass it. The total accreted mass is estimated to be of order ~1/10,000th of the planet’s mass for a Jupiter-mass planet.

The stellar wind streaming away from a post-main-sequence star can impose enough drag to change the orbits of small objects circling the star. These objects could then impact the RGHJ and enhance the abundance of heavy elements in the planet’s atmosphere. Furthermore, a post-main-sequence can exhibit a high carbon-to-oxygen ratio due to carbon dredged-up from the star’s interior. A RGHJ accreting the stellar wind from such a star could acquire a carbon-rich atmosphere.

Figure 5: Artist’s impression of a gas giant planet.

 Figure 6: Artist’s impression of a gas giant planet with a few of its moons appearing as points of light.

The intense stellar irradiation experienced by a RGHJ means that wind speeds in its atmosphere are expected to be faster than on present-day Jupiter. However, wind speeds on a RGHJ would not be as strong as on a typical hot-Jupiter around a main-sequence star because a RGHJ is not tidally-locked, and consequently, would not have a large enough day-night temperature contrast to drive strong winds. Roughly estimating, the expected wind speeds for present-day Jupiter, a RGHJ and a typical hot-Jupiter are 40 m/s, ~100 m/s and ~1000 m/s respectively.

Changes in the incident stellar irradiation around an evolving post-main-sequence star can cause interesting changes in the atmospheric chemical properties of a RGHJ. The present-day Jupiter has an atmospheric temperature of 165 K at the 1 bar level. As the Sun’s luminosity increases considerably during its post-main-sequence phase, an important change for a soon-to-be RGHJ would be the enhancement of the H2O abundance in the planet’s atmosphere as it becomes warm enough (i.e. ~300 K) for water-ice to sublimate. The abundance of H2O drops slights for a brief period during the RGB stage when atmospheric temperatures on the RGHJ exceed ~600 K. At such temperatures, some of the oxygen in H2O becomes bounded in silicates. The same drop in H2O abundance might also occur when temperatures rise again during the AGB stage.

Figure 7: Post-main-sequence evolution of Jupiter’s equilibrium temperature and atmospheric chemical composition. Spiegel & Madhusudhan (2012).

Figure 8: Example spectra of Jupiter as a function of equilibrium temperature, as seen from a distance of 5 AU. Spiegel & Madhusudhan (2012).

- Spiegel & Madhusudhan (2012), “Jupiter will become a hot Jupiter: Consequences of Post-Main-Sequence Stellar Evolution on Gas Giant Planets”, arXiv:1207.2770 [astro-ph.EP]
- Zuckerman & Dyck (1989), “Outflow Velocities from Carbon Stars”, Astronomy and Astrophysics, 209, 119-125

Sunday, July 20, 2014

Life and the Formation of Continents

On Earth, the presence of life plays a major role in determining the chemistry of the atmosphere and oceans. A study by D. Höning et al. (2014) suggests that the presence of life may play an even deeper role in influencing the planet’s evolution. In particular, the presence of life can enhance continental weathering rates, thereby increasing the rates at which sediments wash into and settle on the bottom of the oceans. These sedimentary layers hold within them a significant amount of water and hydrated minerals. Along convergent plate boundaries, the oceanic crust gets subducted into the Earth’s mantle, bringing along the water-rich sedimentary layers. As the subducting oceanic crust dives deeper, the increasing lithostatic pressure squeezes free water out from the sedimentary layers in a process known as shallow dewatering.

The enhanced continental weathering rates due to the presence of life would lead to thicker sedimentary layers, and consequently, increase the amount of water being subducted. In addition, the enhanced continental weathering rates would reduce the amount of shallow dewatering due to a greater abundance of low-permeability deposits such as clay-rich deposits in the sedimentary layers. These low-permeability deposits effectively ‘seal off’ water entailed in the sedimentary layers from being squeezed out, and in doing so, reduces the amount of shallow dewatering, allowing more water to be transported by subduction deeper into the Earth’s mantle. Water that is not squeezed out becomes bound in stable hydrated minerals as it is dragged further down by the subducting oceanic crust.

Figure 1: Artist’s impression of an Earth-like planet hosting a system of rings.

Figure 2: Schematic cartoon depicting Earth’s global water cycle, where water is represented by large and small dots, its path by black arrows, and movement of the oceanic plate by white arrows. Initial water uptake occurs within the submarine oceanic crust and sediments. Water loss first occurs after the subduction trench through dewatering, followed by the formation of the water-rich partial melt. The partial melt drives arc volcanism and continental crust formation. However, a fraction of the water contained in the subducting plate is regassed into the mantle. The water leaves the convecting mantle at mid oceanic ridges (MOR) as free volatiles or becomes part of the newly formed oceanic crust. D. Höning et al. (2014).

At a depth of roughly 100 km, the hydrated minerals brought down by the subducting oceanic crust become unstable and releases water into the surrounding mantle. This lowers the melting temperature of the surrounding mantle and leads to partial melting. Buoyancy drives the partial melt towards the surface, causing surface volcanism and the formation of new continental crust. The amount of newly formed continental crust is directly proportional to the amount of water released to produce partial melting. A larger amount of water driven down by the subduing ocean crust and released to form partial melting would enhance the rate of production of continental crust.

Also, not all the water in the form of hydrated minerals is released to form partial melting. Some of it continues deeper into the Earth’s mantle where it dissolves, hydrating the mantle. This lowers the effective viscosity of the mantle (i.e. makes the mantle more fluid), which has the effect of stabilizing plate tectonics. Process such as subduction, volcanic activity and the formation of new continental crust depends a lot on plate tectonics. If the mantle was dry, plate tectonics might not occur due to the high mantle viscosity. In a way, plate tectonics needs water to operate.

On Earth, the oceans cover 70 percent of the planet’s surface, with land covering the remaining 30 percent. This study by D. Höning et al. (2014) show that the presence of life does play an important role in the formation of continents on Earth and should be applicable to the evolution of other Earth-like planets as well. In the study, a model depicting Earth with present day continental weathering and erosion rates show that after roughly 4 billion years, the planet reaches a steady state with continental area covering 40 percent of the planet’s surface and an upper mantle water concentration of 300 parts per million (ppm). The same model is then ran with 60 percent of present day continental weathering and erosion rates, which one might expect on an abiotic Earth (i.e. a lifeless Earth). After 4 billion years, the planet attains a steady state with continents covering ~5 percent of its surface and an upper mantle water concentration of 40 ppm.

Figure 3: Artist’s impression of an Earth-like world. In this case, it is a moon of a gas giant planet.

Figure 4: Artist’s impression of an Earth-like planet. Image credit: Adrian Thomassen.

These findings suggest that the difference between a life-filled and a lifeless Earth can showup as a significant difference in the extent of continental coverage. On a biotic world (i.e. a life-filled planet), the presence of life enhances the formation of continents and stabilizes plate tectonics. In contrast, on an abiotic world (i.e. a lifeless planet), continental coverage is expected to be less, and the occurrence of plate tectonics would be less likely. If future studies support these notions, the detection of large continental coverage and/or plate tectonics on Earth-like exoplanets could serve as a form of biosignature in the search for life beyond Earth. “If we find a planet somewhere in the universe with a continental coverage similar to the Earth, it may be a good place to search for life,” said lead author of the study, Dennis Höning, a planetary scientist at the German Aerospace Centre’s Institute of Planetary Research in Berlin.

D. Höning et al., “Biotic vs. abiotic Earth: A model for mantle hydration and continental coverage”, Planetary and Space Science 98 (2014) 5-13.

Saturday, July 19, 2014

Hellacious Superrotating Winds on Hot-Jupiters

Hot-Jupiters are a class of exoplanets that share similar characteristics to Jupiter (i.e. they are all gas giant planets), but have extraordinarily high surface temperatures because they orbit very close to their parents stars. On a hot-Jupiter, the intense heating on the planet’s dayside drives powerful winds that tear continually around the planet, transporting heat from the dayside and dumping it on the nightside. These hellacious winds whip around the planet from west to east, generating what is known as superrotation. The winds extend from the planet’s equator to latitudes of typically 20° to 60°. Because a hot-Jupiter orbits so close to its parent star, the planet is most likely tidally-locked and presents the same hemisphere towards its parent star all the time. Superrotation on a tidally-locked hot-Jupiter tends to produce an eastward displacement of the planet’s hottest region from the substellar point by up to 10° to 60° of longitude.

Figure 1: Artist’s impression of a gas giant planet.

S. Faigler & T. Mazeh (2014) analysed the Kepler light-curves of four transiting hot-Jupiters - KOI-13b, HAT-P-7b, TrES-2b and Kepler-76b. The light-curves of these four planets show beaming, ellipsoidal and reflection/emission (BEER) phase modulations. As a hot-Jupiter circles its parent star, it gravitationally tugs at the star, causing the star to “wobble”. From an observer’s point of view, the star would appear to approach and recede in a periodic fashion as the hot-Jupiter orbits around it. In the BEER phase modulations, the beaming effect, also know as Doppler boasting, is caused by an increase (decrease) in the brightness of the parent star as it approaches (recedes from) the observer. The ellipsoidal effect is caused by the tidal distortion of the star by the hot-Jupiter. Both the beaming and ellipsoidal phase modulations are proportional to the hot-Jupiter’s mass. Finally, the reflection/emission phase modulations are the result of a combination of light reflected from the planet’s dayside (reflection) and thermal radiation that is re-emitted by the planet (emission).

The back and forth motion of a star induced by the gravitational tugging of an orbiting hot-Jupiter also causes the star’s spectrum to be blue-shifted (red-shifted) when the star is observed to approach (recede from) the observer. This results in a radial velocity signature that is proportional to the planet’s mass and it serves as an independent measure of the planet’s mass in addition to the BEER phase modulations. Radial velocity measurements available for the hot-Jupiters - HAT-P-7b, TrES-2b and Kepler-76b show that the planetary-mass derived from the beaming amplitude is noticeably larger than from the radial velocity measurements. Also, for all four hot-Jupiters - KOI-13b, HAT-P-7b, TrES-2b and Kepler-76b, the planetary-mass derived from the beaming amplitude is larger than from the ellipsoidal amplitude.

S. Faigler & T. Mazeh (2014) suggests that the apparent planetary-mass discrepancy is found to be caused by superrotation, whereby the eastward displacement of the planet’s hottest spot from the substellar point induces an angle shift in the planet’s reflection/emission phase modulation which “leaks” into the beaming modulation and artificially boosts its observed amplitude. As a consequence, the planetary-mass estimated from the beaming amplitude is somewhat larger than the real one. When the effect of superrotation is included in a modified “BEER model”, the apparent planetary-mass discrepancies disappear. This study shows that hot-Jupiter superrotation may be a rather common phenomenon that can be identified in the Kepler light-curves of hot-Jupiters that exhibit considerable BEER phase modulations. For each of the four hot-Jupiters in this study, the hottest spot is estimated to be displaced eastwards from the substellar point by 0.8° ± 0.9° for KOI-13b, 5.4° ± 1.5° for HAT-P-7b, 38° ± 18° for TrES-2b and 9.2° ± 1.3° Kepler-76b.

Figure 2: The four panels correspond to the four hot-Jupiters - KOI-13b, HAT-P-7b, TrES-2b and Kepler-76b. For each panel, the solid line represents the best-fit superrotating BEER model of the corresponding hot-Jupiter’s light-curve. The residuals are shown below each panel. On each panel, the dashed, dash-dot and dotted lines represent the shifted reflection/emission, beaming and ellipsoidal models, respectively. The vertical red dashed line marks the phase of maximum reflection/emission. Notice that the phase of maximum reflection/emission of each planet comes before phase 0.5. This is consistent a superrotation-induced eastward displacement of the planet’s hottest spot from the planet’s substellar point. S. Faigler & T. Mazeh (2014).

S. Faigler & T. Mazeh (2014), “BEER analysis of Kepler and CoRoT light curves: II. Evidence for emission phase shift due to superrotation in four Kepler hot Jupiters”, arXiv:1407.2361 [astro-ph.EP]

Friday, July 18, 2014

Two Tight Pairs of Low-Mass Binary Brown Dwarfs

Figure 1: Artist’s impression of a brown dwarf. Heat from its warm interior “leaks” out through gaps in its cloud coverage. A cool brown dwarf would resemble Jupiter more than it would resemble a star.

Brown dwarfs are substellar objects that span the gap between the most massive planets and the least massive stars. Like stars, brown dwarfs can also come in pairs. Using a technique known as gravitational microlensing, Choi et al. (2013) reported the discovery of two pairs of very low-mass binary brown dwarfs identified as OGLE-2009-BLG-151 and OGLE-2011-BLG-0420. Gravitational microlensing is observed when the gravity of an intervening object (lens) magnifies the light from a background star (source). It happens as the intervening object crosses the line-of-sight between the observer and the background star.

OGLE-2009-BLG-151 was first detected by the Optical Gravitational Lensing Experiment (OGLE) group and then independently detected by the Microlensing Observations in Astrophysics (MOA) group in 2009, hence its alternate designation - “MOA-2009-BLG-232”. The other gravitational microlensing event, OGLE-2011-BLG-0420, was detected by the OGLE group in 2011. A number of ground-based telescopes also provided follow-up observations for both gravitational microlensing events.

Figure 2: Light-curves of the binary brown dwarf gravitational microlensing events OGLE-2009-BLG-151 and OGLE-2011-BLG-0420. Choi et al. (2013).

The light-curve of OGLE-2009-BLG-151 is characterised by two prominent spikes, consistent with a binary-lens model. Based on the light-curve, OGLE-2009-BLG-151 is inferred to be a tightly-bound pair of very low-mass brown dwarfs with masses 0.018 and 0.0075 times the Sun’s mass. The two brown dwarfs are projected to be 0.31 AU, or 46 million km apart from each other. That is equal to Mercury’s closest distance to the Sun. For comparison, the average Earth-Sun separation distance is 1 AU, or 149.6 million km.

As for OGLE-2011-BLG-0420, its light-curve appears smooth and symmetric. However, upon careful observations, the light-curve shows noticeable deviations indicative of a binary-lens model rather than a single-lens model. Similar to OGLE-2009-BLG-151, OGLE-2011-BLG-0420 consists of two very low-mass brown dwarfs in a tight binary system. The two brown dwarfs are 0.025 and 0.0094 times the Sun’s mass, and are spaced only 0.19 AU, or 28 million km apart. In fact, the two brown dwarfs of OGLE-2011-BLG-0420 are spaced as far apart as Jupiter’s outermost moons are from Jupiter.

The total system masses of OGLE-2009-BLG-151 and OGLE-2011-BLG-0420 are 0.025 and 0.034 times the Sun’s mass, respectively, placing them well below the hydrogen-burning limit of ~0.08 times the Sun’s mass. What makes the discovery of OGLE-2009-BLG-151 and OGLE-2011-BLG-0420 interesting is that the two systems have among the lowest total system masses known for brown dwarf binaries. Additionally, OGLE-2009-BLG-151 and OGLE-2011-BLG-0420 are also the tightest known brown dwarf binaries. The discovery of these two systems among the relatively small sample of gravitational microlensing events involving binary systems shows that tightly-bound pairs of very low-mass brown dwarfs are not uncommon.

Figure 3: Projected separation versus total system mass for a compilation of binaries. Grey circles indicate old field binaries, whereas blue squares indicate young (< 500 million year old) systems. The size of the symbols is proportional to the square root of the mass ratio (i.e. the ratio of the less massive component to the more massive component in the binary system). The red stars correspond to OGLE-2009-BLG-151 and OGLE-2011-BLG-0420. Choi et al. (2013).

Figure 4: Binding energy versus total system mass for the same binaries as shown in Figure 3. Choi et al. (2013).

Choi et al. (2013), “Microlensing Discovery of a Population of Very Tight, Very Low-mass Binary Brown Dwarfs”, arXiv:1302.4169 [astro-ph.SR]