Astrobites RSS

planet transits

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: A Compact Multi-Planet System With A Significantly Misaligned Ultra Short Period Planet
Authors: Joseph E. Rodriguez et al.
First Author’s Institution: Harvard-Smithsonian Center for Astrophysics
Status: Submitted to the AAS Journals

Our solar system is flat: all the planets orbit in the same plane as the Sun’s equator, at 90 degrees to its pole, with low orbital inclination relative to each other. While distant exoplanets can orbit at any inclination relative to the plane of the sky, the exoplanets that we detect via transits nearly always have inclination values very close to 90° — i.e, they orbit perpendicular to the plane of the sky. This is what allows us to see them passing in front of their star.

However, the authors of today’s paper have discovered a transiting exoplanet with an inclination of 76.5°. This is possible because EPIC248435473 b is an ultra short period planet (USP) with a period of just 0.66 days, meaning from our viewpoint the planet still transits in front of its star in spite of its larger inclination angle. What makes this discovery more interesting is that the authors have found up to five other transiting planets in this system, all aligned with inclinations 88-90°, making EPIC248435473 b the odd one out.

Discovering and Modeling the Planets

Figure 1: The full K2 light curve of EPIC248435473 from Campaign 14. Top: light curve corrected for systematics. Bottom: flattened light curve with best fit model from EXOFASTv2 showing modeled transits of its planets. [Rodriguez et al. 2018]

Three planet candidates (later named b, d, e) around EPIC248435473 were picked up automatically by the pipeline run on the K2 data (see Figure 1). The authors identified three additional candidates from visual inspection of the light curves. The phase-folded light curves of the six potential planets can be seen in Figure 2. Only four of these planet candidates could be statistically validated, as noisier light curves were extracted to exclude a close background star, making the weakest two candidates no longer reliable. These two remaining planet candidates will require more data to confirm.

Figure 2: Phase folded light curves for validated planets: b,c,d,e and candidate planets .02 and .06. All the repeating transit data from K2 is folded onto each planet’s period, combining into one transit to show the best period and quality of fit of the EXOFASTv2 model (shown by the red line). [Rodriguez et al. 2018]

The authors then simultaneously modelled the SED, radial velocities (RVs), and flattened K2 photometry for the system of four validated and two candidate planets using a software package called EXOFASTv2. The stellar radius and mass were inferred from the spectroscopy. Simultaneous modeling refined the system properties, including the planet masses and inclinations, to fit all the available data.

Modeling planet b was made tricker as it has a grazing transit — only part of the planet passes in front of the stellar disk — resulting in the V shape transit in Figure 2. In this case, the transit data cannot constrain the planetary radius well. As an approximation, the planetary mass found using RVs can be used to estimate the planet radius. Unfortunately the RV data is not precise enough to constrain the mass (see Figure 3), so only an upper limit is found. Using Chen and Kipping’s exoplanet mass–radius relation results in a wide possible radius distribution between 1–10 Earth radii, shown in Figure 4, with the radius most likely about 3 Earth radii.

Figure 3: Radial-velocity data folded on planet b’s period. RV data is not precise enough to measure the planet mass, as RV error bars are quite large compared to the RV signal. Therefore only an upper limit on mass can be found. [Rodriguez et al. 2018]

Figure 4: Probability of planet b having different radii. Transit depth sets a lower limit of 1 Earth radius while an upper mass limit comes from the non-detection of RV signal which is converted to a radius. [Rodriguez et al. 2018]

A Misaligned Planet in a Multi-Planet System

Explaining how EPIC248435473 b entered its short misaligned orbit is an open question. Theorists do not believe it could have formed so close to its star; they argue it must instead have moved in from further away in the system.

Proto-USP planets have been proposed to start in periods of 5–10 days before being gravitationally tugged into eccentric and inclined orbits by other planets. These orbits later circularize at the point closest to the star, but the planets stay inclined. However, this planet’s radius of 3 Earth radii, if confirmed, is larger than other USP planets and implies a dense and water-rich atmosphere that must have formed much further away from the star to resist the strong photoevaporation.

Additional complications arise as EPIC248435473 b is in a compact system. Generally, the planets moved into ultrashort period orbits by the mechanism described above would only have companions with periods of 10 days or more. Planet c and candidate .02 are closer, with periods of 6 and 7 days, so they must have migrated inwards after the USP planet reached its present location.

Further observations to constrain the mass and radius of EPIC248435473 b, as well as study of its atmosphere, may help narrow down its formation history. The planet bulk density and atmospheric constraints on water could help restrict its origin within the disk and determine when it reached its current location. This information should help explain how this strange system came to exist.

About the author, Emma Foxell:

I am a PhD student at the University of Warwick. My project involves searching for transiting exoplanets around bright stars using telescopes on the ground. Outside of astronomy, I enjoy rock climbing and hiking.

AB Doradus

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: A Young Ultramassive White Dwarf in the AB Doradus Moving Group
Author: Jonathan Gagné, Gilles Fontaine, Amélie Simon, & Jacqueline K. Faherty
First Author’s Institution: Carnegie Institution of Washington DTM
Status: Published in ApJL

Figure 1: Artist’s impression of a white dwarf. [All About Space/Imagine Publishing]

In today’s astrobite, we’re talking about a paper on the AB Doradus co-moving group of stars. The paper shows that a white dwarf named GD 50 is part of the AB Dor group; we’ll be talking about how the authors arrived at this result, and what this can tell us about the history and stellar evolution of the group.

A co-moving group of stars is just what it says on the tin: a group of stars that are all moving in about the same direction at about the same speed. Such a group is normally around a couple of dozen stars. They’re generally not too far away from each other in space as well, but they’re not nearly as tightly bunched up as clusters are; the space inside a co-moving group can contain many stars that aren’t part of the group. For instance, the Ursa Major moving group includes most of the brightest stars in the Big Dipper, and it also extends to include some stars at the far end of the sky in Triangulum Australe, but it does not include the Sun — even though the Sun is between the two. Stars in a co-moving group are generally thought to have formed all at once, perhaps as part of an open cluster that has broken apart. This shared history can make co-moving groups useful for constraining models of stellar evolution.

One of the closest co-moving groups is the AB Doradus moving group, a group of about 30 stars centred about 20 parsecs away from the Earth. A possible member of the group for some time has been a white dwarf called GD 50, but there have never been precise enough data on the star to know for sure. If true, GD 50 would be the only white dwarf in the group. This makes it interesting for stellar evolution purposes, as it would provide a new window onto the stellar history of the group. White dwarf models have their own set of strengths and weaknesses that differ from those for main sequence stars; for instance, the age of a white dwarf can generally be determined much more precisely than a main sequence star.

Testing Membership

Figure 2: The measured velocity of GD 50 (shown as a red star) compared with other members of the AB Dor group of stars. U, V and W refer to velocity components in each of three different directions (see text). Because the spectroscopic radial velocity of GD50 is much less certain than other velocity measurements, the yellow dashed line here shows the direction in which GD 50 would move if its radial velocity were to change, and the spread of purple dots show randomly-generated models spread around the best-fit value to give an idea of the uncertainty. The star called AB Dor, after which the co-moving group is named, is highlighted in green. [Gagné et al. 2018]

Today’s authors set out to test GD 50’s membership of the AB Dor group using new data from Gaia‘s second data release. Regular readers might be growing sick of reading about Gaia by now, but for those who aren’t familiar with the idea, Gaia is a satellite that has measured the distance to over a billion stars by the parallax method

, as well as measuring the proper motions (motion across the sky) of those stars. By combining these with a spectroscopic measurement of the star’s velocity towards or away from the Earth, we can pin the star down quite well in both position and velocity space.

In order to test whether GD 50 really belongs to the AB Dor group, the authors took the position and velocity of the star in three dimensions (away from the Galactic centre, around the Galactic centre, and up-and-down relative to the Galactic plane) to give six properties. For each property, they modelled the group as having a Gaussian distribution, and they found the probability of a group member having the property measured for GD 50. They compared this to the probability of the star being just any old member of the galactic field by modelling the local galactic field as a more complex combination of 10 Gaussians. Overall, they found a 99.7% likelihood that the star is a part of the AB Dor group (see Fig 2).

Figure 3: Histogram of initial stellar masses in the AB Dor group. The right-most bin includes only one object, which is the main sequence star that evolved into GD 50. The red dashed line shows a model distribution for inital stellar masses. For the 4 highest-mass bins the model agrees well. The lowest-mass bin has fewer stars than the model predicts. This is probably due to the selection effect against low-mass, faint stars — ie, these stars exist but we haven’t detected them yet. Source: Figure 3 in today’s paper.

The History of GD 50

So what do we know about GD 50, and what can this tell us about the AB Dor group? Firstly, we know what kind of main sequence star GD 50 evolved from. Stars lose most of their mass as they evolve into white dwarfs, but this process is pretty well understood. GD 50 is a pretty heavy white dwarf, with a mass 1.28 times the mass of the Sun. This means that, as a main-sequence star, it must have come in at about 7.8 times the mass of the Sun. This would make it the only star in the AB Dor group with this much mass, but given models of initial mass functions for clusters, we would expect the group to contain around one star this massive (see Fig 3).

We can also tell the age of GD 50 quite well based on models of stellar evolution and white-dwarf cooling — the first tells you how long a star of 7.8 solar masses would live, and the second tells you how long it must have been a white dwarf in order to match GD 50’s temperature. Today’s authors find a total age for GD 50 of 117 +/- 22 million years. Previous estimates for the age of the AB Dor group gave around 130-200 million years, so the estimate in today’s paper is a touch shorter and quite a lot more tightly constrained. This puts the AB Dor group among the youngest known nearby co-moving groups.

Overall, the identification of GD 50 as part of the AB Dor group allows us to put some constraints on the history of that group, and gives us a way to check previous measurements of properties like the group’s age. The data from Gaia is only beginning to make its impact, and we can hope that, in the future, more results like this will be found and help us to build towards a better understanding of the histories of these stellar groups.

About the author, Matthew Green:

I am a PhD student at the University of Warwick. I work with white dwarf binary systems, and in particular with AM CVn-type binaries. In my spare time I enjoy writing of all kinds, as well as playing music, board games and rock climbing. For more things written by me, take a look at my website.

Gaia

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: Off the beaten path: Gaia reveals GD-1 stars outside of the main stream
Author: Adrian M. Price-Whelan, Ana Bonaca
First Author’s Institution: Princeton University
Status: Submitted to ApJL

The Gaia space telescope is revolutionizing our understanding of the Milky Way. This European satellite (Figure 1) is carefully tracking the positions of over a billion stars over five years, providing us with an evolving map of stellar locations and velocities. Just a couple months ago the second Gaia data catalog was released, including brand new information about the motions of many times more stars than in previous datasets to accuracies never before achieved, launching a scramble to see what exciting surprises this new data would reveal about our galaxy. (For more examples of exciting Gaia science see these Astrobites.)

This new data is a powerful tool for studying the stellar halo of our galaxy — the outer region of the Milky Way system that consists of a spherical cloud of stars extending beyond the central spiral, into the outer reaches of the Milky Way’s dark matter halo. This halo of stars, and all of the systems it contains, are powerful probes of the local dark-matter distribution and contain clues as to the history of the formation and evolution of our galaxy. Today’s paper considers one particular type of system in the stellar halo — stellar streams. These groups of stars, stretched into arcs across the sky, are the remnants of small galaxies and clusters of stars that have been torn apart as they orbit around the Milky Way. Previous Astrobites have considered how stellar streams are essential tools for studying galaxy formation and the nature of dark matter, and today’s paper illustrates how powerful these streams can be with the help of Gaia.

Figure 2. A stellar stream wrapped around a galaxy. [Jon Lomberg in collaboration with David Martinez-Delgado for the Stellar Tidal Stream Survey]

The authors of today’s paper look at one stream in particular, GD-1. This nearby, long stream is on an orbit that makes it ideal for constraining the properties of dark matter in the Milky Way. It also provides an excellent example for how Gaia velocities can enhance our understanding of these exciting systems.

So, how do the velocities from Gaia help in the study of stellar streams? First of all, they help to trace out the path that the stellar streams follow as they orbit around the Milky Way. This improves our understanding of the formation and evolution of the stellar streams, and thereby tightens our constraints on the local distribution of dark matter. Second of all, and perhaps less obviously, they can also help with picking out the stars that belong to these streams, clearing up our view of the systems. These objects are far out in the stellar halo of the galaxy, beyond the extent of the central disk (Figure 2), which means that their velocities are distinct from the coherently rotating disk of the galaxy. By selecting only a small range of velocities we can throw away many of the contaminating foreground stars.

Figure 3 shows this process for the stellar stream GD-1. The right-hand panel of Figure 3 shows velocities along (x-axis) and perpendicular to (y-axis) the length of the stream in the plane of the sky. The little clump of stars highlighted in orange corresponds to the stellar stream and nicely stands out from the larger clump corresponding to all of the other Milky Way stars in the region. As expected, the stream stars aren’t moving much perpendicular to the stream (close to 0 along the y-axis), but are mostly moving along the length of the stream, which traces out their orbit around the galaxy. The left panel shows the positions of the stars selected within the orange box, and the stellar stream clearly stands out!

Figure 3. Left: Positions of stars with velocities near GD-1. The thin stream stands out clearly. Right: Velocities along (x-axis) and perpendicular to (y-axis) the stream on the sky. The orange box outlines a clump of stars corresponding to GD-1. [Price-Whelan & Bonaca 2018]

This is already quite impressive, but our view of this stellar stream really opens up when the Gaia astrometry (the measurement of precise positions and velocities) is combined with photometry (the measurement of light) from other experiments! In Figure 4, the authors have matched the Gaia data with data from Pan-STARRS to consider the brightness of these stars as a function of color (right panel). These properties are related to the age and chemical composition of stars, allowing the authors to select out particular stellar populations. The left-hand panel of Figure 4 shows how clearly the stream stands out when these two types of data are used in tandem.

Figure 4. Left: Positions of stars likely to belong to GD-1 based on astrometry and photometry. This is the most clear view of GD-1 to date. The gray dashed line indicates the portion of the stream that was previously undetected, before Gaia. Other interesting features with unusually high (blobs, spurs) or low (gaps) numbers of stars are labeled along the stream. Right: Color (x-axis) and brightness (y-axis) of stars around GD-1. The orange selected region corresponds to a stellar population with a similar age and chemical composition to the stellar stream. [Price-Whelan & Bonaca 2018]

Now, what exactly have we gained from this clear view of GD-1? First, this is the most complete view of the stream to date. A large segment of the stream is visible that had previously been hidden among the foreground stars (left of the gray line in Figure 4), extending the length of the stream by 20 degrees across the sky.

Second, variations in density along the stream are immediately apparent. Figure 4 points out high-density blobs and low-density gaps, which potentially indicate very exciting science! The blob of stars may be the progenitor — the original cluster of stars that is gradually being torn apart to form the stream. The location of the progenitor is an essential piece of information for accurately modeling the formation and the orbit of this stellar stream, which can be used to study the large-scale distribution of dark matter in our galaxy. The low-density gaps are perhaps even more exciting. As discussed in previous Astrobites, small invisible clumps of dark matter can punch holes through stellar streams, leaving behind gaps. Further work will reveal the nature of these intriguing density features along GD-1, perhaps leading to new insights into the nature of dark matter.

This paper illustrates the power of combining Gaia velocities with photometry to study stellar streams, with the excellent example of GD-1. This procedure can also be applied to a variety of systems in the galactic halo, like dwarf galaxies and globular clusters, to expand our understanding of the Milky Way halo as a whole, ultimately informing theories on the nature of dark matter and the evolution of the Milky Way. This exciting new era of Gaia has only begun — stay tuned to see what other unexpected science this new view of our galaxy will reveal.

About the author, Nora Shipp:

I am a 2nd year grad student at the University of Chicago. I work on combining simulations and observations to learn about the Milky Way and dark matter.

Rho Ophiuchi

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: Compact Dusty Clouds and Efficient H2 Formation in Diffuse ISM
Author: A. V. Ivlev, A. Burkert, A. Vasyunin, and P. Caselli
First Author’s Institution: Max-Planck-Institut for Extraterrestrial Physics, Germany
Status: Accepted to ApJ

An Element of Epic Proportions

Figure 1: A filament of the Taurus Molecular Cloud, which contains both (1) stars that are newly formed and (2) stars that have yet to form. [ESO/APEX (MPIfR/ESO/OSO)/A. Hacar et al./Digitized Sky Survey 2. Acknowledgment: Davide De Martin.]

The simplest known element in the universe, hydrogen, is also the most important. Hydrogen (aka “H”) plays a crucial role for many of the awesome astrophysical phenomena that have happened (and will happen!) across the universe’s history. This element fuses in the cores of stars, lights up ionized regions in supernova remnants, and serves as a building block for other elements in the periodic table — just to name a few of its many talents.

But that’s not all this epic element can do!  Its molecular form, H2, is the primary ingredient for molecular clouds, like the one shown in Figure 1. These molecular clouds are made up of both gas and dust grains. They are often found within spiral galaxies (like our own Milky Way), and they’re interesting in part because they’re the only known sites where glorious star formation occurs.

To better understand molecular clouds and their star formation, scientists have long studied the timescales for these clouds to form within the less dense, diffuse interstellar medium (a collective term for all of the matter and radiation between star systems) of space.

Classic calculations of these timescales have often assumed that the diffuse interstellar medium (aka, ISM) is pretty homogeneous, even when we zoom in. In other words, they assumed that the gas and dust in the diffuse ISM are spread out uniformly, so that every bit of the diffuse ISM looks the same as every other bit. But observations from the last 50 years or so indicate that the diffuse ISM is actually very not homogeneous, with lumps and bumps even across small (relatively speaking) solar-system-sized regions.

There have been numerous studies on what could be causing this non-uniformity at such small scales. But today’s authors look at a possible mechanism that is particularly significant for H2: the formation of tightly-packed dusty clouds and equilibrium gaseous clumps in the diffuse ISM. We’ll dive into what these clouds and clumps are made of in the next section.

Forming Clouds and Clumps

The authors discuss two main bits of physics that could form these aptly-named dusty clouds and gaseous clumps:

Major Physics #1: Attractive shadowing forces between dust grains in the diffuse ISM. These mysterious-sounding forces refer to the collective interactions that can happen between dust grains within a gas. In the diffuse interstellar medium, these forces are attractive: they pull the dust grains towards each other like little magnets, gathering them together into dusty clouds within the diffuse ISM: tightly-packed (aka, compact) dust grains with gas squeezed in between the grains.

Major Physics #2: Efficient thermal coupling between the dust and gas within these dusty clouds. The fancy term in italics here basically means that the gas within a dusty cloud adapts the same temperature as the dust within the dusty cloud. But the temperature of the dust within the cloud, Tdust, is lower than the temperature of the gas outside the cloud, Tgas. That means the gas within the cloud decreases in temperature from Tgas to Tdust; to do that (without breaking the laws of physics), the gas within the cloud also increases in density relative to the gas outside of the cloud. Then we end up with a gaseous clump within and just around the dusty cloud: gas squeezed into the space between the dust grains that is denser than the gas outside of the dusty cloud.

illustrations of the diffuse ISM

Figure 2: Illustrations of the diffuse ISM, if the diffuse ISM were actually homogeneous (left panel) and if the diffuse ISM formed compact dusty clouds (right panel). On the left, we see that the dust (the black dots) and the gas (the pale blue background) are pretty uniformly distributed across the medium. The temperature of the dust grains (Tdust) is smaller than the temperature of the gas (Tgas). On the right, we see dust (again the black dots) has gathered together to form dusty clouds, which are much denser than dust in the surrounding diffuse ISM. The gas has also gathered in gaseous clumps (shaded in dark blue) within and around the edges of the dusty clouds. Here the temperature of the gas within the dusty clouds is about equal to that of the dust within the dusty clouds (Tgas(cl)), and both temperatures are smaller than the temperature of the surrounding gas (still Tgas). [Ivlev et al. 2018]

By these two physics processes, we end up with diffuse ISM that is studded with compact dusty clouds/gaseous clumps, as illustrated in Figure 2.

It All Comes Together

With these dusty clouds embedded within, the diffuse ISM is definitely not homogeneous. This means that the chemistry happening within the diffuse ISM is also not homogeneous, and will unfold differently for the dusty clouds than it will for the surrounding diffuse ISM.

The authors derived how the rate of H2 formation would change due to these dusty clouds and equilibrium gaseous clumps. They found that the new rate of formation, compared to the rate if the diffuse ISM was actually homogeneous, increases by a factor ranging from about ~5–10 for typical gas and dust temperatures.

For the typical diffuse ISM values that the authors explored, this increased rate means that the transition from H to molecular H2 in the diffuse ISM could happen in just a few million years. The authors point out that this reduced timescale has wide astrophysical applications… but first and foremost, it affects how we expect giant molecular clouds, and thus their glorious star formation, to physically evolve. The faster transition from H to H2 could also affect other chemical species in the diffuse ISM, not just our epic element hydrogen. But we’ll need more investigation to see for sure!

About the author, Jamila Pegues:

Hi there! I’m a 2nd-year grad student at Harvard. I focus on the evolution of protoplanetary disks and extra-solar systems. I like using chemical/structural modeling and theory to explain what we see in observations. I’m also interested in artificial intelligence; I like trying to model processes of decision-making and utility with equations and algorithms. Outside of research, I enjoy running, cooking, reading stuff, and playing board/video games with friends. Fun Fact: I write trashy sci-fi novels! Stay tuned — maybe I’ll actually publish one someday!

EPIC view of Earth

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: Using Deep Space Climate Observatory Measurements to Study the Earth as An Exoplanet
Author: Jonathan H. Jiang, Albert J. Zhai, Jay Herman, et al.
First Author’s Institution: Jet Propulsion Laboratory, California Institute of Technology
Status: Submitted to ApJ

The search for habitable, Earth-like planets is high on the agenda of exoplanetary scientists. However, habitability is far more complicated than checking that a planet sits within the goldilocks zone of not-too-hot, not-too-cold temperatures, just right for liquid water to exist on its surface. Dozens of other factors — such as the planet’s atmosphere, seasons, and surface geography — are of critical importance to sustaining a watery planet and creating somewhere suitable for life.

Many previous astrobites have covered studies investigating exoplanetary atmospheres via transmission spectroscopy, but today’s paper takes a slightly different angle by considering the possibilities of direct imaging. Using our own habitable Earth as a “proxy exoplanet”, the authors inspect multi-wavelength observations from the Deep Space Climate Observatory, or DSCOVR, to figure out what we should be looking for further from home.

DSCOVR is largely used for space-weather monitoring, but its Earth-facing instrument EPIC, the Earth Polychromatic Imaging Camera, takes pictures of the sunny side of our planet from Lagrange 1 in ten wavelengths, from the ultraviolet to near-infrared (if it’s not already in your twitter feed, check out the DSCOVR:EPIC bot for a daily dose of pale blue dot).

Observing the Earth as an exoplanet is not a new idea — but DSCOVR has an advantage over many other Earth-observing missions in that the data span a long period of time. The authors analyse over two years’ worth of data from EPIC. By looking at how these images change with time, on periods from hours to years, they work out the kind of imaging we would need of distant exoplanets in order to deduce their rotation periods, seasonal changes, weather, and surface types.

Figure 1: EPIC images of the Earth in 10 different wavelengths. Cloud cover is bright in every wavelength, and continents are brighter in the longest wavelengths: Africa stands out at 779.5 nm. [Jiang et al. 2018]

There are a number of things to be learned from these multi-wavelength images (Figure 1). The reflectance of different surface types varies with wavelength: the ocean, for example, reflects most in the blue-green wavelengths, and vegetation reflects strongly in the near-infrared. So by analysing differences in reflection between long and short wavelengths, it may be possible to identify different types of surface coverage on exoplanets. Clouds reflect light across the spectrum, and the authors go on to separate cloud cover by simply subtracting the emission in a wavelength where there is little reflectance from other components from the other images.

An obvious difference between EPIC images and potential direct imaging of exoplanets is the glorious resolution: each EPIC pixel images a 12×12 km region of the Earth’s surface, compared to the unresolved point source we would detect if we were to place the Earth around our nearest neighboring star, Alpha Centauri. To simulate exoplanet imaging, the authors average their data into one pixel, and then look at the time evolution of the signals in each wavelength.

Figure 2: Time-series data for the “single-point” measurements. On the left are the measurements over the course of a single day (filled circles show February 8th, 2017, and empty circles show August 8th, 2016). On the right is the entire dataset spanning over 800 days. [Jiang et al. 2018]

Figure 2 shows those time variations. Peaks can be seen periodically, from features such as Pacific ocean clouds reflecting on a daily cycle, and seasonal changes depending on the part of the Earth facing EPIC.

Figure 3: Periodograms, showing peaks at the periodic frequencies in the data: notably, the daily cycle at 24 hours, and the yearly at 365 days. (d) shows the signal split into the contribution from clouds, in blue, and from surface features in green. Figure 5 in the paper.

The authors investigate these periodic changes with a Fourier analysis. The periodograms in Figure 3 simply show peaks at the frequencies of periodic features in the data. There are some obvious ones: a peak at 24 hours, and one at 365 days. These come from the daily patterns of cloud and continent passing through the images, and a combination of annual climate cycles such as monsoons as well as the tilt of the Earth. Some periodicities are less obvious: the peak at 12 hours is attributed to the Pacific and Atlantic oceans passing through the field of view 12 hours apart. The longer periodicities, at 90 and 180 days, are most likely due to DSCOVR’s orbital period, but they may also show trends in the growth cycle of crops. By looking at the relative strengths of the peaks across different wavelengths, we can deduce the features responsible for different cycles to build up a picture of the surface composition of the planet.

Another difference between these images and exoplanetary imaging is the phase: EPIC’s view of the Earth is always of a fully sunlit planet. The authors simulate the effect of viewing an exoplanet orbit edge-on, as any exoplanets detected by the transit method would be, by putting artificial phases on the images (figure 4). They successfully recover the same features in the periodogram from these phased images. They also estimate the minimum amount of data required to detect the rotation period, showing that for the Earth, a measurement needs to be made at least every 9 hours.

Figure 4: Artificially “phased” images, to simulate measurements of an exoplanet viewed with its orbit edge-on. [Jiang et al. 2018]

In practice, directly imaging exoplanets is incredibly challenging, and it has only been done on a handful of objects. While we might not be there yet, the techniques explored here could be used to guide future mission design for investigating exoplanet habitability, and they could open up new prospects for time-evolution, multi-wavelength studies of distant worlds.

About the author, Joanna Ramasawmy:

I’m in the second year of my PhD in observational extragalactic astrophysics at the University of Hertfordshire. More specifically, I’m looking into the relationship between supermassive black holes and star formation in the galaxies that host them. In my spare time, I’m learning to make ceramics and to climb rocks!

diffraction pattern

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: Radio interferometric observation of an asteroid occultation
Author: Jorma Harju, Kimmo Lehtinen, Jonathan Romney et al.
First Author’s Institution: University of Helsinki, Finland; Max Planck Institute for Extraterrestrial Physics, Germany
Status: Submitted to AJ

For decades astronomers have used occultations to study celestial bodies such as the moon, asteroids and even quasars. An occultation occurs when some foreground object — the moon, a planet, or an asteroid — crosses in front of a background object — the Sun (in the case of an eclipse), a planet, a star, or a quasar. While rare, these occurrences can tell us a lot about the foreground and/or background objects. For example they have been used to determine the positions of quasars (before radio interferometers were available), detect binary stars, study surface topography of the moon, the atmosphere of the outer planets, and the shapes and sizes of asteroids (including New Horizons‘s next target). Today’s paper uses an intriguing technique to analyze an occultation of an active galactic nucleus (AGN) by an asteroid observed with the Very Long Baseline Array (VLBA).

On May 15, 2017 at UT 14:31:23 an asteroid named Palma crossed between AGN 0141+268 and the Brewster VLBA station. Figure 1 shows the path of the shadow across the northwestern United States. Normally when observing an occultation you would expect to simply see the background source disappear for some period of time and then reappear from behind the occulter — but that was not quite the case for this observation. This observation was special because of the relative size of the asteroid, the distance to the asteroid, and the wavelength of the observation. In this instance, scientists Jorma Harju and collaborators observed not just the shadow of the asteroid as it occulted the AGN, but also the diffraction pattern.

Figure 1: Path of the shadow cast by Palma across the Brewster VLBA station. Grey eclipses indicate the geometric shadow at intervals of one second. Orange eclipses indicate the location of the first maxima of the diffraction pattern four seconds before and after the closest approach. [Harju et al. 2018]

This is essentially the classic diffraction experiment of shining a laser on a small ball bearing, but on a much larger scale and under less-than-ideal conditions. These conditions, however, are precisely what allow us to learn about the asteroid. Since the asteroid is not a perfect sphere, its diffraction pattern is asymmetric; this asymmetry can then be used to infer the shape of the asteroid. The asymmetry can clearly be seen in the bottom center panel of Figure 2.

Since the occultation was observed with a radio interferometer, Harju and collaborators observed not only how the intensity of light changed, but also its phase. This was the first time a measurement of the phase of astronomical diffraction has been measured. This phase information, combined with the amplitude in an amplitude-phase diagram (shown in the bottom left panel of Figure 2), plays a crucial role in determining the size and shape of the asteroid. The authors fit a number of models (circle, ellipse, two overlapping circles they call potato-shaped, a random continuous shape, and a model derived from visible-light occultations) to the combined amplitude and phase observations. With the given data it is difficult to discern the exact shape of the asteroid, though they did find a diameter of 192 km (consistent with previous observations), and determined that the asteroid’s shape deviates 10–20% from a circle.

Figure 2: Top: model of the shape of the asteroid Palma (a), and the amplitude (b) and phase (c) of the resulting diffraction pattern. Bottom: amplitude phase diagram (d) showing the data points and modeled curves (blue and red correspond to ingress and egress, respectively), and amplitude (e) and phase (f) cuts along the path of the Brewster VLBA station. Data are shown in thick lines, models in thin lines and residuals at the bottom. [Harju et al. 2018]

An interesting effect of diffraction around a circular occulter, in this case an asteroid, is that there is a bright spot at the center of the shadow (seen in the center of the two center panels, b and e, in Figure 2). Since the center of the shadow is equidistant from all points on the edge of the sphere, light constructively interferes and creates a light spot in the middle of the dark shadow. This spot, called the Arago-Poisson spot, was an early point of contention when the wave theory of light was first being developed. The measurement of this spot tells you how far you are from the center of the shadow; in other words, how close the center of the occulter is to lining up with the light source. In this case, this corresponds to the exact position of the asteroid relative to the background AGN during the occultation. This extremely precise position measurement can be used to refine the orbit of the asteroid, reducing the long term uncertainty in its position by an order of magnitude.

While the concept of diffraction and the odd effects of the wave nature of light can be difficult to wrap one’s head around, they are important and powerful tools that can be leveraged to perform amazing science from spectroscopy (diffraction gratings) to high-resolution imaging by combining light from multiple telescopes (interferometry).

About the author, Samuel Factor:

Sam Factor is a 3rd year Ph.D. candidate at The University of Texas at Austin studying direct imaging of extrasolar planets and low mass binary stars. He uses an interferometric post processing technique to allow the detection of companions below the diffraction limit of the telescope.

Robo-AO

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: Robo-AO Kepler Planetary Candidate Survey IV: The effect of nearby stars on 3857 planetary candidate systems
Author: Carl Ziegler, Nicholas Law, Christoph Baranec, et al.
First Author’s Institution: University of North Carolina at Chapel Hill
Status: Published in AJ

Introduction

In the 1970s and 80s, scientists working for the U.S. Department of Defense developed a secret technology for imaging Soviet spy satellites from the ground. One of the consultants, Claire Max, realized the technology could benefit astronomy and pushed for its declassification. The concept of adaptive optics with laser-powered guide stars was finally released to the public in 1991.

Adaptive optics (AO) involves the rapid deformation of a mirror to remove distortions from an image. Imagine seeing a rock at the bottom of a clear but fast-moving stream of water. Inverting the jittery, garbled image into a true image of the rock is a significant engineering and computational challenge. First, we need to observe a simple point source behind the same distorting medium.

This is where laser guide stars come in. One type of laser lights up a thin layer of sodium at about 90 km altitude. This creates a “guide star” that experiences much of the distortion seen in a true star. Using this information, a deformable mirror can correct the true star’s image in real-time.

But large AO systems are extremely unwieldy. They require lots of money and manpower, and lots of observing overhead — it can take several minutes to slew to a target, lock onto a guide star, and initialize a correction loop for deforming the mirror. And yet, in today’s era of “big data” astronomy, AO imaging has become ever more relevant for rapidly-growing datasets. There’s a niche for fast, robotic AO. Recently, the optical scientist Christoph Baranec has done just that by building… Robo-AO.

improvement from Robo-AO

Figure 1: Left: An example image of Kepler’s view of a KOI. The large pixels hide an unknown number of stellar companions. Right: The same view, using Robo-AO. [Ziegler et al. 2017]

Today’s Paper

The Kepler space telescope has detected thousands of possible exoplanets transiting in front of their host stars. But the Kepler pixels are huge — about 4 arcseconds across — and there’s always a risk that multiple stars are lurking in a single pixel (Figure 1). This can lead to false positives (like in eclipsing binary systems) or can dilute real planet signals (where the light from another star will make a transiting planet’s radius seem smaller than it actually is). This calls for AO follow-up.

Before Robo-AO, AO follow-up of Kepler objects of interest (KOIs) has been a heterogeneous and incomplete effort using different instruments, reduction pipelines, and analyses. To establish a consistent, complete, and unbiased analysis, Robo-AO has targeted Kepler stars from Palomar Mountain in California and Kitt Peak in Arizona. (Check out an introductory video about the project here, and a time-lapse of Robo-AO hammering targets with a UV laser here.) The most recent paper to be published on the accumulating Robo-AO Kepler data reports on 3857 KOI observations, taken at a rapid-fire cadence of 20 targets per hour.

Robo-AO KOI images

Figure 2: A beauty parlor gallery of Robo-AO KOI images taken at Kitt Peak. [Ziegler et al. 2018]

Within their ranges of sensitivity, the authors found that about 17% of the stars had stellar companions within 4″. In total, they found 610 stellar companions to 559 KOI stars. The authors calculate that if the candidates orbit the target star, then the planet radii have to be multiplied, on average, by a factor of 1.08. If they orbit a bound stellar companion, their radii increase by a factor of 3.29! A total of 35 supposedly rocky planets have newly revised radii of >1.6 Earth radii in either scenario, meaning that they are gaseous giant planets, not rocky ones.

rate of stellar companions

Figure 3: Rate of stellar companions based on separation. Unassociated stars would be expected to increase roughly like a parabola (dashed line), but the observed number is greater, meaning that some of the stellar companions must be gravitationally bound and possibly influence exoplanet types which emerge in the system. [Ziegler et al. 2018]

Planets in Multistellar Systems?

Models suggest that stellar companions could severely perturb planets’ orbits or fling planets out of a system entirely. To determine the effect of stellar companions on planet formation, we first have to determine which of the KOI companions are gravitationally bound with each other, and how many are unbound but are just along close lines of sight.

The authors argue that if companions were unbound, then we would expect a field of view twice as large to lead to about four times as many found companions. But in fact, found companions increase linearly, suggesting that some of the stars are indeed bound (Fig. 3). The authors are preparing another paper that will take a deeper dive into the nature of these companions, which will bring us closer to understanding the effect of multiple stars on planetary systems.

In the meantime, Robo-AO is continuing to pursue many different science cases. And a southern Robo-AO unit at Cerro-Tololo is under development, as is a Robo-AO 2 instrument in Hawaii. Soon we can observe most of the sky with Robo-AO’s automated gaze. Unless, that is, the robots finally rise against us.

About the author, Eckhart Spalding:

I am a graduate student at the University of Arizona, where I am part of the LBT Interferometer group. I went to college in Illinois, was a secondary-school physics and math teacher in Kenya’s Maasailand for two years, and got an M.S. in Physics from the University of Kentucky. My out-of-office interests include the outdoors, reading, and unicycling.

magnetar

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: Sites That Can Produce Left-Handed Amino Acids in the Supernova Neutrino Amino Acid Processing Model
Author: Richard N. Boyd, Michael A. Famiano, Takashi Onaka, and Toshitaka Kajino
First Author’s Institution: The Ohio State University
Status: Published in ApJ

Scientists are enamored with the search for life. They’ve scoured spectra for hints of life-supporting gases in the atmospheres of exoplanets. They’ve assessed the friendliness of galaxies near and far, and found the universe to be an unforgiving place. They’ve plumbed the depths of the oceans and studied pristine Antarctic lakes to understand the harsh conditions life might be able to withstand elsewhere in the cosmos.

The search for life on other worlds has been guided by what we know about life on Earth. Life as we know it depends on amino acids to survive. Today’s paper explores the effects of exotic astrophysical settings on amino acids, which could tell us something about how life came to be on Earth — and where else in the universe life like Earth’s might be found.

The Stuff of Life

Amino acids — important components of proteins — have a property called chirality or handedness. Much like left and right hands, which are mirror images of each other, chiral molecules come in left- and right-handed pairs. These pairs of molecules contain the same atoms and the same chemical bonds, but no amount of turning, bending, or wiggling can make them look exactly the same, as Figure 1 shows.

alanine

Figure 1. A simple amino acid, alanine. The solid triangle indicates that that part of the molecule sticks out of the plane of the page, while the dashed triangle indicates that that part of the molecule is directed backward, into the page. The two forms of alanine, left-handed (left-most illustration) and right-handed (center illustration) alanine, are a chiral pair. Right-handed alanine can’t be converted to left-handed alanine just by turning the molecule over (as illustrated in the right-most illustration).

Right-handed amino acids are rare in nature, and all multicellular Earth life depends on amino acids that are left-handed (with the exception of glycine, which is the same as its mirror image). This is weird. If you whip up a batch of your favorite amino acid in a lab (such as in the famous Miller-Urey experiment and subsequent experiments that demonstrated that lightning could kick off the formation of amino acids in the primordial soup), the molecules will be roughly 50-50 left- and right-handed — so why does life have a preference for left-handed amino acids?

While many scientists believe that the preference for left-handed amino acids arose naturally on Earth, the how is still unclear. Another option is that the bias toward left-handed molecules was introduced from afar — through amino-acid-bearing meteorites crashing to Earth, for example. Many meteorites contain a small excess of left-handed amino acids, which tells us that some process in space can create an excess of left-handedness in chiral molecules.

Where Does Left-Handedness Come From?

The authors of today’s paper suggest that a preference for left-handed amino acids can be caused by a combination of a strong magnetic field and a source of electron antineutrinos.

First, assume that amino acids are orbiting an astrophysical object that emits lots of antineutrinos and has a strong magnetic field, such as an expanding supernova, a massive Wolf-Rayet star, or a newly born neutron star. These amino acids could be present on the surface of dust grains or embedded within meteoroids or planets. Then, as the molecules are pummeled with antineutrinos (\bar{\nu} _{e}), some nitrogen atoms are converted to carbon atoms following this process: \bar{\nu} _{e} +\, _{ }^{14}\textrm{N} \rightarrow e^{+} +\, _{ }^{14}\textrm{C} .

magnetic field of a neutron star

Figure 2. An illustration of the magnetic field of a neutron star. The magnetic field orientation (B) and the antineutrino and nitrogen spins (Sν and SN) are shown with light blue arrows. [richardboydastro.com]

If the nitrogen atom is converted to a carbon atom, the molecule will no longer be an amino acid, since amino acids are defined as having an amine (NH2) and a carboxyl (COOH) group. The likelihood of this process occurring depends on how the spins of the antineutrino and the nitrogen atom are aligned; if the spins are parallel, as shown on the left-hand side of Figure 2, the process is much less likely to happen than if the spins are misaligned.

Here’s where the magnetic field comes in: combined with the motion of the molecules around the star, the magnetic field works to align the nuclear spins of the nitrogen atoms. For left-handed amino acids, the nuclear spin is parallel to the antineutrino spin — so left-handed amino acids are more likely to survive the onslaught of neutrinos. The effect isn’t huge, since most neutrinos pass through without interacting, but the authors estimate that this process could cause a 1% excess of left-handed amino acids — comparable to the few percent excess seen in meteorites. Enough, perhaps, to imprint a bias toward left-handed molecules on the young Earth.

Are We All Aliens, Then?

Does life on Earth owe its origins to meteorites processed by antineutrinos and magnetic fields? Maybe, maybe not. There are many hypotheses for how life acquired its taste for left-handed amino acids, and they don’t have to be mutually exclusive. One thing is for sure, though: we have a lot to learn about chiral molecules — and amino acids in particular — in space. While glycine, an achiral amino acid, has been discovered everywhere from the interstellar medium to the atmospheres of comets, the first chiral molecule in space was discovered only in 2016. Perhaps the first chiral amino acids will be discovered in the tumultuous region around a newborn neutron star. If so, Earthlings can look on from afar and give them a wave. Left-handed, of course.

About the author, Kerrin Hensley:

I am a third year graduate student at Boston University, where I study the upper atmospheres and ionospheres of Venus and Mars. I’m especially interested in how the ionospheres of these planets change as the Sun proceeds through its solar activity cycle and what this can tell us about the ionospheres of planets around other stars. Outside of grad school, you can find me rock climbing, drawing, or exploring Boston.

red dwarf flare

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org!

Title: The First Naked-Eye Superflare Detected from Proxima Centauri
Author: Ward S. Howard, Matt A. Tilley, Hank Corbett, Allison Youngblood, R. O. Parke Loyd, et al.
First Author’s Institution: University of North Carolina at Chapel Hill
Status: Submitted to ApJL

Proxima Centauri is the closest known star to the Sun at just 4.246 light-years (1.302 parsecs) away. It’s a red dwarf of spectral type M6 with about 12% of the Sun’s mass, 1.2 times the diameter of Jupiter, and 0.17% of the Sun’s luminosity. It hosts the closest known exoplanet to us, Proxima Centauri b, which was discovered in 2016 as covered in this Astrobite. Like our Sun, it’s on the main sequence, steadily fusing hydrogen into helium in its core. Yet this tiny star is way more active than the Sun is!

Red dwarfs like Proxima Centauri have interiors that are fully convective, meaning that the energy generated by fusion in their cores is transported to the surface primarily via convection. Like a pot of boiling water, you can think of it as being one giant ball of boiling plasma. This turnover of ionized gas generates powerful magnetic fields, which are carried to the surface along with the bubbles of hot plasma. When these bubbles reach the surface the energy contained in the magnetic fields can be violently released in the form of stellar flares, which can grow as large as Proxima Centauri itself and reach temperatures of up to 27 million K! (Normally its effective surface temperature is around 3,000 K.) These flares from Proxima Centauri have been observed frequently in the past (for instance in this recent Astrobite).

Figure 1: The light curve of Proxima Centauri as seen by the Evryscope around the time of the superflare. Three weaker (but still strong) flares were detected in the aftermath of the superflare, marked by arrows. [Howard et al. 2018]

Erupting With (Super)Flair

In this paper, the authors report the discovery of the first-known superflare from Proxima Centauri (see Figure 1 for the light curve), a flare roughly ten times more powerful than any seen before. Normally Proxima Centauri sits at a visual magnitude of 11.13, approximately 100 times fainter than the human eye can see. But during the superflare the authors calculated that it would have reached an apparent visual magnitude of 6.8 for a few minutes, just bright enough to be seen with the naked eye in extremely dark skies! (No accounts of anyone actually seeing it by eye at the time are known, though.)

To discover this superflare the paper authors used data from the Evryscope, which we’ve covered in this Astrobite. Despite the name it’s not actually a single giant telescope formed out of every telescope on Earth (sadly), but is instead a unique collection of twenty-seven small telescopes all mounted on a single German Equatorial mount in Chile, one of its goals being to catch and record these short-duration transient events. It pans across the sky throughout the night taking images of 8,000 square degrees of the southern night sky simultaneously every two minutes, all night long.

On March 18, 2016, at 8:32:10 UT the Everyscope detected a superflare that lasted for over an hour, though the bulk of the energy was emitted in the first ten minutes. The authors estimate that the total energy radiated at all wavelengths was 1033.5 ergs, ten times more than any previously seen flare. However, based on the many other, less powerful flares observed by the Evryscope, the authors estimate that flares with an energy release of 1033 or more ergs probably occur around 5 times per year.

Ozone? More Like NO-zone

The authors also investigated the effects of so many flares on a hypothetical atmosphere of Proxima Centauri b by running a 1D atmospheric simulation in which the planet is assumed to have an Earth-like atmosphere.  Strong proton fluxes from coronal mass ejections associated with flares can destroy ozone by first breaking nitrogen (N2) apart into nitrogen atoms that react with oxygen (O2) to form NO and O. The NO then reacts with ozone in a catalytic reaction to form NO2, depleting the ozone (O3) layer in a very efficient manner.

The simulation generated a series of flares with a range of energies compatible with what has been observed in the past, which interacted with the model atmosphere over a simulated five-year period. Each flare had an 8% chance to have a strong proton flux associated with it (based on other work). Even this low chance of producing strong proton fluxes, however, depleted the ozone layer by 90% within five years (as shown in Figure 2). Thus it seems highly probable that Proxima Centauri b has no ozone layer to speak of.

Figure 2: The depth of the ozone column in the simulation performed in the paper. The two dashed lines denote one year and five years after starting the simulation. Flares were stopped after five years in the simulation which leads the solid line to recover back up to full, but the authors note that the dashed line is more likely in reality where flares continue. [Howard et al. 2018]

The amount of ultraviolet light reaching the surface in the absence of any ozone in the atmosphere spells bad news for any living thing unlucky enough to be in its path. The superflare over its duration would have deposited an estimated 3.6 J/cm2 of the most dangerous UV-C light to Proxima Centauri b, which is some 65 times greater than the amount needed to kill off the most radiation resistant organism known, the bacterium Deinococcus radiodurans.

Summary

Proxima Centauri is a very active star, and with the Evryscope up and running we’re in a good position to catch the next superflare it gives off (along with any regular flares). And if you were planning a beachside vacation to Proxima Centauri b, you may want to hold off until you find a sunscreen with an SPF of a million or so.

About the author, Daniel Berke:

I’m a first-year grad student at Swinburne University of Technology in Melbourne, where I search for variation in the fine-structure constant on the Galactic scale. When I’m not at uni I enjoy a variety of creative enterprises including photography, blogging, and video editing, or just relaxing with a good video game or some classical music.

Earth-like exoplanet

Editor’s note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org! Normal AAS Nova posting will resume on Friday.

Title: A Revised Exoplanet Yield from the Transiting Exoplanet Survey Satellite (TESS)
Author: Thomas Barclay, Joshua Pepper, Elisa V. Quintana
First Author’s Institution: NASA Goddard Space Flight Center
Status: Submitted to AAS Journals

Exoplanet hunters around the world held their breath while NASA’s Transiting Exoplanet Survey Satellite (TESS) launched last Wednesday. Thankfully the launch was a success, and after 60 days of orbit manoeuvring and engineering tests, TESS is expected to begin its initial two years of science observations. The question is: how many planets do we expect TESS to find?

How To Count Chickens Before They Hatch (or Planets Before They Are Detected)

The TESS mission will observe 90% of the sky to find nearby unknown planets. However, to get mission funding, astronomers need to predict how many planets they expect to identify.

The authors of today’s paper took on this challenge with a three-step modelling plan: 1) predict which stars would be observed, 2) randomly assign planets around them and 3) test if they are detected.

  1. Stars Observed
    Determining which stars TESS is likely to observe is made easier thanks to the Candidate Target List, a ranked list of 3.8 million stars most suitable for detecting small planets (where “small” means a planet with radius smaller than 4 Earth radii). The Candidate Target List contains isolated dwarf stars brighter than TESS magnitude 13, which will be less blended in the giant TESS pixels (each pixel contains 21 arcseconds of sky, compared to 4 arcseconds in Kepler). The authors calculate how long TESS can observe these stars and those most likely to be priority targets. Data for priority targets will be available as observations every two minutes (at a two-minute cadence) whereas other stars will only have data in the full-frame images every 30 minutes.
  2. Planet Assignment
    Each star in their list is assigned 0 or more planets according to a Poisson distribution. Each planet is then given random properties, including inclination, orbital period and radius based on the general trends found by the Kepler Space Telescope. Periods are drawn from distributions between 0.5–85 days for planets orbiting A/F/G/K stars and between 0.5–200 days for planets with M-star hosts. Time of the first transit is then drawn randomly between 0 and the length of the period, which can be after observations finish, meaning no transits will be seen.
  3. Detection test
    Finally, the authors test whether the transit signal is significantly stronger than the noise. The signal strength is determined by considering the number of transits, the transit depth and duration, and the extent of contamination from nearby stars. If the signal is greater than 7.3 times the TESS photometric noise level (7.3 SNR) and at least two transits are seen, this optimistic model claims a detection.

Optimistic Planet Numbers

Figure 1: Planet numbers detected by radius using the optimistic model. Red bars indicate the numbers of planets detected using 2 minute cadence data. Numbers above blue bars show the combined number of planets found in 2 minute cadence or full frame images. Note the log scale of planet numbers. [Barclay et al. 2018]

The optimistic model above identifies ~4,500 planets around stars in the Candidate Target List, shown split into planet radius in Figure 1. For stars with V magnitude brighter than 12, the authors predict the detection of 1,317 small planets. If 20% of these are amenable to radial velocity follow up, this would triple the number of small planets with measured masses and exceed the TESS main science objective of identifying 50 small planets with measurable masses.

Predicting planet numbers means astronomers can plan the number of follow-up observations necessary to confirm planets. Not all transit-like signals will be due to transiting planets; they may be caused by instrumental effects or astrophysical false positives, such as deep eclipsing binary signals blended to give shallower transits, especially with multiple stars on the same pixel. Today’s paper predicts for every true planet found there will be one astrophysical false positive in the 2-minute cadence data, and 5.5 astrophysical false positives in the full-frame images.

More Conservative Planet Numbers

Figure 2: Planet numbers detected by radius using the conservative model. Green bars indicate the numbers of planets detected using 2 minute cadence data. Numbers above orange bars show the combined number of planets found in 2 minute cadence or full frame images. Note the log scale of planet numbers. [Barclay et al. 2018]

Identifying planets based on fewer than three transits and detecting all planets with SNR ≥ 7.3 is very difficult. The Kepler Space Telescope is capable of identifying planets from one or two transits in K2 data, but only with additional investment of limited space-based follow up, or in cases where other planets had already been discovered around that star. Analysis of Kepler data found that below SNR = 8–10, there were many spurious detections, so typically only targets with SNR > 12 were followed up.

In their conservative model, removing planets with fewer than three transits or SNR < 10 reduced the number of planets detected by 60%, as seen in Figure 2. The number of small planets detected around stars brighter than V = 12 halved to 621. The number of habitable-zone planets smaller than 2 Earth radii drops to just 6.

Adding to James Webb Space Telescope Targets

Any small planets detected by TESS may represent new targets for atmospheric characterisation with the James Webb Space Telescope (JWST). Figure 3 shows that the simulated TESS planets greatly increase the number of known small nearby planets, some of which should be amenable to atmospheric characterisation. The authors estimate that on the order of ten super-Earth planets could be found around bright M3 stars in the optimistic habitable zone, adding to JWST’s sample of temperate worlds.

Figure 3: Simulation of small planets TESS may find (orange) as a function of the star’s distance to us. Kepler planet candidates are in blue and planets detected by other telescopes in black. The size of the circle is proportional to transit depth. [Barclay et al. 2018]

Conclusion

Today’s paper will be useful for planning follow up strategies and for identifying potential numbers of planets found in TESS 2-minute cadence data and full-frame images. This is the first paper to predict planets based on the stars most likely to be observed from the Candidate Target List rather than simulated star populations. Where the authors also considered the hotter, fainter, giant or more crowded stars that TESS would observe (excluded from the Candidate Target List), planet number estimates increased to 16,000. However the higher labour intensity to follow up and much higher rates of false positives mean few of these are likely to be confirmed.

Today’s paper shows that TESS should greatly add to the numbers of known small nearby planets which we should be able to investigate further. Now we just need to find them!

About the author, Emma Foxell:

I am a PhD student at the University of Warwick. My project involves searching for transiting exoplanets around bright stars using telescopes on the ground. Outside of astronomy, I enjoy rock climbing and hiking.

1 … 30 31 32 33 34 … 41