Latest / SLEEPY HISTORY / Meet the Smallest Planet Ever in the Universe_(M4A_128K)
Transcript
- 0:01Speaker 1: There is a planet orbiting a star 209 light years
- 0:04Speaker 1: away that is smaller than Mercury. It is barely larger
- 0:07Speaker 1: than our moon. Its entire world could fit inside the
- 0:11Speaker 1: width of Australia. When this planet passes in front of
- 0:14Speaker 1: its star, it blocks less than 12 parts per million
- 0:17Speaker 1: of the starlight. That is a dimming so faint it
- 0:20Speaker 1: would be invisible to every telescope humanity has ever built
- 0:23Speaker 1: except one. Astronomers needed nearly a thousand days of unbroken observation.
- 0:28Speaker 1: and a technique that listens to the vibrations of a
- 0:30Speaker 1: distant star just to prove this world exists. It is
- 0:34Speaker 1: called Kepler-37b. It is the smallest confirmed planet ever found
- 0:39Speaker 1: around a sun-like star, and it should not be possible
- 0:42Speaker 1: to detect. If you want to understand how astronomers found
- 0:45Speaker 1: a world that should be invisible, and why its existence
- 0:48Speaker 1: rewrites what a planet can be, hit that like button
- 0:52Speaker 1: and subscribe. Now, get yourself comfortable, let's begin. Every human
- 1:03Speaker 1: being carries an unconscious picture of what a planet is.
- 1:06Speaker 1: You formed yours before you could spell the word. It
- 1:09Speaker 1: came from posters on classroom walls, from photographs beamed home
- 1:13Speaker 1: by spacecraft, from the heavy certainty of standing on solid
- 1:17Speaker 1: ground and knowing that the rock beneath your feet stretches
- 1:20Speaker 1: thousands of kilometres in every direction. A planet is a
- 1:23Speaker 1: big thing, a world. A place with a horizon and
- 1:27Speaker 1: a sky, a surface you could theoretically stand on, a
- 1:30Speaker 1: gravitational presence that commands respect. Earth, Mars, Jupiter, Saturn, even Mercury,
- 1:37Speaker 1: the smallest planet in our solar system, fits comfortably inside
- 1:41Speaker 1: this mental image. Mercury is a battered, sun-scorched world, roughly 4,880
- 1:49Speaker 1: kilometres across. It has a global magnetic field. It has
- 1:53Speaker 1: towering cliffs where the entire crust buckled as the planet's
- 1:57Speaker 1: iron core cooled and shrank. It has a history written
- 2:00Speaker 1: in craters, in lava plains, in geology that speaks to
- 2:03Speaker 1: billions of years of planetary evolution. NASA's MESSENGER spacecraft, which
- 2:08Speaker 1: orbited Mercury from 2011 to 2015, revealed a planet far
- 2:14Speaker 1: more complex than the battered cratering record visible from Earth.
- 2:17Speaker 1: MESSENGER mapped the entire surface at high resolution and discovered
- 2:21Speaker 1: that smooth volcanic plains cover roughly 27% of the total
- 2:24Speaker 1: surface area. Vast regions where ancient lava flows buried older
- 2:30Speaker 1: craters under. Kilometers of solidified magma. The spacecraft detected unexpectedly
- 2:36Speaker 1: high concentrations of volatile elements including sulfur and potassium in
- 2:40Speaker 1: the surface rocks. Challenging earlier models that predicted Mercury's proximity
- 2:44Speaker 1: to the sun would have stripped those elements away during formation.
- 2:48Speaker 1: MESSENGER also revealed hollows, strange, shallow, irregularly shaped depressions found
- 2:53Speaker 1: nowhere else in the solar system. apparently formed by the
- 2:57Speaker 1: sublimation of volatile materials from just beneath the surface. These
- 3:01Speaker 1: features are geologically young, possibly still forming today. They suggest
- 3:06Speaker 1: a planet that remains active in ways no one predicted.
- 3:09Speaker 1: For a world this small and this old, Mercury's interior
- 3:12Speaker 1: tells an equally compelling story. Its iron core occupies roughly
- 3:16Speaker 1: 85% of the planet's radius, an enormous fraction compared to Earth,
- 3:21Speaker 1: where the core occupies only about 55%. This outsized core
- 3:25Speaker 1: gives Mercury the highest uncompressed density of any planet in
- 3:29Speaker 1: the solar system, meaning that even if you removed the
- 3:32Speaker 1: effect of gravitational compression, Mercury would still be denser than Earth.
- 3:36Speaker 1: The mechanisms that produce this iron-rich composition remain debated. One
- 3:41Speaker 1: hypothesis proposes that a giant impact early in Mercury's history
- 3:45Speaker 1: stripped away much of its silicate mantle, leaving behind a
- 3:49Speaker 1: world that is mostly core. Another suggests that intense solar
- 3:53Speaker 1: radiation during the planet's formation vaporized lighter materials from the
- 3:57Speaker 1: protoplanetary disk before they could be incorporated into the growing planet.
- 4:02Speaker 1: Either way, Mercury is not a simple ball of rock.
- 4:05Speaker 1: It is a geologically differentiated, chemically distinct, structurally complex world
- 4:11Speaker 1: with a formation story that remains actively investigated. This is
- 4:15Speaker 1: the object that defines the lower boundary of what a
- 4:18Speaker 1: planet looks like in our solar system. A world with
- 4:20Speaker 1: volcanic plains, a global magnetic field generated by a liquid
- 4:25Speaker 1: outer core, bizarre surface features found nowhere else, and an
- 4:29Speaker 1: interior structure that defies easy explanation. Mercury earned the title
- 4:34Speaker 1: of planet, not merely through its size but through its complexity,
- 4:38Speaker 1: through the evidence that it formed, evolved, and continues to
- 4:42Speaker 1: change in ways that reward scientific investigation across centuries. Mercury
- 4:46Speaker 1: is small, but it is undeniably a world. Now take
- 4:50Speaker 1: that mental image and compress it. Shrink Mercury by roughly 20%.
- 4:54Speaker 1: Reduce it until it is barely larger than the Moon,
- 4:59Speaker 1: hanging in your night sky. What you are holding in
- 5:02Speaker 1: your mind is no longer Mercury. It is an object
- 5:05Speaker 1: called Kepler-37b, and astronomers call it a planet. Kepler-37b orbits
- 5:11Speaker 1: a star 209 light-years from Earth in the constellation Lyra.
- 5:19Speaker 1: That star, designated Kepler 37, is a G-type main sequence
- 5:24Speaker 1: star similar to our own Sun but slightly smaller and cooler,
- 5:27Speaker 1: with a radius about 77% of the Sun's. The star
- 5:31Speaker 1: is unremarkable. It sits in no special region of the galaxy.
- 5:36Speaker 1: It does not pulse dramatically or flare violently. It is,
- 5:39Speaker 1: by every measure, an ordinary star. And orbiting that ordinary star,
- 5:44Speaker 1: closer than any planet in our solar system orbits our Sun,
- 5:48Speaker 1: is a world so small that it has no right
- 5:50Speaker 1: to be called a planet by any intuition you have
- 5:52Speaker 1: ever developed. The numbers are stark. Kepler-37b has a diameter
- 5:57Speaker 1: of approximately 3,900 kilometers. For context, our moon measures roughly 3,474
- 6:02Speaker 1: kilometers across. Mercury spans approximately 4,880 kilometers. This means Kepler-37b
- 6:05Speaker 1: sits almost exactly between the moon and Mercury in size.
- 6:16Speaker 1: closer in scale to the moon than to any object
- 6:18Speaker 1: we traditionally call a planet. It is roughly 14% larger
- 6:22Speaker 1: than the moon. It is roughly 20% smaller than Mercury.
- 6:27Speaker 1: It occupies a size regime that, in our solar system,
- 6:31Speaker 1: belongs to moons, not planets. Think about what this means.
- 6:36Speaker 1: You have looked at the moon your entire life. You
- 6:38Speaker 1: have watched it rise over rooftops, tracked its phases through
- 6:42Speaker 1: the months, maybe even seen it through a telescope where
- 6:44Speaker 1: its craters become sharp and vivid. That familiar silver disk
- 6:48Speaker 1: in your sky is almost the same size as a
- 6:51Speaker 1: confirmed planet orbiting another star. Not a candidate, not a
- 6:55Speaker 1: theoretical possibility. A confirmed, validated planet detected through one of
- 7:01Speaker 1: the most precise measurements in the history of astronomy. Kepler-37b
- 7:06Speaker 1: completes one orbit around its star every 13.4 days. Mercury,
- 7:11Speaker 1: the closest planet to our Sun, takes 88 days to
- 7:15Speaker 1: complete a single orbit. This means Kepler-37b is far closer
- 7:19Speaker 1: to its star than Mercury is to ours roughly one
- 7:22Speaker 1: quarter the distance. At that proximity, the planet receives a
- 7:26Speaker 1: punishing amount of radiation. Its equilibrium surface temperature is estimated
- 7:30Speaker 1: at approximately 718 Kelvin, which translates to roughly 445 degrees
- 7:37Speaker 1: Celsius or 833 degrees Fahrenheit. That is hot enough to
- 7:42Speaker 1: melt zinc, hot enough to melt lead. hot enough to
- 7:45Speaker 1: make the surface glow faintly in the infrared if you
- 7:48Speaker 1: could somehow get close enough to see it. There is
- 7:51Speaker 1: no atmosphere. There cannot be one. A body this small,
- 7:55Speaker 1: this hot, this close to its star has no mechanism
- 7:58Speaker 1: for holding onto any envelope of gas. Whatever volatiles may
- 8:01Speaker 1: have once existed on this world were stripped away long
- 8:04Speaker 1: ago by stellar radiation and solar wind. There is no water.
- 8:08Speaker 1: There are no clouds. There is no weather. What remains
- 8:12Speaker 1: is an airless, scorched, rocky ball, barely larger than a moon,
- 8:17Speaker 1: circling a distant sun in permanent silence. And yet, we
- 8:21Speaker 1: call it a planet. Not because it looks like one
- 8:24Speaker 1: or feels like one, but because it orbits a star
- 8:28Speaker 1: and it formed the way planets form. It is a
- 8:31Speaker 1: product of the same physical processes that built Mercury and
- 8:34Speaker 1: Earth and Jupiter. Gravity pulled dust and gas together in
- 8:38Speaker 1: a protoplanetary disk. Collisions accumulated material into a larger body.
- 8:43Speaker 1: That body settled into an orbit and has remained there
- 8:46Speaker 1: for billions of years. By the mechanical definition of what
- 8:49Speaker 1: a planet is an object that orbits a star, formed
- 8:52Speaker 1: through accretion, and is not massive enough to ignite nuclear fusion.
- 8:56Speaker 1: Kepler-37b qualifies. But by every human instinct about what a
- 9:01Speaker 1: planet should be, it feels wrong. It feels too small.
- 9:05Speaker 1: It feels like something that slipped through a crack in
- 9:08Speaker 1: our classification system, a rounding error in the cosmic inventory.
- 9:12Speaker 1: This discomfort is the point. Kepler-37b was not discovered because
- 9:18Speaker 1: astronomers expected to find a planet this small. It was
- 9:21Speaker 1: discovered because the Kepler Space Telescope was watching Kepler-37's light,
- 9:26Speaker 1: with enough precision and patience to notice an almost impossibly
- 9:30Speaker 1: faint dimming every 13 days. That dimming, the signal of
- 9:34Speaker 1: the planet's transit across the face of its star, is
- 9:37Speaker 1: so shallow that it represents a change in brightness of
- 9:40Speaker 1: approximately 12 parts per million. 12 parts per million. That
- 9:44Speaker 1: is a decrease of 0.0012%. If you were watching a
- 9:49Speaker 1: light bulb and its output dropped by 12 parts per million,
- 9:53Speaker 1: you would not notice. No human eye could detect it.
- 9:57Speaker 1: No ground-based telescope could reliably detect it. Only a space-based
- 10:02Speaker 1: observatory staring without blinking at the same star for years,
- 10:06Speaker 1: accumulating data from transit after transit, folding dozens of events
- 10:11Speaker 1: together to extract a coherent signal from the noise, could
- 10:14Speaker 1: find something this faint. We will explore exactly how this
- 10:18Speaker 1: detection worked shortly. But first, sit with the implication. The
- 10:22Speaker 1: smallest confirmed planet ever found around a sun-like star is
- 10:26Speaker 1: so small that its discovery required pushing photometric measurement to
- 10:30Speaker 1: the boundary of what is physically possible. It is not
- 10:33Speaker 1: a world that announced itself. It is a world that
- 10:36Speaker 1: had to be painstakingly extracted from data, teased out of noise,
- 10:41Speaker 1: statistically validated against every possible false alarm. And the fact
- 10:45Speaker 1: that it exists raises a question, that is far more
- 10:49Speaker 1: unsettling than the planet itself. If a world this small
- 10:52Speaker 1: can form and survive around a sun-like star, how many
- 10:56Speaker 1: similar worlds exist around the billions of other stars we
- 10:58Speaker 1: have never watched this carefully? But before that question can
- 11:02Speaker 1: land properly, you need to understand the neighborhood Kepler-37b lives in.
- 11:07Speaker 1: It is not alone. The Kepler-37 system contains at least
- 11:11Speaker 1: three known planets. Kepler-37b, the innermost and smallest, orbits every 13.4 days. Kepler-37c,
- 11:21Speaker 1: the middle planet, is slightly smaller than Venus, with a
- 11:25Speaker 1: radius approximately 75% of Earth's, and orbits every 21.3 days. Kepler-37d,
- 11:32Speaker 1: the outermost known planet, is roughly twice the size of
- 11:35Speaker 1: Earth and orbits every 39.8 days. All three planets orbit
- 11:39Speaker 1: their star closer than Mercury orbits our Sun. Number Rhine,
- 11:44Speaker 1: the entire system is compact, tight, a miniature arrangement of
- 11:48Speaker 1: worlds huddled close to their parent star. The orbital periods
- 11:52Speaker 1: of these three planets fall near a 5 to 8
- 11:54Speaker 1: to 15 ratio, hinting at a possible mean motion resonance chain,
- 11:59Speaker 1: a gravitational arrangement where the planets' orbital periods relate to
- 12:03Speaker 1: each other through simple integer ratios. This kind of configuration
- 12:07Speaker 1: is not random. It suggests a formation history, where the
- 12:10Speaker 1: planets migrated inward through the protoplanetary disk, and locked into
- 12:15Speaker 1: a gravitational relationship that has persisted for billions of years.
- 12:19Speaker 1: The architecture of the system tells a story of its own,
- 12:22Speaker 1: but within that story, the innermost planet stands out as
- 12:25Speaker 1: an anomaly of scale. Kepler-37c and d are small by
- 12:30Speaker 1: exoplanet standards, but they are not shocking. A planet three-quarters
- 12:34Speaker 1: the size of Earth is unusual, but not paradigm-breaking. A
- 12:38Speaker 1: planet twice Earth's size is well within the range of
- 12:41Speaker 1: thousands of known super-Earths. These are worlds that fit comfortably
- 12:45Speaker 1: within the mental framework astronomers have built over the past
- 12:48Speaker 1: two decades of exoplanet discovery. Kepler-37b does not fit. It
- 12:54Speaker 1: is an outlier at the extreme small end of the
- 12:56Speaker 1: confirmed planetary radius distribution. As of the most recent data
- 13:00Speaker 1: in the NASA Exoplanet Archive, no other confirmed exoplanet orbiting
- 13:05Speaker 1: a main sequence star has a measured radius smaller than
- 13:09Speaker 1: Kepler-37b's 0.31. Earth radii. The next smallest confirmed transiting planet
- 13:16Speaker 1: is Kepler-444b at approximately 0.4 Earth radii, and Kepler-102b are
- 13:23Speaker 1: meaningfully larger. Kepler-37b stands alone at the bottom of the
- 13:27Speaker 1: size distribution, a data point that stretches the boundary of
- 13:31Speaker 1: what we thought was detectable. Here is where the story
- 13:34Speaker 1: takes a turn that most people never consider. Kepler-37b is
- 13:39Speaker 1: the smallest confirmed exoplanet by radius. That is a specific
- 13:43Speaker 1: claim about a specific measurement. It does not mean it
- 13:47Speaker 1: is the lightest planet, the least massive planet, or the
- 13:50Speaker 1: smallest planet in every possible category. It means that when
- 13:54Speaker 1: astronomers measured the fraction of starlight this object blocks during transit,
- 13:59Speaker 1: the resulting radius is the smallest they have ever confirmed
- 14:02Speaker 1: around a star like our Sun. But its mass is unknown.
- 14:05Speaker 1: The mass of a transiting planet is typically determined through
- 14:09Speaker 1: a second independent measurement called radial velocity. As a planet orbits,
- 14:13Speaker 1: its gravity tugs on its star, causing the star to
- 14:17Speaker 1: wobble very slightly back and forth. This wobble shifts the
- 14:20Speaker 1: wavelengths of the star's light a tiny Doppler effect that
- 14:23Speaker 1: can be measured with extremely sensitive spectrographs. For Jupiter-sized planets,
- 14:28Speaker 1: the stellar wobble can reach tens or hundreds of meters
- 14:31Speaker 1: per second. For Earth-mass planets, it drops to roughly 10
- 14:35Speaker 1: centimeters per second. For a planet the size of Kepler-37b,
- 14:40Speaker 1: assuming a rocky composition, the expected radial velocity signal is
- 14:44Speaker 1: less than 1 cm per second. That is an order
- 14:47Speaker 1: of magnitude below the sensitivity of the best spectrographs ever built.
- 14:52Speaker 1: No instrument on Earth or in space can currently measure
- 14:55Speaker 1: how much Kepler-37b weighs. This means we know the planet's size,
- 15:00Speaker 1: but not its substance. We know how large its shadow is,
- 15:03Speaker 1: but not how much matter it contains. And it is
- 15:06Speaker 1: not only radial velocity that fails at this scale. Every
- 15:09Speaker 1: alternative detection method encounters the same wall. Astrometry, the technique
- 15:14Speaker 1: of measuring the physical displacement of a star on the
- 15:17Speaker 1: sky caused by a planet's gravitational pull, is sensitive to
- 15:22Speaker 1: planetary mass and orbital distance. But the astrometric signal from
- 15:25Speaker 1: Kepler-37b would be measured in fractions of a micro arc second,
- 15:31Speaker 1: thousands of times below the precision achievable by even the
- 15:34Speaker 1: Gaia spacecraft, the most accurate astrometric instrument ever built. Direct imaging,
- 15:40Speaker 1: photographing the planet as a point of light separate from
- 15:42Speaker 1: its star, requires angular separations and contrast ratios that are
- 15:47Speaker 1: physically impossible for a world this small, this faint. And
- 15:50Speaker 1: this close to a star 209 light years away. The
- 15:54Speaker 1: planet's reflected light would be roughly a billion times fainter
- 15:57Speaker 1: than the star, and the angular separation between them would
- 16:00Speaker 1: be barely more than a thousandth of an arc second.
- 16:03Speaker 1: No coronagraph or starshade currently designed could isolate that signal.
- 16:08Speaker 1: Transit timing variations, the technique of inferring the mass of
- 16:12Speaker 1: one planet from the gravitational perturbations it imposes on the
- 16:15Speaker 1: transit times of another planet in the same system, offer
- 16:19Speaker 1: a theoretical path to mass. Measurement, but the expected timing
- 16:23Speaker 1: perturbations from a body as light as Kepler-37b are fractions
- 16:27Speaker 1: of a second, buried in the noise of the transit
- 16:30Speaker 1: timing measurements for the other planets in the system. Every
- 16:33Speaker 1: path to measuring this planet's mass is blocked by the
- 16:36Speaker 1: same fundamental problem, The object is too small to exert
- 16:40Speaker 1: a detectable gravitational influence on anything we can observe. Researchers
- 16:44Speaker 1: have placed an upper limit on its mass roughly 0.79
- 16:48Speaker 1: Earth masses, but this is a boundary, not a measurement.
- 16:51Speaker 1: The actual mass could be anything below that ceiling. If
- 16:54Speaker 1: the planet is composed primarily of iron, like Mercury's core,
- 16:59Speaker 1: it would be denser and more massive for its size.
- 17:02Speaker 1: If it is composed of lighter silicate rock, it would
- 17:04Speaker 1: be less massive. We cannot tell the difference with current technology.
- 17:08Speaker 1: The planet's internal composition, its density, its structure, all of
- 17:13Speaker 1: these remain unknown. We are left with a silhouette, a shadow,
- 17:17Speaker 1: a shape without a weight. This uncertainty matters because mass
- 17:21Speaker 1: is one of the primary ways astronomers determine whether an
- 17:24Speaker 1: object is a planet, a dwarf planet, or something else entirely.
- 17:28Speaker 1: In our solar system, The distinction between a planet and
- 17:31Speaker 1: a dwarf planet rests partly on whether the object has
- 17:34Speaker 1: cleared its orbital neighborhood of other debris, a criterion that
- 17:38Speaker 1: depends heavily on mass. An object with enough mass exerts
- 17:41Speaker 1: sufficient gravitational influence to sweep its orbital zone clean over
- 17:45Speaker 1: billions of years. An object without enough mass cannot, and
- 17:49Speaker 1: it shares its orbit with a swarm of similar-sized bodies.
- 17:52Speaker 1: This is what separates Jupiter from Ceres, Earth from Pluto.
- 17:56Speaker 1: But this classification framework was designed for our solar system,
- 18:00Speaker 1: It does not formally apply to exoplanets. The International Astronomical
- 18:04Speaker 1: Union's 2006 resolution defining a planet specifies that the object
- 18:09Speaker 1: must orbit the Sun, not a star, not any star,
- 18:14Speaker 1: specifically the Sun. The definition is explicitly limited to our
- 18:18Speaker 1: solar system. For exoplanets, a separate working definition exists. Amended
- 18:23Speaker 1: in 2018 by the International Astronomical Union's Commission on Exoplanets
- 18:27Speaker 1: and the Solar System, That definition sets an upper mass
- 18:30Speaker 1: boundary below the deuterium burning limit of roughly 13 Jupiter
- 18:34Speaker 1: masses and requires a specific mass ratio with the central object.
- 18:39Speaker 1: But it sets no lower boundary on size or mass.
- 18:42Speaker 1: There is no minimum radius below which an exoplanet stops
- 18:45Speaker 1: being a planet. There is no minimum mass. As long
- 18:48Speaker 1: as it orbits a star or brown dwarf and is
- 18:50Speaker 1: not massive enough to burn deuterium, it is, technically, a planet.
- 18:55Speaker 1: This is why Kepler-37b can be called a planet despite
- 18:59Speaker 1: being barely larger than the Moon. The classification does not
- 19:03Speaker 1: depend on size. It depends on what the object orbits,
- 19:06Speaker 1: and what it is not massive enough to do. By
- 19:08Speaker 1: this standard, a rocky body the size of a large
- 19:11Speaker 1: asteroid could qualify as a planet if it orbited a star.
- 19:15Speaker 1: The word planet does not carry the intuitive weight we
- 19:18Speaker 1: assign it. At the smallest scales, it becomes a label
- 19:21Speaker 1: rather than a description, and Kepler-37b sits precisely at the
- 19:26Speaker 1: point where the label starts to feel inadequate. Consider the
- 19:30Speaker 1: comparison that makes this discomfort concrete. Ganymede, the largest moon
- 19:34Speaker 1: in our solar system, has a diameter of approximately 5,270 kilometres. Titan,
- 19:38Speaker 1: Saturn's largest moon, spans roughly 5,150 kilometres. Both of these
- 19:41Speaker 1: moons are substantially larger than Kepler-37b. Both are larger than Mercury,
- 19:53Speaker 1: and yet neither Ganymede nor Titan is a planet. They
- 19:57Speaker 1: orbit planets, not stars. Their orbital context determines their classification,
- 20:03Speaker 1: not their size. Now reverse the comparison. Pluto, The object
- 20:08Speaker 1: famously reclassified from planet to dwarf planet in 2006 has
- 20:12Speaker 1: a diameter of approximately 2,380 kilometers. Kepler-37b, at roughly 3,900 kilometers,
- 20:22Speaker 1: is significantly larger than Pluto. It is bigger than the
- 20:25Speaker 1: object whose demotion caused a public outcry about what counts
- 20:28Speaker 1: as a planet. And yet Pluto was reclassified partly because
- 20:32Speaker 1: it had not cleared its orbital neighborhood, a criterion that
- 20:35Speaker 1: does not even apply to exoplanets. under the current working definition.
- 20:39Speaker 1: The classification boundaries do not line up. The rules that
- 20:42Speaker 1: govern what we call a planet in our solar system
- 20:45Speaker 1: are different from the rules that govern what we call
- 20:47Speaker 1: a planet around another star. Kepler-37b exists in this gap,
- 20:53Speaker 1: a world that is technically a planet by exoplanet standards,
- 20:57Speaker 1: but would face serious definitional scrutiny if it orbited our sun.
- 21:01Speaker 1: None of this diminishes the reality of what Kepler-37b is,
- 21:05Speaker 1: It is a real object. It transits its star with
- 21:09Speaker 1: measurable regularity. Its radius has been determined with remarkable precision, 0.3098
- 21:14Speaker 1: Earth radii, with uncertainties of less than 3%. It is
- 21:21Speaker 1: not a candidate. It is not a statistical artifact. It
- 21:24Speaker 1: is a confirmed, validated detection published in the journal Nature
- 21:30Speaker 1: in 2013 by a team led by Thomas Barclay at
- 21:34Speaker 1: NASA's Ames Research Center. The discovery paper describes the detection
- 21:38Speaker 1: as requiring an exceptionally bright and quiet host star, a
- 21:42Speaker 1: precise astro-seismic characterisation of the star's radius, and nearly a
- 21:48Speaker 1: thousand days of continuous photometric monitoring. Every step of the
- 21:52Speaker 1: validation process was designed to ensure that this signal was
- 21:55Speaker 1: not a false positive, not a background eclipsing binary, not
- 22:00Speaker 1: an instrumental artefact, not a statistical fluctuation misidentified as a transit.
- 22:05Speaker 1: The planet is real. The question is what it tells
- 22:09Speaker 1: us about the universe. And what it tells us is
- 22:12Speaker 1: deeply uncomfortable. If planet formation can produce a body this
- 22:15Speaker 1: small around a sun-like star and park it in a
- 22:18Speaker 1: stable orbit for billions of years, then the lower boundary
- 22:22Speaker 1: of what counts as a planet is not set by physics.
- 22:25Speaker 1: It is set by detection. The universe does not stop
- 22:28Speaker 1: making rocky bodies at some convenient minimum size. It produces
- 22:32Speaker 1: objects across a continuous spectrum, from dust grains to super-Earths,
- 22:37Speaker 1: and the word planet is a human boundary drawn across
- 22:40Speaker 1: that spectrum, at a location that reflects our instruments more
- 22:44Speaker 1: than it reflects nature. Kepler-37b was found only because the
- 22:49Speaker 1: circumstances were extraordinarily favourable. The star was bright. The star
- 22:54Speaker 1: was quiet. The planet's orbit was short enough to produce
- 22:58Speaker 1: over 60 transits during Kepler's observation window. The star's size
- 23:02Speaker 1: was measured with rare precision through astroseismology. Remove any one
- 23:07Speaker 1: of these conditions and the planet disappears into the noise, undetectable, uncounted, unknown.
- 23:13Speaker 1: How many stars host similar worlds where the conditions are
- 23:15Speaker 1: not this favourable? How many moon-sized planets orbit in the
- 23:19Speaker 1: darkness of the galaxy, producing transit signals too faint for
- 23:23Speaker 1: any telescope we have ever pointed at the sky? The answer,
- 23:26Speaker 1: according to statistical analyses of Kepler's data, is almost certainly enormous.
- 23:31Speaker 1: small rocky planets appear to be the most common type
- 23:34Speaker 1: of planet in the galaxy. The smaller the radius bin,
- 23:37Speaker 1: the more planets it contains up to the point where
- 23:40Speaker 1: our instruments lose sensitivity and the counts artificially drop. The
- 23:43Speaker 1: true distribution of planetary sizes almost certainly continues downward below
- 23:48Speaker 1: our detection limits, meaning the galaxy is filled with tiny
- 23:52Speaker 1: rocky worlds we have never seen and cannot yet count.
- 23:55Speaker 1: Kepler-37b is the tip of something vast, It is the
- 24:00Speaker 1: one example that happened to be detectable, the single data
- 24:03Speaker 1: point that slipped above the noise floor and into our catalogue.
- 24:07Speaker 1: And the reason it matters is not that one tiny
- 24:09Speaker 1: planet is strange on its own. It matters because it
- 24:12Speaker 1: reveals a boundary the edge of what we can measure,
- 24:15Speaker 1: the place where the word planet starts to buckle under
- 24:18Speaker 1: its own assumptions, and the beginning of a planetary population
- 24:21Speaker 1: that extends far below what our eyes and instruments can reach.
- 24:25Speaker 1: But understanding what Kepler-37b truly represents requires understanding how it
- 24:30Speaker 1: was found, because the story of this planet is not
- 24:33Speaker 1: really about the planet at all. It is about a
- 24:36Speaker 1: measurement so precise, so delicate, and so close to the
- 24:40Speaker 1: limits of physical possibility that the detection itself is as
- 24:43Speaker 1: remarkable as the world it revealed. To understand how astronomers
- 24:47Speaker 1: found Kepler-37b, you first need to understand what a transit
- 24:51Speaker 1: actually is, not in the abstract, not as a diagram
- 24:56Speaker 1: in a textbook. You need to feel the geometry of it,
- 24:59Speaker 1: because the geometry is everything. A planet orbits a star.
- 25:04Speaker 1: From most vantage points in the galaxy, you would never
- 25:06Speaker 1: know the planet was there. It would be lost in
- 25:09Speaker 1: the glare of the star, invisible, undetectable by any change
- 25:13Speaker 1: in the star's brightness. But if your line of sight
- 25:16Speaker 1: happens to align almost exactly with the planet's orbital plane,
- 25:19Speaker 1: if you are watching from just the right angle, then
- 25:22Speaker 1: once per orbit, the planet will cross directly between you
- 25:26Speaker 1: and the star. It will pass across the face of
- 25:28Speaker 1: the stellar disk the way a marble would pass across
- 25:31Speaker 1: the face of a dinner plate if you were watching
- 25:34Speaker 1: from across a football field. During that crossing, the planet
- 25:37Speaker 1: blocks a tiny fraction of the starlight. The star dims,
- 25:41Speaker 1: not by much, not by anything you could see with
- 25:44Speaker 1: your eyes, but by an amount that a sufficiently sensitive
- 25:47Speaker 1: instrument can measure. This is transit photometry. It is the
- 25:51Speaker 1: technique that has discovered the majority of all known exoplanets.
- 25:56Speaker 1: Its power is lies in its simplicity. The fraction of
- 25:59Speaker 1: starlight blocked during a transit depends on exactly one geometric quantity,
- 26:04Speaker 1: the ratio of the planet's cross-sectional area to the star's
- 26:08Speaker 1: cross-sectional area. Because both objects are roughly spherical, this reduces
- 26:12Speaker 1: to the square of the ratio of their radii. If
- 26:15Speaker 1: the planet has a radius one tenth the size of
- 26:18Speaker 1: the star, it will block one percent of the light.
- 26:21Speaker 1: If the planet has a radius one hundredth the size
- 26:23Speaker 1: of the star, It will block 1 100th of 1%.
- 26:27Speaker 1: The transit depth is the planet's silhouette measured in units
- 26:30Speaker 1: of starlight. For large planets, this signal is generous. A
- 26:35Speaker 1: Jupiter-sized planet crossing a sun-like star produces a transit depth
- 26:39Speaker 1: of roughly 1% 10,000 parts per million. Ground-based telescopes can
- 26:44Speaker 1: detect that. It is how the first transiting exoplanet, HD 209458b,
- 26:48Speaker 1: was confirmed in the year 2000. For Earth-sized planets, the
- 26:55Speaker 1: signal is far more subtle. An Earth-sized world crossing a
- 26:59Speaker 1: Sun-like star produces a transit depth of roughly 84 parts
- 27:03Speaker 1: per million, a dimming of 0.0084%. Detecting that requires space-based
- 27:10Speaker 1: telescopes free from the blurring effects of Earth's atmosphere. It
- 27:14Speaker 1: is extraordinarily difficult, but achievable. For Kepler-37b, the signal is
- 27:20Speaker 1: something else entirely. The planet's radius is approximately 0.31 Earth radii.
- 27:25Speaker 1: The host star's radius is approximately 0.77 solar radii. When
- 27:30Speaker 1: you compute the transit depth, the square of the planet
- 27:33Speaker 1: to star radius ratio, you get a value of roughly
- 27:36Speaker 1: 12 parts per million. 12, not 12,000, not 1,200, 12.
- 27:40Speaker 1: A dimming of 0.0012%. To grasp how faint this is,
- 27:41Speaker 1: consider a different kind of measurement. Imagine you are in
- 27:52Speaker 1: a quiet room, listening to a sustained musical note played
- 27:56Speaker 1: by an orchestra of 10,000 instruments. One of those instruments
- 28:00Speaker 1: changes its pitch by a fraction so small that the
- 28:03Speaker 1: overall sound shifts by 12 ten-millionths of its total frequency.
- 28:08Speaker 1: That is the scale of the signal. Kepler-37b produces. It
- 28:12Speaker 1: is not merely small. It is at the frontier of
- 28:15Speaker 1: what physical measurement can achieve with photons collected from a
- 28:19Speaker 1: star 209 light-years away. Detecting a signal this faint requires
- 28:23Speaker 1: three things. An instrument with extraordinary precision, a star that cooperates,
- 28:29Speaker 1: and time. Kepler. 37B had all three, and it still
- 28:33Speaker 1: nearly escaped detection. The instrument was the Kepler Space Telescope.
- 28:38Speaker 1: Launched by NASA in March 2009, Kepler was designed with
- 28:42Speaker 1: one overriding purpose. To stare at a single patch of
- 28:46Speaker 1: sky in the constellations Cygnus and Lyra, monitoring the brightness
- 28:50Speaker 1: of over 150,000 stars simultaneously, watching for the periodic dimming
- 28:55Speaker 1: caused by transiting planets. Kepler did not scan the sky.
- 28:59Speaker 1: It did not slew from target to target. It locked
- 29:02Speaker 1: onto one field and watched it without interruption for four years,
- 29:06Speaker 1: accumulating brightness measurements with a cadence of roughly one data
- 29:10Speaker 1: point every 30 minutes for most targets. The telescope's photometric
- 29:14Speaker 1: precision was remarkable. For bright, quiet stars, Kepler could measure
- 29:19Speaker 1: changes in brightness at the level of a few tens
- 29:22Speaker 1: of parts per million over a period of several hours.
- 29:25Speaker 1: Over longer baselines, averaging multiple transits together, the effective precision
- 29:30Speaker 1: improved further. The instrument achieved this by orbiting the Sun
- 29:34Speaker 1: rather than Earth, eliminating the day-night cycle and atmospheric interference
- 29:38Speaker 1: that plague ground-based observations. Its detector was a mosaic of
- 29:42Speaker 1: charge-coupled devices, covering a field of view spanning 115 square degrees,
- 29:48Speaker 1: an area of sky roughly the size of your hand,
- 29:52Speaker 1: held at arm's length. Within that field, every star bright
- 29:55Speaker 1: enough to register on the detector was monitored continuously. Kepler
- 30:00Speaker 1: did not choose its targets based on whether they were
- 30:02Speaker 1: likely to have planets. It watched everything and let the
- 30:05Speaker 1: data speak. But raw instrumental precision is not enough to
- 30:09Speaker 1: detect a 12 part per million transit. Stars are not
- 30:13Speaker 1: perfect light bulbs. They flicker. They pulsate. They have spots
- 30:18Speaker 1: that rotate in and out of view. They experience flares,
- 30:21Speaker 1: convective granulation, and long-term brightness variations driven by magnetic activity cycles.
- 30:27Speaker 1: For many stars in Kepler's field of view, this intrinsic
- 30:31Speaker 1: stellar variability produces noise at the level of hundreds or
- 30:34Speaker 1: thousands of parts per million, easily drowning a signal as
- 30:38Speaker 1: small as Kepler-37b's transit. To find a planet this small,
- 30:43Speaker 1: you need a star that is not only bright but
- 30:45Speaker 1: also remarkably stable, a star that sits still, photometrically speaking,
- 30:50Speaker 1: while you watch it for years. Kepler-37 is that star.
- 30:55Speaker 1: It is a G-type main sequence star, somewhat cooler and
- 30:58Speaker 1: smaller than our sun, with an effective temperature of approximately 5,300 Kelvin. Critically,
- 31:04Speaker 1: it is one of the brightest stars in Kepler's field
- 31:07Speaker 1: at visual magnitude 9.8, meaning the telescope collected a large
- 31:11Speaker 1: number of photons from it during every exposure. More photons
- 31:15Speaker 1: means lower shot noise, which means better precision on each
- 31:18Speaker 1: individual brightness measurement. But brightness alone is not the key advantage.
- 31:23Speaker 1: Kepler-37 is also an exceptionally quiet star. Its magnetic activity
- 31:28Speaker 1: level is low. Its surface does not produce large star
- 31:32Speaker 1: spots that rotate in and out of view and modulate
- 31:34Speaker 1: the brightness. Its rotation period is estimated at roughly 29 days,
- 31:39Speaker 1: and the amplitude of its rotational brightness variation is small
- 31:43Speaker 1: enough that it does not overwhelm the planetary transit signal.
- 31:47Speaker 1: In a field of over 150,000 stars, Kepler-37 was one
- 31:51Speaker 1: of the few that offered a noise floor low enough
- 31:54Speaker 1: for a 12 part per million signal to be detectable.
- 31:57Speaker 1: Even on this unusually cooperative star, a single transit of
- 32:00Speaker 1: Kepler-37b would be invisible. One 12 part per million dip,
- 32:05Speaker 1: lasting perhaps an hour, buried in the noise of stellar
- 32:08Speaker 1: variability and instrumental systematics, would be indistinguishable from random fluctuation.
- 32:14Speaker 1: The detection required something that Kepler was uniquely positioned to provide, time.
- 32:19Speaker 1: Because Kepler-37b orbits every 13.4 days, The planet transits its
- 32:25Speaker 1: star roughly 27 times per year. Over the course of
- 32:28Speaker 1: Kepler's primary mission, the telescope accumulated 978 days of nearly
- 32:34Speaker 1: continuous observations of this star. During that interval, Kepler-37b transited
- 32:40Speaker 1: its star more than 60 times. Each individual transit is noisy,
- 32:45Speaker 1: but when you know the orbital period, you can fold
- 32:48Speaker 1: all the transits on top of each other. You align
- 32:51Speaker 1: them in time so that the midpoint of every transit
- 32:54Speaker 1: event overlaps. Then you average. The random noise, the stellar flickering,
- 32:59Speaker 1: the instrumental drift, the photon statistics averages down. The transit signal,
- 33:04Speaker 1: which repeats at the same depth and duration every time,
- 33:08Speaker 1: does not average down. It reinforces. With each additional transit
- 33:12Speaker 1: folded into the stack, the signal grows stronger relative to
- 33:16Speaker 1: the noise. After 62 transits, the accumulated signal rises above
- 33:20Speaker 1: the noise threshold with sufficient statistical confidence to claim a detection.
- 33:25Speaker 1: This is the core technique. It is elegant in principle
- 33:28Speaker 1: and brutal in practice. The folding process requires knowing the
- 33:32Speaker 1: orbital period to extraordinary precision, a fraction of a second
- 33:35Speaker 1: over 13 days, so that the transits align correctly. Any
- 33:40Speaker 1: error in the period smears the signal across multiple phase
- 33:43Speaker 1: bins and destroys the detections. The process also requires removing
- 33:47Speaker 1: systematic trends from the data. Slow drifts in the detector response,
- 33:51Speaker 1: quarterly rotations of the spacecraft that shifted stars onto different pixels,
- 33:56Speaker 1: thermal variations that changed the instrument's sensitivity over time. Each
- 34:00Speaker 1: of these systematics can mimic or mask a transit signal,
- 34:04Speaker 1: if not carefully modeled and removed. One of the most
- 34:07Speaker 1: persistent challenges was the quarterly rotation of the spacecraft. Every
- 34:11Speaker 1: 90 days, Kepler rotated 90 degrees around its pointing axis
- 34:15Speaker 1: to keep its solar panels oriented toward the Sun. This
- 34:18Speaker 1: rotation shifted every star in the field of view onto
- 34:21Speaker 1: a different region of the detector, where the pixel response characteristics, sensitivity,
- 34:27Speaker 1: charge bleeding, interpixel capacitance differed from the previous quarter. Each
- 34:32Speaker 1: time the spacecraft rotated, the photometric baseline for every star
- 34:36Speaker 1: changed discontinuously. These jumps had to be identified and corrected
- 34:41Speaker 1: before any transit search could proceed. For a star like Kepler-37,
- 34:45Speaker 1: where the transit signal was only 12 parts per million,
- 34:49Speaker 1: even a small uncorrected offset at a quarterly boundary could
- 34:52Speaker 1: introduce or erase a signal at the level of the
- 34:55Speaker 1: planetary transit. Thermal variations added another layer of complexity. The
- 35:00Speaker 1: telescope's focus changed subtly with temperature fluctuations as the spacecraft's
- 35:04Speaker 1: orientation relative to the Sun shifted over its orbital period.
- 35:08Speaker 1: These focus changes altered the distribution of starlight across detector pixels,
- 35:13Speaker 1: modulating the measured brightness in ways that correlated with spacecraft
- 35:16Speaker 1: temperature rather than with any astrophysical signal. Correcting for these
- 35:20Speaker 1: focus-driven systematics required modeling the relationship between pixel-level flux distributions
- 35:26Speaker 1: and thermal telemetry data from the spacecraft, a process that
- 35:30Speaker 1: was computationally intensive and imperfect. Residual systematics at the level
- 35:35Speaker 1: of a few parts per million persisted even after correction.
- 35:39Speaker 1: And for the smallest planets, these residuals represented a significant
- 35:43Speaker 1: fraction of the transit depth itself. The data processing pipeline
- 35:46Speaker 1: for Kepler was one of the most sophisticated ever built
- 35:50Speaker 1: for an astronomical mission, involving multiple stages of detrending, outlier rejection,
- 35:56Speaker 1: and systematic correction before any planet search could begin. But
- 36:00Speaker 1: even after folding 62 transits and extracting a statistically significant signal,
- 36:05Speaker 1: the detection of Kepler 37b faced another challenge that no
- 36:10Speaker 1: amount of data processing could resolve on its own. How
- 36:13Speaker 1: do you know the signal is actually a planet? Transit
- 36:16Speaker 1: photometry is vulnerable to a specific class of false positives
- 36:19Speaker 1: called astrophysical blends. Imagine a faint background star, too dim
- 36:24Speaker 1: to resolve separately from Kepler-37, that happens to lie nearly
- 36:29Speaker 1: along the same line of sight. If that background star
- 36:32Speaker 1: is an eclipsing binary two stars orbiting each other, With
- 36:35Speaker 1: one periodically blocking the other, the resulting brightness variation could
- 36:39Speaker 1: mimic a planetary transit when diluted by the brighter foreground star.
- 36:43Speaker 1: The eclipse of the binary would be deep, perhaps 10
- 36:46Speaker 1: or 20%, but blended with the much brighter Kepler-37. The
- 36:51Speaker 1: apparent dimming would be reduced to a few parts per million.
- 36:54Speaker 1: It would look exactly like a tiny planet transiting a
- 36:57Speaker 1: bright star. Ruling out this scenario requires additional analysis. The
- 37:02Speaker 1: Kepler team used a technique called Statistical validation. Applying a
- 37:07Speaker 1: software tool known as Blender that systematically models all plausible
- 37:11Speaker 1: astrophysical scenarios, background eclipsing binaries, hierarchical triple star systems, grazing
- 37:18Speaker 1: stellar eclipses, and compares the probability of each scenario to
- 37:23Speaker 1: the probability that the signal is a genuine planet. For Kepler-37b...
- 37:28Speaker 1: the validation analysis concluded that the planetary interpretation was overwhelmingly
- 37:33Speaker 1: more probable than any false positive scenario. The blender analysis
- 37:37Speaker 1: was not a single test, but a systematic exploration of
- 37:41Speaker 1: every plausible astrophysical configuration that could mimic the observed signal.
- 37:45Speaker 1: The software modelled scenarios in which a faint background eclipsing
- 37:49Speaker 1: binary star system lay behind Kepler-37 at various distances, blending
- 37:54Speaker 1: its deep eclipses with the bright foreground star's light to
- 37:58Speaker 1: produce an apparent shallow transit. It modeled hierarchical triple star
- 38:03Speaker 1: configurations in which a physically bound companion to Kepler-37 was
- 38:07Speaker 1: itself being eclipsed by a low mass stellar or substellar object.
- 38:12Speaker 1: It modeled grazing eclipses where a stellar companion just barely
- 38:15Speaker 1: skimmed the edge of Kepler-37's disk, producing a shallow V-shaped
- 38:20Speaker 1: dimming that could resemble a planetary transit. For each scenario,
- 38:24Speaker 1: Blender generated synthetic light curves compared them to the observed
- 38:28Speaker 1: transit shape, and calculated the statistical likelihood that the scenario
- 38:32Speaker 1: could reproduce the data. One by one, the false positive
- 38:36Speaker 1: configurations were eliminated. Their predicted light curve shapes were too V-shaped,
- 38:42Speaker 1: or too deep in certain wavelength bands, or inconsistent with
- 38:45Speaker 1: the absence of secondary eclipses in the Kepler data. The
- 38:49Speaker 1: planetary model survived every test, but the team did not
- 38:52Speaker 1: stop at statistical modelling. They pointed NASA's Spitzer Space Telescope
- 38:56Speaker 1: at Kepler-37 to observe a transit of the innermost planet
- 39:00Speaker 1: at an infrared wavelength of 4.5 micrometers, a completely different
- 39:05Speaker 1: part of the electromagnetic spectrum from Kepler's optical bandpass. This
- 39:10Speaker 1: was a critical independent check. If the transit signal were
- 39:14Speaker 1: actually caused by a background eclipsing binary blended with Kepler-37,
- 39:17Speaker 1: the two stars would almost certainly have different surface temperatures,
- 39:22Speaker 1: and therefore different brightness ratios at optical and infrared wavelengths.
- 39:27Speaker 1: The blend fraction would change with wavelength, and the apparent
- 39:30Speaker 1: transit depth measured by Spitzer would differ from the depth
- 39:33Speaker 1: measured by Kepler. A genuine planet, by contrast, blocks starlight
- 39:38Speaker 1: by the same geometric fraction regardless of wavelength its silhouette
- 39:41Speaker 1: is colourless. The Spitzer observation detected the transit of Kepler-37b
- 39:46Speaker 1: at a depth consistent with the Kepler measurement. The signal
- 39:50Speaker 1: was the same in the infrared as it was in
- 39:51Speaker 1: visible light. This wavelength-independent depth eliminated the remaining blend scenarios
- 39:57Speaker 1: with a directness that statistical modelling alone could not achieve.
- 40:00Speaker 1: The planet was real. Two telescopes operating at different wavelengths,
- 40:05Speaker 1: separated by years of observation, agreed on the same answer.
- 40:09Speaker 1: The team also used Spitzer Space Telescope observations at infrared
- 40:12Speaker 1: wavelengths to confirm that the transit depth was consistent across
- 40:16Speaker 1: different colours of light. A background eclipsing binary would typically
- 40:20Speaker 1: produce a different depth at different wavelengths because the two
- 40:23Speaker 1: blended stars would have different colours. The fact that the
- 40:26Speaker 1: transit depth, measured by Spitzer at 4.5 micrometres, matched the
- 40:32Speaker 1: Kepler depth in visible light, provided strong independent evidence that
- 40:36Speaker 1: the signal originated from a single star with a genuine
- 40:39Speaker 1: transiting planet. The validation was rigorous, careful and conservative. It
- 40:45Speaker 1: had to be. The claim of discovering the smallest planet
- 40:49Speaker 1: ever found around a main sequence star invites intense scrutiny.
- 40:53Speaker 1: Every alternative explanation had to be tested and rejected before
- 40:57Speaker 1: the discovery could be announced. When the team published their
- 41:00Speaker 1: results in the journal Nature in February 2013, the paper
- 41:05Speaker 1: presented not just a detection but a comprehensive argument that
- 41:09Speaker 1: no plausible non-planetary explanation could account for the observed signal.
- 41:14Speaker 1: At the heart of this argument was one technique that
- 41:16Speaker 1: made the entire discovery possible, a technique most people have
- 41:20Speaker 1: never heard of, applied to a purpose it was not
- 41:23Speaker 1: originally designed for. That technique is astroseismology, and without it,
- 41:28Speaker 1: Kepler-37b could not exist in our catalogue. Stars vibrate, not visibly,
- 41:34Speaker 1: not dramatically, but continuously. Deep inside a star like Kepler-37,
- 41:40Speaker 1: convective motions in the outer layers generate pressure waves, sound
- 41:44Speaker 1: waves that propagate through the stellar interior. These pressure waves,
- 41:48Speaker 1: called p-modes, behave much like sound waves inside a musical instrument,
- 41:52Speaker 1: except the instrument is a sphere of plasma 700,000 kilometers across.
- 41:58Speaker 1: Different oscillation modes penetrate to different depths within the star.
- 42:02Speaker 1: Low order modes with longer wavelengths reach deep into the
- 42:05Speaker 1: stellar core, sampling the density and temperature conditions near the center.
- 42:09Speaker 1: Higher order modes with shorter wavelengths remain confined to the
- 42:13Speaker 1: outer layers, sampling the convective envelope in the region just
- 42:16Speaker 1: below the surface. Each mode vibrates at a characteristic frequency
- 42:20Speaker 1: determined by the speed of sound along its path through
- 42:23Speaker 1: the interior. Because the speed of sound depends on temperature, density,
- 42:28Speaker 1: and chemical composition, the pattern of oscillation frequencies encodes a
- 42:32Speaker 1: detailed portrait of the star's internal structure, like a fingerprint
- 42:36Speaker 1: made of sound. For sun-like stars, The dominant oscillation periods
- 42:41Speaker 1: fall in the range of roughly 3 to 15 minutes,
- 42:44Speaker 1: producing brightness variations of only a few parts per million.
- 42:47Speaker 1: Detecting these oscillations requires the same kind of sustained, high-precision
- 42:52Speaker 1: photometry that transit detection demands. Kepler was ideally suited for
- 42:56Speaker 1: this dual purpose. The same uninterrupted time series that revealed
- 43:00Speaker 1: planetary transits also captured the star's acoustic spectrum. astronomers extracted
- 43:06Speaker 1: the oscillation frequencies from the Fourier transform of the brightness data,
- 43:10Speaker 1: identifying individual modes separated by just a few microhertz frequency differences,
- 43:16Speaker 1: so small that resolving them required hundreds of days of
- 43:19Speaker 1: continuous observation. The critical quantity derived from this analysis is
- 43:23Speaker 1: the large frequency separation, the regular spacing between consecutive oscillation
- 43:28Speaker 1: modes of the same angular degree. This separation is directly
- 43:32Speaker 1: proportional to the square root of the star's mean density.
- 43:35Speaker 1: Measure the spacing, and you measure the density. Combine the
- 43:39Speaker 1: density with the effective temperature derived from spectroscopy, and you
- 43:43Speaker 1: determine the radius. For Kepler-37, this process yielded a stellar
- 43:48Speaker 1: radius of 0.77 solar radii, with an uncertainty of roughly 3%,
- 43:54Speaker 1: a precision that was, at the time, a record for
- 43:57Speaker 1: a star this small. At the time of the discovery,
- 44:00Speaker 1: Kepler-37 was the densest star in which solar-like oscillations had
- 44:05Speaker 1: been detected through astro-seismology. The measurement was at the edge
- 44:09Speaker 1: of what was achievable. The oscillation amplitudes were tiny, the
- 44:13Speaker 1: frequencies were high, and the signal required careful extraction from
- 44:17Speaker 1: a power spectrum dominated by granulation noise from the star's
- 44:22Speaker 1: convective surface. Without this astro-seismic anchor, The stellar radius would
- 44:27Speaker 1: have been known only to perhaps 10 or 15% from
- 44:30Speaker 1: spectroscopy alone, and the planetary radius would have been correspondingly uncertain.
- 44:35Speaker 1: Too uncertain to claim a record. These waves bounce off
- 44:38Speaker 1: the surface, reflect back through the core, and set up
- 44:42Speaker 1: standing oscillation patterns that cause the entire star to expand
- 44:46Speaker 1: and contract by tiny amounts. The star rings like a
- 44:49Speaker 1: bell struck by the turbulence of its own convection. The
- 44:52Speaker 1: resulting brightness variations are extraordinarily small. typically a few parts
- 44:57Speaker 1: per million, but they occur at specific frequencies determined by
- 45:01Speaker 1: the star's internal structure, density and size. By measuring these
- 45:06Speaker 1: oscillation frequencies, astronomers can determine the star's properties with a
- 45:10Speaker 1: precision that no other technique can match. The spacing between
- 45:13Speaker 1: oscillation modes reveals the star's mean density. The overall frequency
- 45:17Speaker 1: pattern reveals the star's radius and mass. For Kepler-37, astrozismic
- 45:23Speaker 1: analysis determined the stellar radius to within 3%, an uncertainty
- 45:27Speaker 1: of roughly 0.02 solar radii. This was the most precise
- 45:32Speaker 1: stellar radius measurement ever achieved for a star this small
- 45:35Speaker 1: at the time of the discovery. Why does the star's
- 45:38Speaker 1: radius matter so much? Because the transit depth gives you
- 45:42Speaker 1: only the ratio of the planet's radius to the star's radius.
- 45:46Speaker 1: If you know the transit depth is 12 parts per million,
- 45:49Speaker 1: you know that the planet's cross-sectional area is 12 millionths
- 45:52Speaker 1: of the star's cross-sectional area. But to convert that ratio
- 45:56Speaker 1: into an actual planetary radius in kilometres, you need to
- 45:59Speaker 1: know the star's radius in kilometres. A 10% error in
- 46:03Speaker 1: the stellar radius propagates directly into a 10% error in
- 46:07Speaker 1: the planetary radius. For a planet this small, a 10%
- 46:11Speaker 1: error could mean the difference between a world slightly larger
- 46:13Speaker 1: than the Moon and a world the size of Mars.
- 46:16Speaker 1: The interpretation changes completely. Astroseismology gave the Kepler team the
- 46:21Speaker 1: precision they needed. It anchored the entire measurement chain. Without it,
- 46:26Speaker 1: the transit depth would have been just a number, a
- 46:29Speaker 1: ratio without a scale. With it, the ratio became a size, 0.31
- 46:34Speaker 1: Earth radii, approximately 3,900 kilometers in diameter, a world between
- 46:40Speaker 1: the Moon and Mercury. The confidence in that size comes
- 46:44Speaker 1: directly from the confidence in the star. And the confidence
- 46:47Speaker 1: in the star comes from listening to its internal vibrations,
- 46:51Speaker 1: its stellar heartbeat, captured in the same photometric data stream
- 46:56Speaker 1: that revealed the planet's transit. This is one of the
- 46:59Speaker 1: most remarkable aspects of the discovery. The same instrument, the
- 47:03Speaker 1: same data, the same stream of photons collected by Kepler
- 47:07Speaker 1: over nearly a thousand days simultaneously revealed the planet's existence
- 47:11Speaker 1: through its transits and the star's size through its oscillations.
- 47:15Speaker 1: The planet and the star were characterised together from the
- 47:18Speaker 1: same observations, in a single self-consistent analysis. Nothing about the
- 47:24Speaker 1: measurement was borrowed from external assumptions or indirect proxies. The
- 47:28Speaker 1: star told astronomers its own size, and the planet told
- 47:32Speaker 1: astronomers its own shadow, and together those two pieces of
- 47:35Speaker 1: information yielded a planetary radius measured to better than 6% precision.
- 47:40Speaker 1: There is one final dimension to this detection that is
- 47:43Speaker 1: easy to overlook but impossible to ignore once you understand it.
- 47:46Speaker 1: Kepler-37b has no measured mass. The standard method for measuring
- 47:51Speaker 1: a planet's mass is radial velocity spectroscopy. As a planet orbits,
- 47:56Speaker 1: its gravitational pull causes the star to wobble toward and
- 48:00Speaker 1: away from the observer. This wobbled Doppler shifts the star's
- 48:03Speaker 1: spectral lines by a tiny amount a few meters per
- 48:06Speaker 1: second for Jupiter-mass planets, a few centimeters per second for
- 48:10Speaker 1: Earth-mass planets. By measuring these shifts over many orbits, astronomers
- 48:15Speaker 1: can determine the planet's mass. For Kepler-37b, the expected radial
- 48:20Speaker 1: velocity signal is less than one centimeter per second. One
- 48:24Speaker 1: centimeter per second. That is the speed at which a
- 48:27Speaker 1: snail moves across a leaf. It is roughly two orders
- 48:30Speaker 1: of magnitude below the sensitivity of even ESPRESSO, the most
- 48:34Speaker 1: precise optical spectrograph in operation, which achieves approximately 10 centimeters
- 48:39Speaker 1: per second precision on its best targets. On a star
- 48:43Speaker 1: as faint as Kepler-37, the practical radial velocity precision of
- 48:47Speaker 1: the best available spectrographs is closer to one and a
- 48:50Speaker 1: half to two meters per second, hundreds of times larger
- 48:53Speaker 1: than the signal this planet would produce. Even assuming a
- 48:56Speaker 1: composition of pure iron, the densest plausible material the planet's
- 49:01Speaker 1: mass would be only about three hundredths of an Earth mass,
- 49:04Speaker 1: producing a radial velocity signal far too small for any
- 49:08Speaker 1: existing instrument. The team placed an upper limit on the
- 49:11Speaker 1: mass at roughly 0.79 Earth masses, but this is a constraint,
- 49:16Speaker 1: not a detection. The actual mass could be a tenth
- 49:19Speaker 1: of that limit or a hundredth. There is no way
- 49:21Speaker 1: to know. This means Kepler-37b is a planet defined entirely
- 49:26Speaker 1: by its shadow. We know its shape. We know its size.
- 49:30Speaker 1: We know its orbit. We know its temperature. But we
- 49:33Speaker 1: do not know its weight. We do not know its density.
- 49:37Speaker 1: We cannot determine whether it is a ball of iron,
- 49:39Speaker 1: porous silicate rubble pile or something in between. Every statement
- 49:44Speaker 1: about its composition is an inference based on what is
- 49:46Speaker 1: physically plausible for a body of this size, not a measurement.
- 49:50Speaker 1: For most exoplanets, the combination of radius and mass provides
- 49:54Speaker 1: the density, and density reveals composition. A dense planet is rocky.
- 49:59Speaker 1: A fluffy planet has a thick atmosphere or volatile rich envelope.
- 50:04Speaker 1: This is how astronomers classify worlds into categories terrestrial, super-Earth, mini-Neptune,
- 50:10Speaker 1: gas giant. Without mass, Kepler-37b resists classification beyond probably rocky.
- 50:18Speaker 1: It is too small to retain a hydrogen envelope. It
- 50:21Speaker 1: is too hot for surface ices. Rocky is the only
- 50:24Speaker 1: plausible composition. But probably rocky is not the same as
- 50:28Speaker 1: measured to be rocky. The distinction matters because it illustrates
- 50:32Speaker 1: a deeper truth about the smallest exoplanets. They are the
- 50:35Speaker 1: hardest to characterize. The same properties that make them faint
- 50:39Speaker 1: in transit, their tiny radii, their minuscule gravitational influence make
- 50:44Speaker 1: them invisible to every complementary measurement technique. They exist at
- 50:48Speaker 1: the edge of what our instruments can confirm, known by
- 50:51Speaker 1: their shadows and nothing more. Kepler-37b was found by a
- 50:54Speaker 1: telescope that stared at one patch of sky for nearly
- 50:57Speaker 1: three years. Watching a star that happened to be bright
- 51:00Speaker 1: and quiet enough to permit a 12 part per million
- 51:03Speaker 1: signal to emerge from the noise after 62 transits were
- 51:06Speaker 1: folded together, anchored by an asterosismic measurement of the star's
- 51:11Speaker 1: vibrations that determined its radius to 3% precision, validated against
- 51:15Speaker 1: every plausible false positive scenario, using statistical modeling and infrared
- 51:21Speaker 1: confirmation from a second space telescope. Remove any link in
- 51:25Speaker 1: that chain and the discovery does not happen. The planet
- 51:28Speaker 1: remains hidden, uncounted, one more invisible world in a galaxy
- 51:32Speaker 1: that may contain billions like it. And this chain of
- 51:35Speaker 1: dependencies reveals something important. The discovery of Kepler-37b is not
- 51:42Speaker 1: just a story about a small planet. It is a
- 51:44Speaker 1: story about a measurement pushed to the absolute boundary of
- 51:47Speaker 1: what is possible. The planet was not found because it
- 51:50Speaker 1: was easy or obvious. It was found because every condition aligned.
- 51:54Speaker 1: The right star, the right instrument, the right observation strategy,
- 51:59Speaker 1: the right analysis techniques in a combination that is extraordinarily rare.
- 52:03Speaker 1: Most stars are not this bright. Most stars are not
- 52:06Speaker 1: this quiet. Most planets do not transit from our line
- 52:09Speaker 1: of sight. Most transit signals this faint are swallowed by
- 52:13Speaker 1: noise before they can ever be extracted. This raises a
- 52:16Speaker 1: question that transforms the story from one planet into something
- 52:19Speaker 1: far larger. We found Kepler-37b because the circumstances were nearly perfect.
- 52:26Speaker 1: But the galaxy contains hundreds of billions of stars where
- 52:29Speaker 1: the circumstances are not perfect. How many of those stars
- 52:33Speaker 1: host moon-sized planets producing transit signals of 10, 8, five
- 52:37Speaker 1: parts per million signals that sit permanently below our detection threshold?
- 52:42Speaker 1: How many worlds exist at scales our instruments cannot reach,
- 52:45Speaker 1: forming a vast invisible population that our catalogues will never contain?
- 52:50Speaker 1: To answer that question, you need to understand how the
- 52:52Speaker 1: word planet itself starts to fracture at these scales, and
- 52:57Speaker 1: why what we call the smallest planet depends entirely on
- 53:00Speaker 1: which question you are asking. The title of this episode
- 53:03Speaker 1: contains a claim, the smallest planet we have ever found.
- 53:07Speaker 1: That claim feels simple. It suggests a clear record, a
- 53:12Speaker 1: definitive answer, a single object sitting at the bottom of
- 53:15Speaker 1: a ranked list. But the moment you press on it,
- 53:18Speaker 1: the claim fractures into pieces, and each piece depends on
- 53:22Speaker 1: a question you have to answer before the ranking means anything.
- 53:25Speaker 1: Smallest by what measure? Confirmed by which method? Planet by
- 53:29Speaker 1: whose definition? Around what type of star? Kepler, 37b-1. is
- 53:35Speaker 1: the smallest confirmed exoplanet by measured radius orbiting a main
- 53:39Speaker 1: sequence star. That sentence contains five qualifiers and every one
- 53:44Speaker 1: of them matters. Remove any single qualifier and the record
- 53:48Speaker 1: either changes hands or dissolves into ambiguity. This is not
- 53:51Speaker 1: a weakness in the discovery. It is a fundamental feature
- 53:55Speaker 1: of what it means to study planets at the smallest scales.
- 53:58Speaker 1: The closer you look at the boundary between planet and
- 54:01Speaker 1: not planet, the less solid that boundary becomes. Start with
- 54:04Speaker 1: the most basic distinction. Radius and mass are not the
- 54:08Speaker 1: same measurement, and they do not rank planets the same way.
- 54:11Speaker 1: Transit photometry, the method that found Kepler-37b, measures radius. It
- 54:17Speaker 1: tells you how large the planet's shadow is. Radial velocity
- 54:20Speaker 1: spectroscopy measures mass. It tells you how strongly the planet
- 54:24Speaker 1: pulls on its star. For most well-characterized exoplanets, astronomers have
- 54:29Speaker 1: both measurements, and the combination yields density, which reveals composition.
- 54:34Speaker 1: But for the smallest worlds, this combination breaks down. The
- 54:38Speaker 1: transit signal becomes vanishingly faint. The radial velocity signal drops
- 54:42Speaker 1: below instrumental sensitivity. You are left with one measurement or
- 54:46Speaker 1: the other, rarely both. Kepler-37b has a measured radius of 0.31
- 54:52Speaker 1: Earth radii. Its mass is unknown. Only an upper limit exists.
- 54:57Speaker 1: This means it holds the record for smallest confirmed radius.
- 55:01Speaker 1: but it cannot compete in a ranking of smallest confirmed
- 55:03Speaker 1: mass because it has no confirmed mass to rank. The
- 55:06Speaker 1: two records, smallest by radius and smallest by mass, are
- 55:10Speaker 1: held by different objects measured by different methods, and they
- 55:13Speaker 1: tell you different things about what small means for a planet.
- 55:17Speaker 1: Consider a world that illustrates the opposite case. GJ 367b
- 55:23Speaker 1: is an ultra-short period planet orbiting a red dwarf star,
- 55:27Speaker 1: roughly 31 light years from Earth. It was discovered by
- 55:30Speaker 1: TESS in 2021 and subsequently characterized with intensive radial velocity
- 55:35Speaker 1: observations using the HARPS spectrograph. Unlike Kepler-37b, GJ367b has both
- 55:43Speaker 1: a measured radius and a measured mass. Its radius is
- 55:47Speaker 1: approximately 0.7 Earth radii, more than twice the radius of
- 55:51Speaker 1: Kepler-37b and somewhat larger than Mars. By radius alone, it
- 55:57Speaker 1: is not particularly remarkable. Dozens of confirmed exoplanets have similar sizes,
- 56:02Speaker 1: but its mass is striking. GJ367b weighs approximately 0.63 Earth masses,
- 56:10Speaker 1: and its bulk density is roughly 10 grams per cubic centimetre,
- 56:14Speaker 1: nearly twice the average density of Earth and denser than
- 56:17Speaker 1: uncompressed iron. Interior models suggest this planet has an iron
- 56:21Speaker 1: core occupying roughly 86% of its radius, a structure strikingly
- 56:26Speaker 1: similar to Mercury's but more extreme. It completes one orbit
- 56:30Speaker 1: every 7.7 hours. Not days, hours. Its dayside temperature exceeds 1,700 Kelvin.
- 56:39Speaker 1: It is a ball of molten iron smaller than Mars,
- 56:42Speaker 1: whipping around its star faster than the hour hand on
- 56:45Speaker 1: a clock. GJ 367b is not the smallest exoplanet by radius,
- 56:52Speaker 1: but it is one of the smallest with a precisely
- 56:54Speaker 1: measured mass, and that mass measurement reveals something the radius
- 56:58Speaker 1: alone cannot. It tells you the planet is real in
- 57:00Speaker 1: a way that a shadow measurement, however precise, does not
- 57:04Speaker 1: fully convey. Mass is substance. Mass is gravitational influence. Mass
- 57:09Speaker 1: is the quantity that determines whether an object dominates its
- 57:12Speaker 1: orbital neighbourhood, holds an atmosphere, or generates internal heat through
- 57:17Speaker 1: radioactive decay. Radius without mass is geometry. Mass without radius
- 57:22Speaker 1: is weight. You need both to understand what a world
- 57:25Speaker 1: actually is. This distinction might sound academic. but it has
- 57:29Speaker 1: practical consequences for how we interpret the smallest planets. When
- 57:33Speaker 1: astronomers say Kepler-37b is the smallest confirmed planet, they are
- 57:37Speaker 1: making a statement about a geometric measurement the size of
- 57:40Speaker 1: a shadow cast onto a detector. They are not making
- 57:43Speaker 1: a statement about the planet's gravitational significance, its internal structure,
- 57:48Speaker 1: or its place in a formation model. They are saying,
- 57:52Speaker 1: this is the smallest confirmed silhouette. It is an important record,
- 57:56Speaker 1: but it is a specific kind of record, and conflating
- 58:00Speaker 1: it with smallest planet in every sense would be misleading.
- 58:03Speaker 1: The confusion deepens when you look at the full roster
- 58:06Speaker 1: of small confirmed exoplanets. Below Kepler-37b 0.3, one Earth radii,
- 58:13Speaker 1: there are no confirmed planets. The next smallest confirmed transiting
- 58:17Speaker 1: planets are Kepler-444b at approximately 0.4 Earth radii, and Kepler-1
- 58:25Speaker 1: O2b at a similar size. The Kepler 444 system deserves
- 58:29Speaker 1: particular attention because it demonstrates that sub-Earth planets are not
- 58:33Speaker 1: confined to isolated examples. Kepler 444 is a K-type star,
- 58:39Speaker 1: roughly 117 light years from Earth, in the constellation Lyra,
- 58:43Speaker 1: the same constellation as Kepler 37. The star is ancient,
- 58:48Speaker 1: estimated at roughly 11 billion years old, making it one
- 58:51Speaker 1: of the oldest known planet-hosting stars in the galaxy. It
- 58:55Speaker 1: harbors five confirmed transiting planets, all with orbital periods shorter
- 58:59Speaker 1: than 10 days and radii ranging from approximately 0.4 to 0.7
- 59:04Speaker 1: Earth radii. Every planet in this system is smaller than Earth.
- 59:09Speaker 1: Several are smaller than Venus. The two innermost worlds, Kepler-444b
- 59:14Speaker 1: and Kepler-444c, are among the smallest confirmed exoplanets after Kepler-37b.
- 59:21Speaker 1: Their existence in a system this old demonstrates that sub-Earth
- 59:25Speaker 1: planets can form early in cosmic and survive for the
- 59:29Speaker 1: better part of the age of the universe. They are
- 59:31Speaker 1: not transient or fragile objects. They are stable, long-lived, and
- 59:36Speaker 1: apparently common enough that Kepler found multiple examples in a
- 59:40Speaker 1: single field of view. Then come a handful of others
- 59:43Speaker 1: in the 0.4 to 0.5 Earth radii range, Kepler-444c, a
- 59:49Speaker 1: few ultra-short period worlds detected in recent reanalyses of Kepler data.
- 59:54Speaker 1: One study using advanced computational techniques identified candidates such as Kepler-879c, Kepler-158d,
- 1:00:02Speaker 1: and Kepler-963c, with radii between 0.5 and 0.7 Earth radii,
- 1:00:10Speaker 1: orbiting K and G dwarf stars in ultra-short periods of
- 1:00:13Speaker 1: less than one day. These are among the smallest ultra-short
- 1:00:16Speaker 1: period planets ever identified, but many of these remain candidates
- 1:00:21Speaker 1: requiring further validation, not yet promoted to the status of
- 1:00:24Speaker 1: confirmed planets. The boundary between candidate and confirmed is itself
- 1:00:30Speaker 1: a judgment call, a threshold of statistical confidence that different
- 1:00:35Speaker 1: teams may set differently. And here lies a subtlety that
- 1:00:38Speaker 1: most discussions of exoplanet records gloss over. Confirmation is not
- 1:00:42Speaker 1: a binary property of nature. It is a judgment made
- 1:00:46Speaker 1: by scientists about how confident they are that a signal
- 1:00:49Speaker 1: is real. Different validation methods yield different levels of confidence.
- 1:00:53Speaker 1: A planet confirmed by both transit photometry and radial velocity
- 1:00:58Speaker 1: mass measurement is confirmed in a stronger sense than a
- 1:01:01Speaker 1: planet validated only by statistical arguments against false positives. Kepler-37b
- 1:01:08Speaker 1: falls into the second category. Its confirmation rests on the
- 1:01:11Speaker 1: statistical improbability of all non-planetary explanations for its transit signal,
- 1:01:16Speaker 1: combined with the consistency of its transit depth across optical
- 1:01:20Speaker 1: and infrared wavelengths. This is rigorous and widely accepted, but
- 1:01:25Speaker 1: it is not the same as an independent mass detection.
- 1:01:28Speaker 1: If a future instrument with sufficient sensitivity were to fail
- 1:01:31Speaker 1: to detect any radial velocity signal from Kepler-37b, it would
- 1:01:36Speaker 1: not necessarily disprove the planet. The expected signal is simply
- 1:01:40Speaker 1: too small, but it would leave the confirmation resting entirely
- 1:01:43Speaker 1: on the photometric evidence. This matters because it highlights the
- 1:01:47Speaker 1: asymmetry that defines our knowledge of the smallest exoplanets. The
- 1:01:51Speaker 1: smallest worlds are the hardest to confirm independently. They produce
- 1:01:54Speaker 1: the faintest transit signals, the weakest radial velocity signatures, and
- 1:01:59Speaker 1: the smallest astrometric wobbles. Every complementary detection method becomes less
- 1:02:04Speaker 1: effective as the planet shrinks. At some point, you reach
- 1:02:08Speaker 1: a scale where the only evidence for a planet's existence
- 1:02:11Speaker 1: is a barely detectable dip in starlight, validated against false
- 1:02:15Speaker 1: positives by statistical argument alone. That is the regime Kepler-37b inhabits.
- 1:02:21Speaker 1: And it is the regime where the word confirmed starts
- 1:02:24Speaker 1: to carry a weight that depends on how much confidence
- 1:02:27Speaker 1: you are willing to place in statistical validation without independent corroboration.
- 1:02:32Speaker 1: None of this casts doubt on Kepler-37b specifically. The discovery
- 1:02:36Speaker 1: paper is thorough, the validation is rigorous, and the astronomical
- 1:02:41Speaker 1: community has broadly accepted the result. But it does illustrate
- 1:02:45Speaker 1: a general principle At the smallest planetary scales, our knowledge
- 1:02:50Speaker 1: becomes thinner. We know less about each individual world because
- 1:02:54Speaker 1: our instruments can measure less. The record for smallest planet
- 1:02:57Speaker 1: is held by the object about which we know the least.
- 1:03:00Speaker 1: That is not a coincidence. It is a consequence of
- 1:03:03Speaker 1: the physics of detection. This brings us to the force
- 1:03:06Speaker 1: that shapes our entire understanding of small planets more than
- 1:03:10Speaker 1: any other factor, detection bias. Every method for finding exoplanets
- 1:03:14Speaker 1: has blind spots. Transit photometry cannot detect planets whose orbits
- 1:03:19Speaker 1: are not aligned with our line of sight. For a
- 1:03:21Speaker 1: randomly oriented planetary system, the probability that a given planet
- 1:03:25Speaker 1: transits as seen from Earth depends on the ratio of
- 1:03:28Speaker 1: the star's radius to the planet's orbital distance. The relationship
- 1:03:31Speaker 1: is straightforward. The transit probability is approximately equal to the
- 1:03:36Speaker 1: star's radius divided by the planet's orbital distance. For a
- 1:03:40Speaker 1: Sun-like star with a radius of about 700,000 kilometers, and
- 1:03:44Speaker 1: a planet orbiting at one astronomical unit roughly 150 million kilometres,
- 1:03:50Speaker 1: the probability is 700,000 divided by 150 million, which gives
- 1:03:55Speaker 1: approximately 0.005, or one half of one percent. That means
- 1:04:01Speaker 1: for every 200 randomly oriented planetary systems with a planet
- 1:04:05Speaker 1: at Earth's orbital distance, only one will be aligned, so
- 1:04:09Speaker 1: that we see a transit. The other 199 systems contain
- 1:04:13Speaker 1: the planet but hide it from our view. Move the
- 1:04:16Speaker 1: planet to Mercury's orbital distance, about 0.4 astronomical units, and
- 1:04:21Speaker 1: the probability rises to roughly 1.2%. Move it to 0.1.
- 1:04:28Speaker 1: Astronomical units, the orbital distance of Kepler-37b, and it climbs
- 1:04:33Speaker 1: to roughly 4%, around a star the size of Kepler-37.
- 1:04:37Speaker 1: Closer orbits improve the odds, which is one reason transit
- 1:04:41Speaker 1: surveys are so heavily biased towards short-period planets. But even
- 1:04:45Speaker 1: at 4%, 96 out of every 100 Kepler-37b analogues would
- 1:04:50Speaker 1: be invisible to transit photometry regardless of how sensitive the
- 1:04:53Speaker 1: telescope is. The geometry alone hides the vast majority. The
- 1:04:58Speaker 1: relationship is straightforward. The transit probability is approximately equal to
- 1:05:03Speaker 1: the star's radius divided by the planet's orbital distance. For
- 1:05:07Speaker 1: a Sun-like star with a radius of about 700,000 kilometres
- 1:05:11Speaker 1: and a planet orbiting at one astronomical unit, roughly 150
- 1:05:16Speaker 1: million kilometres, the probability is 700,000 divided by 150 million,
- 1:05:23Speaker 1: which gives approximately 0.005, or one half of one percent.
- 1:05:29Speaker 1: That means for every 200 randomly oriented planetary systems with
- 1:05:32Speaker 1: a planet at Earth's orbital distance, only one will be aligned.
- 1:05:37Speaker 1: so that we see a transit. The other 199 systems
- 1:05:40Speaker 1: contain the planet, but hide it from our view. Move
- 1:05:43Speaker 1: the planet to Mercury's orbital distance, about 0.4 astronomical units,
- 1:05:49Speaker 1: and the probability rises to roughly 1.2%. Move it to Kepler,
- 1:05:54Speaker 1: 37B' s orbital distance of 0.1 astronomical units, and it climbs
- 1:06:00Speaker 1: to roughly 5%. Closer orbits improve the odds, which is
- 1:06:04Speaker 1: one reason transit surveys are so heavily biased toward short-period planets.
- 1:06:08Speaker 1: But even at 5%, 95 out of every 100 Kepler-37b
- 1:06:14Speaker 1: analogues would be invisible to transit photometry, regardless of how
- 1:06:18Speaker 1: sensitive the telescope is. The geometry alone hides the vast majority.
- 1:06:23Speaker 1: For a planet orbiting at Earth's distance from a sun-like star,
- 1:06:26Speaker 1: the transit probability is roughly one-half of 1%. Move the
- 1:06:31Speaker 1: planet closer and the probability rises. Move it farther and
- 1:06:35Speaker 1: it drops. This geometric selection effect means that transit surveys
- 1:06:40Speaker 1: overwhelmingly detect planets on short-period orbits. Long-period planets, those orbiting
- 1:06:46Speaker 1: at Earth-like distances or beyond transit so rarely from any
- 1:06:49Speaker 1: given line of sight that they are severely underrepresented in
- 1:06:52Speaker 1: transit catalogues. But the bias does not stop at orbital period.
- 1:06:57Speaker 1: Transit depth depends on the square of the planet-to-star radius ratio.
- 1:07:01Speaker 1: A planet half the radius of Earth produces a transit
- 1:07:05Speaker 1: only one-quarter as deep. A planet one-third the radius of
- 1:07:08Speaker 1: Earth produces a transit one-ninth as deep. The signal drops
- 1:07:12Speaker 1: precipitously with decreasing size, and at some point it falls
- 1:07:16Speaker 1: below the noise floor of the instrument. For the Kepler
- 1:07:19Speaker 1: Space Telescope observing a typical sun-like star, The practical detection
- 1:07:24Speaker 1: limit for individual transit events was roughly in the range
- 1:07:27Speaker 1: of 20 to 50 parts per million, depending on stellar
- 1:07:30Speaker 1: brightness and variability. Only by folding many transits together could
- 1:07:34Speaker 1: Kepler push below this threshold, and even then, the star
- 1:07:38Speaker 1: had to be exceptionally bright and quiet. This means the
- 1:07:41Speaker 1: catalogue of confirmed exoplanets is not a random sample of
- 1:07:45Speaker 1: what exists. It is a filtered sample, shaped by the
- 1:07:49Speaker 1: geometry of orbital alignment, the physics of transit depth, the
- 1:07:53Speaker 1: brightness and behaviour of host stars, and the duration and
- 1:07:56Speaker 1: sensitivity of the surveys that produce the data. Every exoplanet
- 1:08:00Speaker 1: catalogue in existence is a distorted mirror of reality, reflecting
- 1:08:05Speaker 1: the large and the close-in with high fidelity, while barely
- 1:08:09Speaker 1: hinting at the small and the distant. The distortion is
- 1:08:12Speaker 1: most severe at the smallest sizes. Statistical analyses of Kepler
- 1:08:16Speaker 1: data attempted to correct for detection bias and estimate the
- 1:08:21Speaker 1: true occurrence rates of planets at various sizes and orbital periods.
- 1:08:25Speaker 1: These studies consistently find that small rocky planets are the
- 1:08:29Speaker 1: most common type of planet in the galaxy. The occurrence
- 1:08:32Speaker 1: rate rises steeply as planet radius decreases from roughly four
- 1:08:36Speaker 1: Earth radii down to about one Earth radius. Below one
- 1:08:39Speaker 1: Earth radius the measured occurrence rate drops, but this drop
- 1:08:43Speaker 1: is almost certainly an artifact of declining sensitivity rather than
- 1:08:47Speaker 1: a real absence of planets. The instruments are losing their
- 1:08:51Speaker 1: grip on the signal, and the counts fall not because
- 1:08:53Speaker 1: the planets stop existing, but because the telescopes stop seeing them.
- 1:08:57Speaker 1: One of the most influential occurrence rate studies using the
- 1:09:01Speaker 1: full Kepler dataset, estimated that roughly 1 in 4 Sun-like
- 1:09:05Speaker 1: stars hosts a planet with a radius between 1 and
- 1:09:08Speaker 1: 2 Earth radii, and an orbital period shorter than roughly
- 1:09:13Speaker 1: 400 days. For planets between 0.5 and 1 Earth radii,
- 1:09:18Speaker 1: the sub-Earth regime, the estimates become less certain, but they
- 1:09:21Speaker 1: suggest these worlds may be even more common. The data
- 1:09:25Speaker 1: hint that the true planet radius distribution continues rising below
- 1:09:28Speaker 1: 1 Earth radius, meaning sub-Mercury worlds may be abundant. But
- 1:09:33Speaker 1: the error bars grow enormous because Kepler simply could not
- 1:09:36Speaker 1: detect enough of them to pin the numbers down. The
- 1:09:39Speaker 1: method astronomers use to reach these conclusions deserves attention. because
- 1:09:43Speaker 1: it transforms the bias argument from an abstract claim into
- 1:09:46Speaker 1: a measured quantity. The technique is called injection recovery testing,
- 1:09:51Speaker 1: and it works by deliberately lying to the detection pipeline,
- 1:09:55Speaker 1: then measuring how often the pipeline catches the lie. Here
- 1:09:58Speaker 1: is what the Kepler team did. They took the real
- 1:10:01Speaker 1: photometric data, the actual brightness measurements of real stars, observed
- 1:10:05Speaker 1: over the mission lifetime, and injected thousands of synthetic transit
- 1:10:09Speaker 1: signals into it. These fake transits had known depths, known
- 1:10:13Speaker 1: periods and known durations, corresponding to simulated planets of specified
- 1:10:18Speaker 1: radii and orbital configurations. The modified data, now containing both
- 1:10:23Speaker 1: real signals and planted fakes, was then processed through the
- 1:10:26Speaker 1: same automated detection pipeline used to find genuine planet candidates.
- 1:10:32Speaker 1: At the end, the team compared how many injected signals
- 1:10:35Speaker 1: the pipeline successfully recovered against how many it missed. The
- 1:10:38Speaker 1: results were sobering. For planets with radii above 2 Earth
- 1:10:41Speaker 1: radii and orbital periods shorter than about 100 days, the
- 1:10:45Speaker 1: recovery rate was high, typically above 80%. The pipeline found
- 1:10:50Speaker 1: most of what was planted. For Earth-sized planets at similar periods,
- 1:10:54Speaker 1: the recovery rate dropped to roughly 50-60%. The pipeline was
- 1:10:59Speaker 1: losing about half the signals. For sub-Earth planets radii between 0.5
- 1:11:03Speaker 1: and 1 Earth radius, the recovery rate fell further, in
- 1:11:07Speaker 1: many cases below 30%. And for planets at the scale
- 1:11:10Speaker 1: of Kepler-37b, 0.3 Earth radii, recovery was possible only under
- 1:11:16Speaker 1: the most favorable circumstances. Very bright, very quiet stars with
- 1:11:20Speaker 1: many accumulated transits. Across the broader stellar sample, the pipeline
- 1:11:25Speaker 1: would miss nearly all of them. This is not speculation.
- 1:11:29Speaker 1: It is a direct quantitative measurement of incompleteness. It proves
- 1:11:33Speaker 1: that when the observed planet count drops sharply below one
- 1:11:36Speaker 1: Earth radius, The drop is driven overwhelmingly by the pipeline
- 1:11:39Speaker 1: losing sensitivity, not by a real absence of planets in nature.
- 1:11:44Speaker 1: The planets are there. The signals are in the data.
- 1:11:47Speaker 1: The detection machinery simply cannot extract them. Every occurrence rate
- 1:11:51Speaker 1: estimate for small rocky planets depends on this injection recovery
- 1:11:55Speaker 1: calibration to correct the raw counts upward. And the corrections
- 1:11:58Speaker 1: at the smallest radii are enormous factors of 3, 5,
- 1:12:01Speaker 1: 10 or more. The true population of sub-Earth worlds is
- 1:12:05Speaker 1: not slightly larger than what the raw catalogue shows. It
- 1:12:08Speaker 1: is vastly larger. TESS, the transiting exoplanet survey satellite that
- 1:12:13Speaker 1: succeeded Kepler, brought different strengths and different limitations to the
- 1:12:18Speaker 1: search for small planets. Launched in April 2018, TESS surveys
- 1:12:23Speaker 1: nearly the entire sky rather than staring at a single field.
- 1:12:27Speaker 1: It observes much brighter and closer stars than Kepler did.
- 1:12:30Speaker 1: making follow-up characterisation including radial velocity mass measurements far more practical.
- 1:12:35Speaker 1: But TESS observes most of its sky for only about
- 1:12:38Speaker 1: 27 days at a time before moving to the next sector.
- 1:12:41Speaker 1: This short baseline means TESS typically sees only one or
- 1:12:44Speaker 1: two transits for planets on orbits longer than about two weeks.
- 1:12:48Speaker 1: For short-period planets, TESS can detect multiple transits, but its
- 1:12:52Speaker 1: smaller aperture and broader survey strategy mean it generally achieves
- 1:12:56Speaker 1: lower photometric precision per star than Kepler did. The result is...
- 1:13:00Speaker 1: that TESS excels at finding relatively large planets on short
- 1:13:04Speaker 1: orbits around bright, nearby stars worlds ripe for atmospheric characterisation
- 1:13:10Speaker 1: by the James Webb Space Telescope and ground-based spectrographs. But
- 1:13:15Speaker 1: TESS struggles with the smallest planets. The reason is partly
- 1:13:19Speaker 1: aperture and partly observation duration. Kepler's primary mirror measured 0.95
- 1:13:23Speaker 1: metres in diameter. and collected light from the same stars
- 1:13:28Speaker 1: continuously for up to four years. For a bright, quiet
- 1:13:31Speaker 1: star like Kepler-37, this combination yielded an effective photometric precision
- 1:13:37Speaker 1: after folding dozens of transits that pushed below 10 parts
- 1:13:40Speaker 1: per million. TESS uses four cameras with lens apertures of
- 1:13:44Speaker 1: only 10.5 centimeters each. Its per exposure photometric precision for
- 1:13:49Speaker 1: a given star is considerably lower than Kepler's, and its
- 1:13:52Speaker 1: standard observation window of 27 days limits the number of
- 1:13:56Speaker 1: transits that can be folded together. For a planet with
- 1:13:59Speaker 1: a 13-day orbital period like Kepler-37b, TESS would capture at
- 1:14:03Speaker 1: most two transits in a single sector, not nearly enough
- 1:14:07Speaker 1: to extract a 12-part-per-million signal from the noise. Stars in
- 1:14:11Speaker 1: TESS's continuous viewing zones near the ecliptic poles receive up
- 1:14:15Speaker 1: to a year of coverage, which improves the situation. But
- 1:14:18Speaker 1: even there, the effective sensitivity for submercury transits around sun-like
- 1:14:23Speaker 1: stars remains substantially worse than what Kepler achieved for its
- 1:14:27Speaker 1: best targets. TS compensates by surveying far more stars across
- 1:14:32Speaker 1: nearly the entire sky, increasing the chance of finding rare
- 1:14:36Speaker 1: favourable configurations. But the trade-off is real. Breadth for depth,
- 1:14:40Speaker 1: coverage for sensitivity. The smallest worlds require depth. Even GJ 367b,
- 1:14:48Speaker 1: discussed earlier and the smallest planet confirmed through test data,
- 1:14:52Speaker 1: has a radius of 0.7 Earth radii, still more than
- 1:14:56Speaker 1: twice the radius of Kepler-37b. Pushing test detections into the
- 1:15:01Speaker 1: true sub-Mercury regime below 0.4 Earth radii remains beyond the
- 1:15:06Speaker 1: mission's practical sensitivity for most of its target stars. Pushing
- 1:15:10Speaker 1: test detections below the sub-Earth threshold requires exceptionally favourable conditions.
- 1:15:16Speaker 1: Ultra-short orbital periods that produce many transits during a single
- 1:15:20Speaker 1: sector observation host stars that are both bright and inactive,
- 1:15:24Speaker 1: and planetary radii large enough to produce detectable transit depths
- 1:15:28Speaker 1: on the TESS aperture. Ultra-short period planets, worlds with orbital
- 1:15:33Speaker 1: periods shorter than one day, represent one window into this regime.
- 1:15:38Speaker 1: Because they transit so frequently, dozens of events can accumulate
- 1:15:42Speaker 1: even during short observing windows, improving the signal-to-noise ratio for
- 1:15:46Speaker 1: tiny planets. Several ultra-short period planets with sub-Earth radii have
- 1:15:51Speaker 1: been identified in Kepler data through reanalysis with advanced computational techniques.
- 1:15:56Speaker 1: These include some of the smallest detected transit signals ever reported,
- 1:16:00Speaker 1: with radii as low as 0.5 to 0.6 Earth radii,
- 1:16:05Speaker 1: orbiting G and K dwarf stars. The planets orbit so
- 1:16:09Speaker 1: close to their stars, sometimes within five stellar radii of
- 1:16:13Speaker 1: the photosphere, that they exist in an environment of extreme
- 1:16:16Speaker 1: radiation and tidal forces. Their surfaces are almost certainly molten.
- 1:16:21Speaker 1: Their compositions may be dominated by refractory materials that survive
- 1:16:25Speaker 1: the intense heat. They are not habitable. They are not Earth-like.
- 1:16:29Speaker 1: They are tiny remnants orbiting in regimes so extreme that
- 1:16:33Speaker 1: they test the limits of planet formation theory. But the
- 1:16:36Speaker 1: very existence of these objects makes a powerful statistical argument.
- 1:16:40Speaker 1: If planet formation can produce bodies this small on ultra-short
- 1:16:44Speaker 1: period orbits, the orbits most favourable for transit detection, then
- 1:16:48Speaker 1: it almost certainly produces similar bodies on longer period orbits
- 1:16:51Speaker 1: where detection is far harder. The ultra-short period planets we
- 1:16:56Speaker 1: see are the survivors of an observational filter so aggressive
- 1:16:59Speaker 1: that it admits only the most detectable fraction of a
- 1:17:02Speaker 1: much larger population. For every sub-Earth world on a six-hour
- 1:17:06Speaker 1: orbit that we manage to spot, there may be hundreds
- 1:17:09Speaker 1: on 13-day orbits like Kepler-37b, thousands on orbits of months,
- 1:17:14Speaker 1: and untold numbers on orbits of years all producing transit
- 1:17:17Speaker 1: signals too faint, too infrequent, or too poorly aligned to
- 1:17:21Speaker 1: ever reach our catalogues. This is the core of the
- 1:17:24Speaker 1: detection bias problem. Our picture of the planetary population is
- 1:17:28Speaker 1: drawn by the instruments we use, not by the universe
- 1:17:31Speaker 1: we are trying to map. It is as if you
- 1:17:33Speaker 1: tried to census the wildlife in a forest by counting
- 1:17:36Speaker 1: only the animals that crossed a particular clearing during daylight hours.
- 1:17:41Speaker 1: You would find deer and rabbits and birds. You would
- 1:17:44Speaker 1: conclude that the forest is full of large, conspicuous animals.
- 1:17:48Speaker 1: You would entirely miss the mice, the insects, the creatures
- 1:17:51Speaker 1: that come out only at night, the species that live
- 1:17:54Speaker 1: in the canopy and never touch the ground. Your census
- 1:17:57Speaker 1: would not be wrong about the animals it did find,
- 1:18:00Speaker 1: but it would be catastrophically incomplete about the ones it missed.
- 1:18:03Speaker 1: and the species it missed would outnumber the ones it
- 1:18:05Speaker 1: found by orders of magnitude. The exoplanet census is in
- 1:18:09Speaker 1: exactly this position. We have confirmed more than 5,000 exoplanets.
- 1:18:15Speaker 1: The vast majority are larger than Earth. The vast majority
- 1:18:18Speaker 1: orbit closer to their stars than Earth orbits our Sun.
- 1:18:21Speaker 1: The vast majority were found around stars bright enough and
- 1:18:24Speaker 1: quiet enough to permit transit detection or radial velocity measurement.
- 1:18:28Speaker 1: The catalogue is real. Every confirmed planet in it exists.
- 1:18:33Speaker 1: but the catalogue is not representative. It is dominated by
- 1:18:36Speaker 1: the large and the close-in, because those are the worlds
- 1:18:39Speaker 1: our instruments find most easily. The small, the distant, the faint,
- 1:18:44Speaker 1: the Kepler-37b analogues on longer orbits around dimmer, noisier stars
- 1:18:49Speaker 1: are almost entirely absent. Not because they do not exist,
- 1:18:52Speaker 1: but because we cannot see them yet. The implications are staggering.
- 1:18:57Speaker 1: If occurrence rate estimates are even approximately correct, the Milky
- 1:19:01Speaker 1: Way may contain tens of billions of sub-Earth rocky planets.
- 1:19:05Speaker 1: Worlds smaller than Mercury. Worlds the size of the Moon.
- 1:19:09Speaker 1: Worlds at scales we instinctively associate with moons and asteroids
- 1:19:13Speaker 1: rather than planets. These objects would be legitimate planets by
- 1:19:16Speaker 1: any current working definition bodies formed through accretion, orbiting stars,
- 1:19:22Speaker 1: too small for nuclear fusion, but they would be invisible
- 1:19:24Speaker 1: to every survey we have conducted, every telescope we have built,
- 1:19:29Speaker 1: every catalogue we have compiled. Kepler-37b is the single representative
- 1:19:34Speaker 1: of this population that happened to be detectable It is
- 1:19:37Speaker 1: not the archetype. It is the exception, the one member
- 1:19:41Speaker 1: of a vast class of objects that was orbiting the
- 1:19:44Speaker 1: right star, at the right period, with the right alignment,
- 1:19:48Speaker 1: observed by the right instrument for the right duration. Everything
- 1:19:52Speaker 1: about its detection was contingent on circumstances that will not
- 1:19:56Speaker 1: repeat for most of the tiny planets in the galaxy.
- 1:19:59Speaker 1: And yet it is treated as a record holder, as
- 1:20:02Speaker 1: if it occupies the extreme tail of the distribution. The
- 1:20:05Speaker 1: truth may be the opposite. Kepler-37b may be perfectly ordinary
- 1:20:11Speaker 1: among worlds its size. It only looks extraordinary because it
- 1:20:15Speaker 1: is the only one we manage to see. The boundary
- 1:20:17Speaker 1: between what we detect and what exists is not a
- 1:20:20Speaker 1: minor technical detail. It is the central tension of modern
- 1:20:24Speaker 1: exoplanet science. Every conclusion we draw about the frequency, distribution
- 1:20:29Speaker 1: and properties of small planets is shaped by the instruments
- 1:20:32Speaker 1: that collected the data. Change the instrument and you change
- 1:20:35Speaker 1: the picture. Extend the observation baseline and new populations appear.
- 1:20:40Speaker 1: Improve the photometric precision by a factor of two and
- 1:20:44Speaker 1: an entire hidden layer of small worlds rises above the
- 1:20:47Speaker 1: noise floor. The planetary population we know is a function
- 1:20:51Speaker 1: of our technology, not a function of reality. And our technology,
- 1:20:55Speaker 1: as precise as it has become, is still profoundly limited
- 1:20:59Speaker 1: when it comes to the smallest worlds. This is why
- 1:21:02Speaker 1: future missions matter so much. Future missions may begin to
- 1:21:05Speaker 1: change this, New space telescopes designed for higher precision transit
- 1:21:09Speaker 1: photometry and entirely different detection methods like gravitational microlensing are
- 1:21:14Speaker 1: expected to push the sensitivity floor substantially lower within the
- 1:21:18Speaker 1: next decade, potentially revealing sub-Earth populations that current surveys cannot reach.
- 1:21:24Speaker 1: But those missions have not launched yet. For now, we
- 1:21:27Speaker 1: are working with catalogues shaped almost entirely by Kepler and
- 1:21:31Speaker 1: TESS catalogues, that are complete for large planets and profoundly
- 1:21:35Speaker 1: incomplete for small ones. Each new mission will push the
- 1:21:38Speaker 1: detection boundary slightly lower, revealing another slice of the planetary
- 1:21:42Speaker 1: population hidden beneath our current noise floor. But none of
- 1:21:46Speaker 1: them will find everything. The smallest planets, worlds with radii
- 1:21:50Speaker 1: below the moons, worlds with masses below a hundredth of
- 1:21:53Speaker 1: an Earth mass, may remain permanently beyond reach. There is
- 1:21:57Speaker 1: no instrument currently planned or proposed that could detect a
- 1:22:01Speaker 1: series-sized planet transiting a sun-like star at an Earth-like orbital distance.
- 1:22:07Speaker 1: The transit depth would be less than one part per million.
- 1:22:10Speaker 1: The radial velocity signal would be measured in micrometers per second.
- 1:22:14Speaker 1: These worlds may exist in enormous numbers, orbiting quietly, pulling weakly,
- 1:22:20Speaker 1: casting shadows too faint for any conceivable photometer to catch.
- 1:22:24Speaker 1: The word planet was not designed for this regime. It
- 1:22:27Speaker 1: was coined by ancient Greeks who watched bright points of
- 1:22:30Speaker 1: light wander against the fixed stars. It was refined by
- 1:22:33Speaker 1: centuries of solar system observation, where planets were large, bright
- 1:22:38Speaker 1: and unmistakable. It was extended to exoplanets as telescopes grew
- 1:22:42Speaker 1: powerful enough to detect Jupiter-sized companions around nearby stars. At
- 1:22:46Speaker 1: every stage, the word carried implicit assumptions about size and
- 1:22:50Speaker 1: significance that the smallest worlds now violate. Kepler, 37b is
- 1:22:56Speaker 1: called a planet. but it occupies a size regime where
- 1:22:59Speaker 1: the distinction between planet and large asteroid depends on context,
- 1:23:03Speaker 1: on definition, on what question you happen to be asking.
- 1:23:07Speaker 1: This is not a problem that better definitions will solve.
- 1:23:10Speaker 1: It is a feature of nature. Rocky bodies form across
- 1:23:14Speaker 1: a continuous spectrum of sizes, from dust grains to super-Earths.
- 1:23:19Speaker 1: There is no physical threshold where small rocky body suddenly
- 1:23:23Speaker 1: becomes planet. The transition is gradual. and any line we
- 1:23:27Speaker 1: draw across it is arbitrary. The question of how small
- 1:23:30Speaker 1: a planet can be is ultimately not a question about physics.
- 1:23:34Speaker 1: It is a question about language, and the language is
- 1:23:37Speaker 1: starting to fail. What remains is the deeper question, the
- 1:23:41Speaker 1: one that Kepler-37b forces us to confront, not as a
- 1:23:45Speaker 1: definition problem, but as a statement about the universe itself.
- 1:23:49Speaker 1: If planet formation produces bodies at every scale, and our
- 1:23:53Speaker 1: instruments can only detect the largest fraction of that production,
- 1:23:56Speaker 1: then what does the existence of one barely detectable moon-sized
- 1:24:00Speaker 1: world tell us about the invisible population it represents? Planets
- 1:24:05Speaker 1: are not placed into orbits by design. They are built
- 1:24:08Speaker 1: from the bottom up, grain by grain, collision by collision,
- 1:24:12Speaker 1: in a process that begins with particles smaller than sand
- 1:24:16Speaker 1: and ends with worlds that can span 100,000 kilometers. Understanding
- 1:24:21Speaker 1: how Kepler-37b came to exist requires understanding this construction process,
- 1:24:27Speaker 1: because the process itself explains why tiny rocky planets are
- 1:24:31Speaker 1: not anomalies. They are inevitable. Every planet begins in a
- 1:24:35Speaker 1: protoplanetary disk, a flat rotating cloud of gas and dust
- 1:24:39Speaker 1: surrounding a newly formed star. The disk forms simultaneously with
- 1:24:43Speaker 1: the star, collapsing from the same molecular cloud. Within the disk,
- 1:24:48Speaker 1: solid particles of silicate metal and ice orbit the young
- 1:24:52Speaker 1: star at various distances, drifting through the gas, colliding with
- 1:24:56Speaker 1: each other. These collisions are the first step. When two
- 1:25:00Speaker 1: dust grains collide at low enough velocity, they stick. Electrostatic
- 1:25:04Speaker 1: forces and surface chemistry hold them together. The resulting clump
- 1:25:08Speaker 1: is slightly larger, slightly more massive, and slightly better at
- 1:25:11Speaker 1: sweeping up additional grains. This process continues through orders of
- 1:25:15Speaker 1: magnitude and size. Millimeter scale particles grow to centimeter scale pebbles.
- 1:25:21Speaker 1: Pebbles accumulate into meter scale boulders. Boulders aggregate into kilometer
- 1:25:26Speaker 1: scale planetesimals. At the planetesimal stage, gravity takes over. A
- 1:25:31Speaker 1: body a few kilometers across has enough mass that its
- 1:25:34Speaker 1: own gravitational pull, though feeble, begins to attract nearby material.
- 1:25:39Speaker 1: It no longer needs to rely on random collisions. It
- 1:25:42Speaker 1: reaches out and pulls. The more mass it accumulates, the
- 1:25:46Speaker 1: stronger its gravitational reach becomes and the faster it grows.
- 1:25:50Speaker 1: This runaway accretion transforms kilometre-scale seeds into embryos hundreds or
- 1:25:56Speaker 1: thousands of kilometres across, over timescales of roughly 100,000 to
- 1:26:01Speaker 1: a few million years. These embryos are the building blocks
- 1:26:04Speaker 1: of rocky planets. What happens next depends on where the
- 1:26:08Speaker 1: embryo is and how much material surrounds it. In the
- 1:26:11Speaker 1: inner regions of a protoplanetary disk, where temperatures are high
- 1:26:14Speaker 1: and only refractory materials like rock and metal can survive
- 1:26:17Speaker 1: in solid form, the available mass is limited. Embryos compete
- 1:26:22Speaker 1: for a finite supply of solids. They collide with each
- 1:26:25Speaker 1: other in giant impacts that can shatter, merge, or partially
- 1:26:29Speaker 1: strip both bodies. Over tens of millions of years, these
- 1:26:32Speaker 1: violent collisions reduced dozens of embryos into a handful of survivors.
- 1:26:37Speaker 1: Those survivors are the terrestrial planets. In our solar system,
- 1:26:40Speaker 1: this process produced four rocky worlds. Mercury, Venus, Earth, and Mars.
- 1:26:46Speaker 1: Their sizes span a range from roughly 5,000 kilometers to
- 1:26:50Speaker 1: nearly 13,000 kilometers in diameter. But this range is not
- 1:26:54Speaker 1: a fundamental limit. It reflects the specific mass budget and
- 1:26:58Speaker 1: orbital architecture of our particular protoplanetary disk. A disk with
- 1:27:02Speaker 1: less solid material in its inner region would produce smaller
- 1:27:06Speaker 1: terrestrial planets. A disk with different embryo distributions would produce
- 1:27:10Speaker 1: different final sizes. The physics of accretion does not impose
- 1:27:14Speaker 1: a sharp lower cut-off on how small a rocky planet
- 1:27:17Speaker 1: can be. It imposes a probability gradient. Smaller planets are
- 1:27:21Speaker 1: easier to build because they require less material and fewer
- 1:27:24Speaker 1: giant impacts. Larger planets are harder to build because they
- 1:27:28Speaker 1: demand more mass converging in the same orbital zone. This
- 1:27:32Speaker 1: means moon-sized rocky planets are not only possible, they may
- 1:27:35Speaker 1: be common. A body the size of Kepler-37b, roughly 3,900
- 1:27:38Speaker 1: kilometres across, could form from a single large embryo that
- 1:27:45Speaker 1: never experienced the giant impacts necessary to grow further. Perhaps
- 1:27:50Speaker 1: it formed in a region of the disk where solid
- 1:27:52Speaker 1: material was sparse. Perhaps it was gravitationally scattered inward by
- 1:27:56Speaker 1: interactions with larger embryos and stranded in a close-in orbit
- 1:28:00Speaker 1: before it could accumulate more mass. Perhaps the disk dispersed
- 1:28:03Speaker 1: before accretion could continue. Any of these pathways could produce
- 1:28:07Speaker 1: a body in the size range of Kepler-37b, and none
- 1:28:11Speaker 1: of them requires exotic physics or unusual initial conditions. The
- 1:28:16Speaker 1: distinction between pebble accretion and planetesimal accretion is relevant here
- 1:28:20Speaker 1: because the two models predict different size distributions for the
- 1:28:23Speaker 1: smallest rocky planets. In the classical planetesimal accretion model, planets
- 1:28:28Speaker 1: grow primarily through collisions between kilometer-scale bodies. This process is
- 1:28:33Speaker 1: relatively slow and inefficient in the inner disk, where orbital
- 1:28:36Speaker 1: velocities are high and the available mass is limited. The
- 1:28:40Speaker 1: result tends to be a small number of moderate-sized embryos
- 1:28:43Speaker 1: that undergo a final phase of giant impacts to produce
- 1:28:46Speaker 1: the finished planets. This model naturally explains the terrestrial planets
- 1:28:50Speaker 1: of our solar system, but tends to produce planets clustered
- 1:28:53Speaker 1: around a characteristic size set by the local mass budget.
- 1:28:57Speaker 1: Pebble accretion offers an alternative pathway. In this model, a
- 1:29:02Speaker 1: growing embryo sweeps up centimeter to meter scale pebbles that
- 1:29:06Speaker 1: drift inward through the disc due to aerodynamic drag from
- 1:29:09Speaker 1: the gas. Because pebbles are continuously supplied by the outer disc,
- 1:29:13Speaker 1: a well-positioned embryo can grow rapidly, potentially reaching Earth mass
- 1:29:17Speaker 1: in less than a million years. But the efficiency of
- 1:29:20Speaker 1: pebble accretion depends sensitively on the properties of the gas disc,
- 1:29:24Speaker 1: particularly its turbulent viscosity. In low viscosity discs, Pebble accretion
- 1:29:29Speaker 1: is efficient and produces relatively massive planets. In high viscosity disks,
- 1:29:34Speaker 1: pebble accretion is suppressed and embryos stall at much smaller
- 1:29:38Speaker 1: masses potentially in the sub-Earth or sub-Mercury range. Computer simulations
- 1:29:43Speaker 1: of pebble accretion in high viscosity disks have shown that
- 1:29:46Speaker 1: no planet more massive than roughly one half an Earth
- 1:29:49Speaker 1: mass forms under those conditions. The result is a population
- 1:29:53Speaker 1: of small rocky worlds. exactly in the size range occupied
- 1:29:57Speaker 1: by Kepler-37b. If high-viscosity disks are common, submercury planets may
- 1:30:02Speaker 1: be not just possible but a routine outcome of planet formation.
- 1:30:06Speaker 1: They require only a protoplanetary disk that did not deliver
- 1:30:10Speaker 1: enough material to that particular orbital zone to build anything bigger.
- 1:30:14Speaker 1: The Kepler-37 system itself offers circumstantial evidence for this scenario.
- 1:30:19Speaker 1: The three known planets increase in size outward, the innermost
- 1:30:22Speaker 1: is moon-sized, The middle planet is Venus-sized, and the outermost
- 1:30:27Speaker 1: is roughly twice Earth-size. This architecture is consistent with a
- 1:30:31Speaker 1: formation history in which the inner disk was mass-poor and
- 1:30:34Speaker 1: the outer disk was mass-rich, a gradient that would naturally
- 1:30:38Speaker 1: produce smaller planets closer in and larger planets further out.
- 1:30:42Speaker 1: The near-resonant orbital period ratios suggest the planets migrated inward
- 1:30:46Speaker 1: through the disk and locked into their current configuration. Kepler-37b
- 1:30:52Speaker 1: may have been the runt of the litter, The embryo
- 1:30:55Speaker 1: that grew in the most material starved zone completed its
- 1:30:58Speaker 1: formation early and settled into a tight orbit where it
- 1:31:02Speaker 1: has remained for billions of years. But formation models tell
- 1:31:06Speaker 1: us something broader than the history of one system. They
- 1:31:09Speaker 1: tell us that the accretion process has no built-in prohibition
- 1:31:12Speaker 1: against producing bodies at this scale. The lower boundary of
- 1:31:15Speaker 1: planet formation is not set by physics. It is set
- 1:31:19Speaker 1: by how we define the word planet. and by whether
- 1:31:22Speaker 1: the body ends up orbiting a star or orbiting another planet.
- 1:31:26Speaker 1: A moon-sized object orbiting Jupiter is a moon. The same
- 1:31:30Speaker 1: object orbiting a sun-like star is a planet. The physics
- 1:31:34Speaker 1: of its formation may be identical. Only the orbital context changes.
- 1:31:39Speaker 1: This realisation leads to an unsettling inference about what the
- 1:31:42Speaker 1: galaxy actually contains. If the accretion process readily produces bodies
- 1:31:47Speaker 1: in the sub-Mercury size range, and if the detection bias
- 1:31:50Speaker 1: analyses discussed earlier are even approximately correct, then the Milky
- 1:31:54Speaker 1: Way is teeming with tiny rocky planets we have never seen.
- 1:31:58Speaker 1: The numbers, even conservatively estimated, are staggering. Kepler occurrence rate
- 1:32:03Speaker 1: studies suggest that roughly 20-30% of sun-like stars host at
- 1:32:08Speaker 1: least one planet with a radius between 0.5 and 1
- 1:32:12Speaker 1: Earth radius and an orbital period shorter than about one year.
- 1:32:16Speaker 1: Below 0.5 Earth radii, the estimates become far less certain
- 1:32:20Speaker 1: because Kepler's sensitivity drops sharply. But extrapolations based on the
- 1:32:25Speaker 1: trend of increasing occurrence with decreasing size suggest that sub-Mercury worlds,
- 1:32:30Speaker 1: planets with radii below 0.4 Earth radii, may orbit a
- 1:32:35Speaker 1: substantial fraction of all, Sun-like stars. If even 10% of
- 1:32:40Speaker 1: the roughly 10 to 20 billion Sun-like stars in the
- 1:32:43Speaker 1: Milky Way host one sub-Mercury planet, That is one to
- 1:32:47Speaker 1: two billion tiny rocky worlds around sun-like stars alone. Extend
- 1:32:51Speaker 1: the estimate to include the far more numerous red dwarf stars,
- 1:32:55Speaker 1: which also host rocky planets, and the count rises to
- 1:32:59Speaker 1: tens of billions or more. If the true fraction is higher,
- 1:33:03Speaker 1: as some models suggest, the number could reach tens of
- 1:33:06Speaker 1: billions or more. These are worlds we have almost no
- 1:33:09Speaker 1: hope of detecting with current technology. Their transits produce signals
- 1:33:13Speaker 1: of a few parts per million or less against sun-like stars.
- 1:33:16Speaker 1: Their radial velocity signatures are measured in fractions of a
- 1:33:19Speaker 1: centimeter per second. They are gravitationally insignificant on any scale
- 1:33:24Speaker 1: we can presently measure from interstellar distances. They orbit in silence,
- 1:33:29Speaker 1: carrying no electromagnetic signature that distinguishes them from the emptiness
- 1:33:33Speaker 1: between stars. They exist in all probability, but they are invisible.
- 1:33:38Speaker 1: The question of how to detect them drives the design
- 1:33:40Speaker 1: of future missions. The European Space Agency's PLATO mission currently
- 1:33:45Speaker 1: expected to launch in the late 2020s, will conduct long-duration,
- 1:33:49Speaker 1: high-precision transit photometry of bright stars across a wide field
- 1:33:53Speaker 1: of view. Its multi-telescope array is designed to achieve the
- 1:33:57Speaker 1: kind of sustained photometric performance that Kepler demonstrated was necessary
- 1:34:01Speaker 1: for sub-Earth detections, but applied to stars bright enough for
- 1:34:05Speaker 1: ground-based follow-up. Plato's primary goal is to find Earth-sized planets
- 1:34:10Speaker 1: in habitable zone orbits around Sun-like stars. but its sensitivity
- 1:34:14Speaker 1: will extend to smaller worlds on shorter orbits, potentially pushing
- 1:34:18Speaker 1: the confirmed radius record below Kepler-37b. NASA's Nancy Grace Roman
- 1:34:23Speaker 1: Space Telescope, also expected to launch in the coming years,
- 1:34:27Speaker 1: will bring an entirely different approach. One of its survey
- 1:34:31Speaker 1: programs will use gravitational microlensing to detect planets across a
- 1:34:34Speaker 1: wide range of masses and orbital distances. Microlensing occurs when
- 1:34:38Speaker 1: a foreground star and its planetary system pass between Earth
- 1:34:42Speaker 1: and a more distant background star. The gravity of the
- 1:34:45Speaker 1: foreground system acts as a lens, briefly magnifying the background
- 1:34:49Speaker 1: star's light. If a planet orbits the foreground star, it
- 1:34:53Speaker 1: creates a characteristic secondary brightening in the light curve. Crucially,
- 1:34:57Speaker 1: microlensing sensitivity does not depend on the planet producing a
- 1:35:01Speaker 1: transit or a radial velocity wobble. It depends on the
- 1:35:04Speaker 1: planet's mass and its projected separation from the host star
- 1:35:08Speaker 1: at the moment of alignment. Roman's microlensing survey is expected
- 1:35:12Speaker 1: to detect roughly 2,600 planets, including worlds as light as
- 1:35:16Speaker 1: a few tenths of an Earth, mass lighter than Mars,
- 1:35:20Speaker 1: potentially comparable to Kepler-37b. The geometry of microlensing makes this possible.
- 1:35:26Speaker 1: When a foreground star passes almost directly between Earth and
- 1:35:29Speaker 1: a distant background star, the foreground star's gravity bends and
- 1:35:33Speaker 1: magnifies the background star's light, producing a characteristic brightening event
- 1:35:37Speaker 1: that typically lasts several weeks. If a planet orbits the
- 1:35:41Speaker 1: foreground star, its own gravity creates a brief secondary distortion
- 1:35:45Speaker 1: in the light curve, a sharp spike or dip lasting
- 1:35:47Speaker 1: hours to days, depending on the planet's mass and its
- 1:35:51Speaker 1: projected separation from the host star at the moment of alignment.
- 1:35:54Speaker 1: The key advantage of microlensing is that it responds to
- 1:35:58Speaker 1: mass directly. not to the planet's luminosity or its effect
- 1:36:02Speaker 1: on the star's radial velocity. A planet that is completely dark,
- 1:36:06Speaker 1: atmosphereless and gravitationally negligible from interstellar distances still bends spacetime,
- 1:36:13Speaker 1: and that bending is detectable if the alignment is favourable.
- 1:36:16Speaker 1: Roman will monitor roughly 200 million stars in the galactic
- 1:36:20Speaker 1: bulge over multiple observing seasons, watching for microlensing events and
- 1:36:25Speaker 1: scanning the resulting light curves for planetary signatures. Because the
- 1:36:29Speaker 1: survey covers such an enormous number of source stars, it
- 1:36:32Speaker 1: will detect planets across a statistical cross-section of the galaxy's
- 1:36:36Speaker 1: stellar population, not just the bright, nearby stars that transit
- 1:36:40Speaker 1: and radial velocity surveys are limited to. Crucially, microlensing sensitivity
- 1:36:45Speaker 1: peaks at orbital separations of roughly 1 to 10 astronomical
- 1:36:48Speaker 1: units and extends to sub-Earth masses, covering precisely the region
- 1:36:52Speaker 1: of parameter space where transit surveys are weakest. The two
- 1:36:56Speaker 1: methods are complementary in the deepest sense. What one misses,
- 1:37:00Speaker 1: the other finds. For the first time, we may be
- 1:37:03Speaker 1: able to measure the mass function of sub-Earth planets directly,
- 1:37:07Speaker 1: filling in the part of the distribution that transit surveys
- 1:37:10Speaker 1: cannot reach. But even these next generation missions will not
- 1:37:13Speaker 1: find everything. The smallest rocky bodies objects in the size
- 1:37:16Speaker 1: range of Ceres, Vesta, or Pallas will remain below the
- 1:37:21Speaker 1: detection floor of any foreseeable instrument when they orbit distant stars.
- 1:37:25Speaker 1: The transit depths they would produce are measured in fractions
- 1:37:28Speaker 1: of a part per million. The microlensing signals they would
- 1:37:31Speaker 1: generate are buried in noise. These objects may be extraordinarily common.
- 1:37:36Speaker 1: The asteroid belt in our solar system contains millions of
- 1:37:39Speaker 1: rocky bodies, ranging from a few metres to nearly a
- 1:37:42Speaker 1: thousand kilometres across. If similar populations orbit other stars, and
- 1:37:47Speaker 1: there is no reason to think they do not, then
- 1:37:50Speaker 1: the number of rocky bodies in the galaxy exceeds any
- 1:37:52Speaker 1: count we are likely to compile. The word planet may
- 1:37:55Speaker 1: not apply to most of them, but the continuum they
- 1:37:58Speaker 1: inhabit extends without interruption from dust to super-Earths, and the
- 1:38:02Speaker 1: boundaries we draw within that continuum are conventions, not laws.
- 1:38:06Speaker 1: This is where the story of Kepler-37b intersects with something
- 1:38:10Speaker 1: deeper than planetary science. It intersects with the limits of
- 1:38:14Speaker 1: human categorisation itself. We name things in order to understand them.
- 1:38:19Speaker 1: The word planet is a container, a mental box into
- 1:38:23Speaker 1: which we place certain objects and from which we exclude others.
- 1:38:27Speaker 1: The box has been resized repeatedly throughout history. For the
- 1:38:31Speaker 1: ancient Greeks, planets were wandering lights in the sky, a
- 1:38:34Speaker 1: category that included Mercury, Venus, Mars, Jupiter, Saturn, the Sun
- 1:38:41Speaker 1: and the Moon. For post-Copernican astronomers, planets became worlds orbiting
- 1:38:45Speaker 1: the Sun and Earth joined the list while the Sun
- 1:38:48Speaker 1: and Moon were removed. For two centuries, Ceres was considered
- 1:38:52Speaker 1: a planet after its discovery in 1801, until the growing
- 1:38:56Speaker 1: population of similar objects in the asteroid belt forced a reclassification.
- 1:39:00Speaker 1: Pluto was a planet for 76 years, before the International
- 1:39:04Speaker 1: Astronomical Union redrew the boundary in 2006. At every stage,
- 1:39:09Speaker 1: the word planet has been stretched, contracted, redefined, and contested
- 1:39:15Speaker 1: not because the objects changed, but because our knowledge of
- 1:39:18Speaker 1: them deepened. and the old categories no longer fit, Kepler-37b
- 1:39:24Speaker 1: applies a new kind of pressure to this category. It
- 1:39:27Speaker 1: is not a borderline case in the way Pluto was
- 1:39:29Speaker 1: an object near a definitional threshold that could fall on
- 1:39:32Speaker 1: either side depending on the criteria. It is a confirmed
- 1:39:35Speaker 1: planet by every current standard. It orbits a star. It
- 1:39:40Speaker 1: is below the deuterium burning mass limit. It formed through
- 1:39:44Speaker 1: accretion in a protoplanetary disk. Nothing in the current definitions
- 1:39:48Speaker 1: excludes it and yet it occupies a size regime that
- 1:39:52Speaker 1: the word planet was never designed to describe. It is
- 1:39:55Speaker 1: a world smaller than every planet in our solar system,
- 1:39:59Speaker 1: smaller than both of the moons that exceed Mercury in size.
- 1:40:03Speaker 1: It is a planet that feels intuitively like it should
- 1:40:06Speaker 1: be something else, an asteroid, a captured moon, a leftover
- 1:40:10Speaker 1: fragment of formation rather than a product of it. This
- 1:40:13Speaker 1: intuitive resistance is revealing. It tells us that our mental
- 1:40:17Speaker 1: model of what a planet is carries assumptions about scale
- 1:40:21Speaker 1: and significance that are not part of any formal definition.
- 1:40:24Speaker 1: When we hear the word planet, we imagine a world
- 1:40:27Speaker 1: with a horizon, a sky, a landscape, something we could,
- 1:40:32Speaker 1: in principle, stand on and look around. Kepler-37b technically satisfies
- 1:40:37Speaker 1: this criterion. It has a surface, it has gravity, assuming
- 1:40:42Speaker 1: a rocky composition similar to the moon, roughly 15-20% of Earth's.
- 1:40:45Speaker 1: If you could stand on it, You would have a horizon,
- 1:40:50Speaker 1: though it would curve sharply away beneath you because the
- 1:40:52Speaker 1: world is so small. You would weigh roughly 12 to
- 1:40:55Speaker 1: 15 kilograms if you weigh 80 on Earth. You could
- 1:40:59Speaker 1: throw a baseball and watch it arc lazily for more
- 1:41:02Speaker 1: than a kilometre before touching down. It is a world
- 1:41:05Speaker 1: you could walk across in a matter of weeks. The
- 1:41:08Speaker 1: entire equator spans roughly 12,200 kilometres less than the distance
- 1:41:13Speaker 1: from New York to Tokyo and back. It is a
- 1:41:15Speaker 1: world you could circumnavigate on foot in roughly a year
- 1:41:18Speaker 1: without stopping, assuming you survived the 400 degree surface temperature
- 1:41:23Speaker 1: and the total absence of air. The point is not
- 1:41:26Speaker 1: that Kepler-37b is hospitable, it is catastrophically hostile. The point
- 1:41:32Speaker 1: is that even at this scale, the object retains some
- 1:41:35Speaker 1: recognisable features of a world. It has a surface shaped
- 1:41:38Speaker 1: by thermal stress and possibly ancient impacts. It has an
- 1:41:42Speaker 1: interior differentiated by density, heavier materials sinking. Lighter materials rising.
- 1:41:47Speaker 1: The same process that gave Earth an iron core and
- 1:41:50Speaker 1: a silicate mantle. It is a product of the same
- 1:41:53Speaker 1: physics that built the world beneath your feet. It is
- 1:41:56Speaker 1: just smaller. Much smaller. Small enough that calling it a
- 1:42:00Speaker 1: planet requires you to accept that planet has no lower
- 1:42:04Speaker 1: boundary beyond the ability to orbit a star and be
- 1:42:07Speaker 1: too small to fuse hydrogen. And this is the realisation
- 1:42:10Speaker 1: that makes Kepler-37b genuinely unsettling. It is not unsettling because
- 1:42:15Speaker 1: it is one unusual world. It is unsettling because it
- 1:42:19Speaker 1: reveals that the category planet extends downward into a realm
- 1:42:24Speaker 1: where our instruments lose their grip and our intuitions lose
- 1:42:27Speaker 1: their footing. If a world 3,900 kilometers across counts as
- 1:42:29Speaker 1: a planet, what about a world 3,000 kilometers across? 2,000? 1,000?
- 1:42:31Speaker 1: Where does planet end and debris begin? There is no
- 1:42:34Speaker 1: physical threshold. There is no moment where an accreting body
- 1:42:46Speaker 1: crosses a line and becomes a planet. There is only
- 1:42:48Speaker 1: a gradient of size and complexity, and the word planet
- 1:42:53Speaker 1: is a label we have pasted over one segment of
- 1:42:55Speaker 1: that gradient with boundaries that reflect our detection capabilities more
- 1:42:59Speaker 1: than any natural distinction. The universe does not organise itself
- 1:43:03Speaker 1: into the categories we prefer. It builds rocky bodies at
- 1:43:06Speaker 1: every scale, from micron-sized dust grains to super-Earths ten times
- 1:43:11Speaker 1: the mass of our world. and it distributes them across
- 1:43:14Speaker 1: an enormous range of orbital configurations. Some of these bodies
- 1:43:18Speaker 1: we call planets, some we call moons, some we call
- 1:43:21Speaker 1: dwarf planets, some we call asteroids. The distinctions among these
- 1:43:25Speaker 1: categories are real, in the sense that they describe different
- 1:43:28Speaker 1: formation pathways and orbital contexts, but they are not as
- 1:43:32Speaker 1: sharp as our vocabulary implies. At the boundaries, and Kepler-37b
- 1:43:38Speaker 1: sits squarely at a boundary, the labels start to blur.
- 1:43:41Speaker 1: The Kepler Space Telescope watched one fraction of 1% of
- 1:43:45Speaker 1: the sky for four years and found one planet this small. One.
- 1:43:50Speaker 1: Out of the billions of stars it could not observe,
- 1:43:52Speaker 1: the trillions of orbital alignments it could not catch, the
- 1:43:56Speaker 1: infinite gradations of transit depth it could not measure, Kepler-37b
- 1:44:02Speaker 1: is not a statistical outlier at the extreme end of
- 1:44:05Speaker 1: a well-sampled distribution. It is a single data point pulled
- 1:44:09Speaker 1: from the edge of an enormous dark region In our knowledge,
- 1:44:13Speaker 1: a region where an entire population of tiny worlds presumably exists,
- 1:44:17Speaker 1: uncounted and uncharacterised, forming and orbiting and evolving in a
- 1:44:22Speaker 1: silence our instruments cannot yet penetrate. What would it mean
- 1:44:26Speaker 1: for our understanding of the universe if we could suddenly
- 1:44:28Speaker 1: see all of them? If every sub-mercury planet in the
- 1:44:32Speaker 1: galaxy lit up on a map, the sky would not
- 1:44:34Speaker 1: look the way we imagine it. We picture a cosmos
- 1:44:37Speaker 1: of gas giants and rocky earths, of habitable zones and
- 1:44:41Speaker 1: warm Jupiters, because those are the worlds our surveys have found.
- 1:44:45Speaker 1: The true picture would be dominated by something else entirely,
- 1:44:49Speaker 1: an overwhelming population of small, scorched, airless rocks orbiting their
- 1:44:53Speaker 1: stars in tight, fast orbits, producing transit signals measured in
- 1:44:58Speaker 1: single-digit parts per million, carrying masses too small to tug
- 1:45:02Speaker 1: their stars by any measurable amount. These worlds would not
- 1:45:05Speaker 1: be habitable. They would not be interesting in the way
- 1:45:08Speaker 1: that Earth-like planets are interesting. They would not make headlines.
- 1:45:12Speaker 1: But they would be the most common type of planet
- 1:45:14Speaker 1: in the galaxy. And they would demonstrate, with a clarity
- 1:45:18Speaker 1: that no single detection can provide, that the universe builds
- 1:45:21Speaker 1: worlds at every scale it can, not just the scales
- 1:45:25Speaker 1: we have learned to detect. Kepler-37b is the first confirmed
- 1:45:29Speaker 1: member of this population to have a name. It is
- 1:45:31Speaker 1: the one that slipped above the noise floor because everything aligned.
- 1:45:35Speaker 1: The right star, the right instrument, the right orbital period,
- 1:45:39Speaker 1: the right photometric stability. It sits in our catalogues as
- 1:45:43Speaker 1: a record holder, the smallest confirmed exoplanet by radius orbiting
- 1:45:48Speaker 1: a main sequence star. But the record it holds is
- 1:45:51Speaker 1: not a statement about nature. It is a statement about us,
- 1:45:55Speaker 1: about our instruments, our methods, our ability to extract impossibly
- 1:45:59Speaker 1: faint signals from the light of distant stars. Nature did
- 1:46:03Speaker 1: not set the lower boundary of the planetary catalogue at 0.31
- 1:46:06Speaker 1: Earth radii. Kepler's noise floor did. Below that floor, the
- 1:46:11Speaker 1: universe continues. Rocky bodies smaller than the moon orbit stars
- 1:46:15Speaker 1: we have observed without ever knowing they were there. Worlds
- 1:46:19Speaker 1: we would call planets if we could detect them. Drift-through
- 1:46:23Speaker 1: systems we have classified as planet-free, based on null results
- 1:46:27Speaker 1: that mean only that our instruments found nothing, not that
- 1:46:29Speaker 1: nothing exists. The absence of evidence is not evidence of absence,
- 1:46:34Speaker 1: and at the smallest planetary scales, the absence of evidence
- 1:46:37Speaker 1: is almost entirely an artifact of insufficient sensitivity. This is
- 1:46:42Speaker 1: the truth that Kepler-37b carries, not in its scorched surface
- 1:46:46Speaker 1: or its airless sky, but in its position at the
- 1:46:49Speaker 1: boundary of the detectable. It is a messenger from the
- 1:46:52Speaker 1: edge of the invisible. It tells us that the word
- 1:46:54Speaker 1: planet does not end where our telescopes stop seeing. It
- 1:46:58Speaker 1: tells us that the population of worlds in the galaxy
- 1:47:00Speaker 1: is vaster, more varied, and more weighted toward the small
- 1:47:04Speaker 1: than any catalogue we have ever compiled. It tells us
- 1:47:07Speaker 1: that the universe's inventory of rocky bodies extends downward in
- 1:47:11Speaker 1: size far beyond the point where our instruments can follow,
- 1:47:14Speaker 1: and that every survey we conduct, no matter how sensitive,
- 1:47:18Speaker 1: reveals only the upper portion of a distribution whose true
- 1:47:21Speaker 1: extent we can only estimate. The smallest planet we have
- 1:47:24Speaker 1: ever found is not the smallest planet that exists. It
- 1:47:28Speaker 1: is merely the smallest planet we have managed to catch
- 1:47:31Speaker 1: in the act of crossing a star. Somewhere in the galaxy,
- 1:47:34Speaker 1: right now, a world smaller than Kepler-37b is orbiting a
- 1:47:39Speaker 1: star no one is watching, producing a transit signal no
- 1:47:43Speaker 1: instrument can detect, existing without a name or a catalogue
- 1:47:46Speaker 1: number or a place in any human database. It is
- 1:47:49Speaker 1: a planet by every definition we have written. It is
- 1:47:52Speaker 1: also completely invisible, and it is almost certainly not alone.
- 1:47:56Speaker 1: The terrifying part is not that one tiny planet pushes
- 1:47:59Speaker 1: our categories to the breaking point. The terrifying part is
- 1:48:02Speaker 1: that the universe may be filled with worlds at scales
- 1:48:06Speaker 1: our intuition rejects, and our instruments cannot confirm. Worlds that
- 1:48:10Speaker 1: sit in the space between planet and almost too small
- 1:48:13Speaker 1: to count. Worlds that form the same way Earth formed,
- 1:48:17Speaker 1: from the same physics, the same dust, the same gravity,
- 1:48:19Speaker 1: only less of it. Worlds that are real and numerous
- 1:48:22Speaker 1: and permanent, orbiting patiently in the dark while we argue
- 1:48:26Speaker 1: about what to call them. Kepler, 37b, is the doorway.
- 1:48:31Speaker 1: What lies beyond it is not another record. It is
- 1:48:34Speaker 1: the recognition that even the word planet, the oldest, most
- 1:48:38Speaker 1: familiar term in astronomy, becomes fragile when you press it
- 1:48:42Speaker 1: against the smallest things the universe can build.