by DC Catling · 2009 · Cited by 64 — The Planetary Air Leak. As Earth’s atmosphere slowly trickles away into space, will our planet come to look like Venus? By David C. Catling and Kevin J. Zahnle

153 KB – 8 Pages

PAGE – 1 ============
36 SCIENTIFIC AMERIC AN May 2009 ALFRED T. K AMAJI ANOne of the most remarkable features of the solar system is the variety of planetary atmospheres. Earth and Venus are of comparable size and mass, yet the surface of Ve -nus bakes at 460 degrees Celsius under an ocean of carbon dioxide that bears down with the weight of a kilometer of water. Callisto and Ti -tan Šplanet-size moons of Jupiter and Saturn, respectively Šare nearly the same size, yet Titan has a nitrogen-rich atmosphere thicker than our own, whereas Callisto is essentially airless. What causes such extremes? If we knew, it would help explain why Earth teems with life while its planetary siblings appear to be dead. Knowing how atmospheres evolve is also essential to de -termining which planets beyond our solar sys -tem might be habitable. A planet can acquire a gaseous cloak in many ways: it can release vapors from its interior, it can capture volatile materials from comets and aster -oids when they strike, and its gravity can pull in gases from interplanetary space. But planetary scientists have begun to appreciate that the es -cape of gases plays as big a role as the supply. Al -though Earth™s atmosphere may seem as perma -nent as the rocks, it gradually leaks back into space. The loss rate is currently tiny, only about three kilograms of hydrogen and 50 grams of he -lium (the two lightest gases) per second, but even that trickle can be signi˜cant over geologic time, and the rate was probably once much higher. As Benjamin Franklin wrote, fiA small leak can sink a great ship.fl The atmospheres of terrestrial planets and outer-planet satellites we see today are like the ruins of medieval castles Šremnants of riches that have been subject to histories of plunder and decay. The atmospheres of smaller bodies are more like crude forts, poorly defended and extremely vulnerable. Recognizing the importance of atmospheric escape changes our perspective on the solar sys -tem. For decades, scientists have pondered why Mars has such a thin atmosphere, but now we wonder: Why does it have any atmosphere left at all? Is the difference between Titan and Callisto a consequence of Callisto™s losing its atmosphere, rather than of Titan having been born of airier stuff? Was Titan™s atmosphere once even thicker than it is today? How did Venus steadfastly cling to its nitrogen and carbon dioxide yet thorough -ly lose its water? Did escape of hydrogen help to set the stage for complex life on Earth? Will it one day turn our planet into another Venus? When the Heat Is On A spaceship that reaches escape velocity is mov -ing fast enough to break free of a planet™s grav -ity. The same is true of atoms and molecules, although they usually reach escape velocity less purposefully. In thermal escape, gases get too hot to hold on to. In nonthermal processes, chemical or charged-particle reactions hurl out atoms and molecules. And in a third process, asteroid and comet impacts blast away the air. Thermal escape is, in some ways, the most common and straightforward of the three. All bodies in the solar system are heated by sunlight. They rid themselves of this heat in two ways: by emitting infrared radiation and by shedding mat -ter. In long-lived bodies such as Earth, the former KEY CONCE PTSMany of the gases that ˜make up Earth™s atmo -sphere and those of the other planets are slowly leaking into space. Hot gases, especially light ones, evaporate away; chemical reactions and particle collisions eject atoms and molecules; and asteroids and comets occasionally blast out chunks of atmosphere. This leakage explains ˜many of the solar system™s mysteries. For instance, Mars is red because its water vapor got broken down into hydrogen and oxygen, the hydrogen drift -ed away, and the surplus oxygen oxidized Šin es -sence, rusted Šthe rocks. A similar process on Venus let carbon dioxide build up into a thick ocean of air; ironically, Venus™s huge atmosphere is the result of the loss of gases. ŠThe Editors PLANETARY SCIENCE The Planetary Air Leak As Earth™s atmosphere slowly trickles away into space, will our planet come to look like Venus? By David C. Catling and Kevin J. Zahnle

PAGE – 2 ============ EARTH PAST: 3 BILLION YEARS AGO EARTH FU TURE: 3 BILLION YEARS F RO M NO WEARTH PRESENT LOSS OF CERTAIN GASES , especially hydrogen, has transformed Earth. It is one of the reasons that oxygen built up in the atmosphere. In the future, the depletion of hydrogen will dry out our oceans and all but shut down geologic cycles that stabilize the climate. Life may still be able to hold out in the polar regions.

PAGE – 3 ============
38 SCIENTIFIC AMERIC AN May 2009 ˜rst, called Jeans escape, after James Jeans, the English astronomer who described it in the early 20th century, air literally evaporates atom by atom, molecule by molecule, off the top of the at -mosphere. At lower altitudes, collisions con˜ne particles, but above a certain altitude, known as the exobase, which on Earth is about 500 kilo -meters above the surface, air is so tenuous that gas particles hardly ever collide. Nothing stops an atom or molecule with suf˜cient velocity from ˚ying away into space. As the lightest gas, hydrogen is the one that most easily overcomes a planet™s gravity. But ˜rst it must reach the exobase, and on Earth that is a slow process. Hydrogen-bearing molecules tend not to rise above the lowest layer of atmosphere: water vapor (H 2O) condenses out and rains back down, and methane (CH 4) is oxidized to form carbon dioxide (CO 2). Some water and methane molecules reach the stratosphere and decompose, releasing hydrogen, which slowly diffuses up -ward until it reaches the exobase. A small amount clearly makes it out because ultraviolet images reveal a halo of hydrogen atoms surrounding our planet [ see illustration on opposite page ].The temperature at Earth™s exobase oscillates but is typically about 1,000 kelvins, implying that hydrogen atoms have an average speed of ˜ve kilometers per second. That is less than Earth™s escape velocity at that altitude, 10.8 ki -lometers per second, but the average conceals a wide range, so some hydrogen atoms still man -age to break free of our planet™s gravity. This loss of particles from the energetic tail of the speed distribution explains about 10 to 40 percent of Earth™s hydrogen loss today. Jeans escape also partly explains why our moon is airless. Gases GEOR GE RETSEC K (illustration ); ALFRED T. K AMAJI AN ( graph )process prevails; for others, such as comets, the latter dominates. Even a body the size of Earth can heat up quickly if absorption and radiation get out of balance, and its atmosphere Šwhich typically has very little mass compared with the rest of the planet Šcan slough off in a cosmic in -stant. Our solar system is littered with airless bodies, and thermal escape seems to be a com -mon culprit. Airless bodies stand out as those where solar heating exceeds a certain threshold, which depends on the strength of the body™s gravity [ see illustration above ].Thermal escape occurs in two ways. In the One major cause of air loss is solar heating. Heat can drive out air in one of two ways. Exosphere [ESCAPE M ECHANIS M #1] PLANET ARY TE AKETTLE AIR EVAPORATES MOLEC ULE BY MOLEC ULE In an atmosphere™s uppermost layer, or exosphere, nothing stops the fastest-moving atoms and molecules from ˜ying off into space. This process, known as Jeans escape, accounts for much of the leakage of hydrogen from our planet. HEATED AIR FLO WS OUT IN A WIND Air heated by sunlight rises, accelerates and attains escape velocity. This process, known as hydrodynamic escape, was particularly impor -tant on early Earth and VenusŠin fact, it may be why Venus became what it is today. EVIDENCE F OR THER MAL ESCAPE comes from considering which planets and satellites have atmospheres and which do not. The deciding factor appears to be the strength of stellar heat -ing ( vertical axis ) relative to the strength of a body™s gravity (horizontal axis ). Airless worlds have strong heating and weak gravity ( left of line ). Bodies with atmospheres have weak heating and strong gravity (right of line ).Molecule Stellar HeatingAirless bodiesBodies with atmospheres Asteroids Jupiter SaturnUranus NeptuneEarthVenus Mars TitanMercury MoonTriton PlutoIoCallistoGanymedeEuropa Extrasolar planetsLapetusCharon Strength of Gravity WHICH BODIES HAVE LOST ALL THEIR AIR ?

PAGE – 4 ============ SCIENTIFIC A MERIC AN 39NASA/UNIVERSI TY OF IOW Areleased from the lunar surface easily evaporate off into space. A second type of thermal escape is far more dramatic. Whereas Jeans escape occurs when a gas evaporates molecule by molecule, heated air can also ˚ow en masse. The upper atmosphere can absorb ultraviolet sunlight, warm up and ex -pand, pushing air upward. As the air rises, it ac -celerates smoothly through the speed of sound and then attains the escape velocity. This form of thermal escape is called hydrodynamic escape or, more evocatively, the planetary wind Šthe latter by analogy to the solar wind, the stream of charged particles blown from the sun into inter -planetary space. Dust in the Wind Atmospheres rich with hydrogen are the most vulnerable to hydrodynamic escape. As hydro -gen ˚ows outward, it can pick up and drag along heavier molecules and atoms with it. Much as the desert wind blows dust across an ocean and sand grains from dune to dune, while leaving cobbles and boulders behind, the hydrogen wind carries off molecules and atoms at a rate that diminishes with their weight. Thus, the present composition of an atmosphere can reveal wheth -er this process has ever occurred. In fact, astronomers have seen the telltale signs of hydrodynamic escape outside the solar system, on the Jupiter-like planet HD 209458b. Using the Hubble Space Telescope, Alfred Vidal- Madjar of the Paris Astrophysics Institute and his colleagues reported in 2003 that the planet has a puffed-up atmosphere of hydrogen. Subse -quent measurements discovered carbon and ox -ygen in this in˚ated atmosphere. These atoms are too heavy to escape on their own, so they must have been dragged there by hydrogen. Hy -drodynamic loss would also explain why astron -omers ˜nd no large planets much closer to their stars than HD 209458b is. For planets that orbit within three million kilometers or so of their stars (about half the orbital radius of HD 209458b), hydrodynamic escape strips away the entire atmosphere within a few billion years, leaving behind only a scorched remnant. This evidence for planetary winds lends cre -dence to ideas put forth in the 1980s about hy -drodynamic escape from ancient Venus, Earth and Mars. Three clues suggest this process once operated on these worlds. The ˜rst concerns no -ble gases. Were it not for escape, chemically un -reactive gases such as neon or argon would re -main in an atmosphere inde˜nitely. The abun -dances of their different isotopes would be similar to their original values, which in turn are similar to that of the sun, given their common origin in the solar nebula. Yet the abundances differ. Second, youthful stars are strong sources of ultraviolet light, and our sun was probably no exception. This radiation could have driven hy -drodynamic escape. Third, the early terrestrial planets may have had hydrogen-rich atmospheres. The hydrogen could have come from chemical reactions of wa -ter with iron, from nebular gases or from water molecules broken apart by solar ultraviolet radi -ation. In those primeval days, asteroids and com -ets hit more frequently, and whenever they smacked into an ocean, they ˜lled the atmosphere with steam. Over thousands of years the steam condensed and rained back onto the surface, but Venus is close enough to the sun that water vapor may have persisted in the atmosphere, where so -lar radiation could break it down. Under such conditions, hydrodynamic escape would readily operate. In the 1980s James F. Kasting, now at Pennsylvania State University, showed that hydrodynamic escape on Venus could have carried away an ocean™s worth of hy -drogen within a few tens of millions of years [see fiHow Climate Evolved on the Terrestrial Plan -ets,fl by James F. Kasting, Owen B. Toon and James B. Pollack; S˛˝˙ˆˇ˝˘˝˛ A˙˝˛ˆ, Feb -ruary 1988]. Kasting and one of us (Zahnle) sub -sequently showed that escaping hydrogen would have dragged along much of the oxygen but left carbon dioxide behind. Without water to medi -ate the chemical reactions that turn carbon di -oxide into carbonate minerals such as limestone, the carbon dioxide built up in the atmosphere and created the hellish Venus we see today. To a lesser degree, Mars and Earth, too, ap -pear to have suffered hydrodynamic losses. The telltale signature is a de˜cit of lighter isotopes, which are more easily lost. In the atmospheres of Earth and Mars, the ratio of neon 20 to neon 22 is 25 percent smaller than the solar ratio. On Mars, argon 36 is similarly depleted relative to argon 38. Even the isotopes of xenon Šthe heavi -est gas in Earth™s atmosphere apart from pollut -ants Šshow the imprint of hydrodynamic es -cape. If hydrodynamic escape were vigorous enough to sweep up xenon, why did it not sweep up everything else in the atmosphere along with it? To solve this puzzle, we may need to construct a different history for xenon than for the other gases now in the atmosphere. Hydrodynamic escape may have stripped Ti -LEAKING HYDROGEN ATO MS give off a red glow in this ultraviolet image of Earth™s night side, taken by NASA ™s Dynamic Explor -er I satellite in 1982. Oxygen and nitrogen account for the band around the North Pole and the wisps in the tropics.

PAGE – 5 ============
40 SCIENTIFIC AMERIC AN May 2009 heavier ions with them. This process may ex -plain the xenon puzzle: if the polar wind was more vigorous in the past, it could have dragged out xenon ions. One piece of evidence is that krypton does not have the same isotopic pattern as xenon does, even though it is a lighter gas and, all else being equal, ought to be more prone to escape. The difference is that krypton, unlike xe -non, resists ionization, so even a strong polar wind would have left it unaffected. A third nonthermal process known as photo -chemical escape operates on Mars and possibly on Titan. Oxygen, nitrogen and carbon monox -ide molecules drift into the upper atmosphere, where solar radiation ionizes them. When the ionized molecules recombine with electrons or collide with one another, the energy released splits the molecules into atoms with enough speed to escape. Mars, Titan and Venus lack global magnetic ˜elds, so they are also vulnerable to a fourth nonthermal process known as sputtering. With -out a planetary ˜eld to shield it, the upper atmo -sphere of each of these worlds is exposed to the full brunt of the solar wind. The wind picks up ions, which then undergo charge exchange and escape. Mars™s atmosphere is enriched in heavy nitrogen and carbon isotopes, suggesting that it ALFRED T. K AMAJI AN ( graph ); GEOR GE RETSEC K (illustration )tan of much of its air, too. When it descended through Titan™s atmosphere in 2005, the Euro -pean Space Agency™s Huygens probe found that the ratio of nitrogen 14 to nitrogen 15 is 70 per -cent of that on Earth. That is a huge disparity given that the two isotopes differ only slightly in their tendency to escape. If Titan™s atmosphere started with the same nitrogen isotopic compo -sition as Earth™s, it must have lost a huge amount of nitrogen Šseveral times the substantial amount it currently has Što bring the ratio down to its present value. In short, Titan™s atmosphere might once have been even thicker than it is to -day, which only heightens its mystery. Better Escaping through Chemistry On some planets, including modern Earth, ther -mal escape is less important than nonthermal escape. In nonthermal escape, chemical reactions or particle-particle collisions catapult atoms to escape velocity. What nonthermal escape mech -anisms have in common is that an atom or mol -ecule reaches a very high velocity as the outcome of a single event that takes place above the exobase, so that bumping into something does not thwart the escapee. Many types of nonther -mal escape involve ions. Ordinarily these charged particles are tethered to a planet by its magnetic ˜eld, either the global (internally generated) mag -netic ˜eld Šif there is one Šor the localized ˜elds induced by the passage of the solar wind. But they ˜nd ways to slip out. In one type of event, known as charge ex -change, a fast hydrogen ion collides with a neu -tral hydrogen atom and captures its electron. The result is a fast neutral atom, which is im -mune to the magnetic ˜eld. This process ac -counts for 60 to 90 percent of the present loss of hydrogen from Earth and most of the hydrogen loss from Venus. Another way out exploits a weak spot Šdare we say a loophole Šin the planet™s magnetic trap. Most magnetic ˜eld lines loop from one magnet -ic pole to the other, but the widest ˜eld lines are dragged outward by the solar wind and do not loop back; they remain open to interplanetary space. Through this opening, ions can escape. To be sure, the ions must still overcome gravity, and only the lightest ions such as hydrogen and helium make it. The resulting stream of charged particles, called the polar wind (not to be con -fused with the planetary wind), accounts for 10 to 15 percent of Earth™s hydrogen loss and al -most its entire helium leak. In some cases, these light ions can sweep up [T HE AU THORS ]Planetary scientist David C. Catling studies the coupled evolu -tion of planetary surfaces and atmospheres. Formerly at the NASA Ames Research Center, he joined the faculty at the University of Washington in 2001. He is a co- investigator for NASA ™s Phoenix lander, which completed its mission last December. Kevin J. Zahnle has been a research scien -tist at the NASA Ames center since 1989. Even by the eclectic stan -dards of planetary science, he has an unusually wide range of inter -ests, from planetary interiors to surfaces to atmospheres. In 1996 Zahnle received the NASA Excep -tional Achievement Medal for his work on the impact of Comet Shoemaker-Levy 9 into Jupiter. LIGHT GASES such as hydrogen are more footloose than heavier ones such as oxygen. Their susceptibility to Jeans escape depends on the temperature at the top of a body™s atmosphere or, for airless bodies such as the moon, at its surface ( vertical axis ) and on the strength of its gravity ( horizontal axis ). If a body lies to the right of the line for a gas, it holds on to the gas; to the left, it loses the gas. For example, Mars loses hydro -gen and helium, retains oxygen and carbon dioxide, and barely retains water. Hydrogen HeliumWaterOxygenCarbon dioxideTemperature Jupiter SaturnUranus NeptuneEarthVenus Mars Mercury MoonTitanPlutoStrength of Gravity WHICH GASES CAN ESCAPE ?

PAGE – 6 ============
The second broad way that air escapes is through charged particle reactions. Electric ˜elds readily accelerate ions to escape velocity. The planet™s magnetic ˜eld traps them, but they have various tricks to slip out. SNEAK O UT ALONG OPEN FIELD LINES Another escape route is along magnetic ˚eld lines at high latitudes, which do not loop back to the ground but instead link up with inter -planetary ˚elds. GET WHISKED A WAY BY SOLAR WIND A third process known as sputtering operates on worlds that lack magnetic ˚elds of their own. The solar wind, which is magnetized, picks up and spirits off ions. STEAL AN ELECTRON AND MAKE A GETA WAY One way for an ion to break free of the magnetic ˚eld is to bump into an uncharged atom and abscond with its electron, thereby becoming electrically neutral. [ESCAPE M ECHANIS M #2] The Houdini Particles Magnetic ˚eld line Ion Ion Electron Solar wind Mars Ion Magnetic ˚eld line SCIENTIFIC A MERIC AN 41

PAGE – 8 ============ SCIENTIFIC A MERIC AN 43 ALFRED T. K AMAJI AN ( icons )Mars has such a thin atmosphere. Scientists have long hypothesized that chemical reactions among water, carbon dioxide and rock turned the original thick atmosphere into carbonate minerals. The carbonates were never recycled back into carbon dioxide gas because Mars, be -ing so small, cooled quickly and its volcanoes stopped erupting. The trouble with this scenar -io is that spacecraft have so far found only a single small area on Mars with carbonate rock, and this outcrop probably formed in warm sub -surface waters. Moreover, the carbonate theory offers no explanation for why Mars has so little nitrogen or noble gases. Escape provides a bet -ter answer. The atmosphere did not get locked away as rock; it dissipated into space. A nagging problem is that impact erosion ought to have removed Mars™s atmosphere alto -gether. What stopped it? One answer is simple chance. Large impacts are inherently rare, and their frequency fell off rapidly about 3.8 billion years ago, so Mars may have been spared the ˜ -nal devastating blow. A large impact of an icy as -teroid or comet could have deposited more vola -tiles than subsequent impacts could remove. Al -ternatively, remnants of Mars™s atmosphere may MORE T O ˚ EXPL OREOrigins of Atmospheres. K. J. Zahnle in Origins . Edited by C. E. Woodward, J. M. Shull and H. A. Thronson. Astro -nomical Society of the Paci˜c Conference Series, Vol. 148, 1998. An Extended Upper Atmosphere around the Extrasolar Planet HD209458b. Alfred Vidal-Madjar et al. in Nature, Vol. 422, pages 143Œ146; March 13, 2003. Planetary Atmospheres and Life. D. C. Catling and J. F. Kasting in Planets and Life: The Emerging Science of Astrobiology. Edited by W. T. Sullivan and J. A. Baross. Cambridge University Press, 2007. have survived underground and leaked out after the bombardment had subsided. Although Earth seems comparatively un -scathed by escape, that will change. Today hy – drogen escape is limited to a trickle because the principal hydrogen-bearing gas, water vapor, condenses in the lower atmosphere and rains back to the surface. But our sun is slowly bright -ening at about 10 percent every billion years. That is imperceptibly slow on a human timescale but will be devastating over geologic time. As the sun brightens and our atmosphere warms, the atmosphere will get wetter, and the trickle of hy -drogen escape will become a torrent. This process is expected to become important when the sun is 10 percent brighter Šthat is, in a billion years Šand it will take another billion years or so to desiccate our planet™s oceans. Earth will become a desert planet, with at most a shrunken polar cap and only traces of precious liquid. After another two billion years, the sun will beat down on our planet so mercilessly even the polar oases will fail, the last liquid water will evaporate and the greenhouse effect will grow strong enough to melt rock. Earth will have fol -lowed Venus into a barren lifelessness. ˜[W HICH M ECHANIS MS APPLY W HERE ]A Litany of Losses The three escape processes operate to different degrees on different planets and at different points in their history. THER MAL NONTHER MAL IMPACT BODY PERIOD KEY GASES LOST JEANS ESCAPE HYDRO- DYNA MIC CHARGE EXCHANGE POLAR W IND PHOTO- CHE MICAL SPUT TER ING Earth Now Hydrogen ˛˛˛Helium ˛˛Prim ordial Hydrogen, neon ˛Venus Now Hydrogen, helium ˛˛Prim ordial Hydrogen, oxygen ˛Mars Now Hydrogen ˛Carbon, oxygen, nitrogen, argon ˛˛Prim ordial All gases ˛Hydrogen, carbon dioxide ˛Jupiter™s satellites Prim ordial All gases ˛˛Titan Now Hydrogen ˛˛Methane, nitrogen ?˛˛Prim ordial Hydrogen, methane, nitrogen ˛Pluto Now Hydrogen, methane, nitrogen ?HD 209458b Now Hydrogen, carbon, oxygen ˛

153 KB – 8 Pages