Earth as a Planet: Atmosphere and Oceans

Timothy E Dowling & Adam P Showman. Encyclopedia of the Solar System. Editor: Lucy-Ann McFadden, et al., 2nd Edition, Elsevier, 2007.

Earth is the only planet that orbits the Sun in the distance range within which water occurs in all three of its phases at the surface (as solid ice caps, liquid oceans, and atmospheric water vapor), which results in several unusual characteristics. Earth is unique in the solar system in exhibiting a global ocean at the surface, which covers almost three quarters of the planet’s area (such that the total amount of dry land is about equal to the surface area of Mars). The ocean exerts a strong control over the planet’s climate by transporting heat from equator to pole, interacting with the atmosphere chemically and mechanically, and, on geological timescales, influencing the exchange of volatiles between the planet’s atmosphere and interior. The Earth’s atmosphere follows the general pattern of a troposphere at the bottom, a stratosphere in the middle, and a thermosphere at the top. There is the usual east-west organization of winds, but with north-south and temporal fluctuations that are larger than found in any other atmosphere. Many of the atmospheric weather patterns (jet streams, Hadley cells, vortices, thunderstorms) occur on other planets too, but their manifestation on Earth is distinct and unique. The Earth’s climate has varied wildly over time, with atmospheric CO₂ and surface temperature fluctuating in response to ocean chemistry, planetary orbital variations, feedbacks between the atmosphere and interior, and a 30% increase in solar luminosity over the past 4.6 billion years (Ga). Despite these variations, the Earth’s climate has remained temperate, with at least partially liquid oceans, over the entire recorded ∼3.8 Ga geological record of the planet. Life has had a major influence on the ocean-atmosphere system, and as a result it is possible to discern the presence of life from remote spacecraft data. Global biological activity is indicated by the presence of atmospheric gases such as oxygen and methane that are in extreme thermodynamic disequilibrium, and by the widespread presence of a red-absorbing pigment (chlorophyll) that does not match the spectral signatures of any known rocks or minerals. The presence of intelligent life on Earth can be discerned from stable radio-wavelength signals emanating from the planet that do not match naturally occurring signals but do contain regular pulsed modulations that are the signature of information exchange.

Overview of Planetary Characteristics

Atmospheres are found on the Sun, 8 planets, and 7 of the 60-odd satellites, for a total count of 16—in addition to the atmospheres that exist around the ∼200 known gas giant planets orbiting other stars. Each has its own brand of weather and its own unique chemistry. They can be divided into two major classes: the terrestrial-planet atmospheres, which have solid surfaces or oceans as their lower boundary condition, and the gas giant atmospheres, which are essentially bottomless. Venus and Titan form one terrestrial subgroup that is characterized by a slowly rotating planet, and interestingly, both exhibit a rapidly rotating atmosphere. Mars, Io, Triton, and Pluto form a second terrestrial subgroup that is characterized by a thin atmosphere, which in large measure is driven by vapor-pressure equilibrium with the atmosphere’s solid phase on the surface. Both Io and Triton have active volcanic plumes. Earth’s weather turns out to be the most unpredictable in the solar system. Part of the reason is that its mountain ranges frustrate the natural tendency for winds to settle into steady east-west patterns, and a second reason is that its atmospheric eddies, the fluctuating waves and storm systems that deviate from the average, are nearly as big as the planet itself and as a result strongly interfere with each other.

Earth has many planetary attributes that are important to the study of its atmosphere and oceans, and conversely there are several ways in which its physically and chemically active fluid envelope directly affects the solid planet. Earth orbits the Sun at a distance of only 108 times the diameter of the Sun. The warmth from the Sun that Earth receives at this distance, together with a 30 K increase in surface temperature resulting from the atmospheric greenhouse effect, is exactly what is needed for H₂O to appear in all three of its phases. This property of the semimajor axis of Earth’s orbit is the most important physical characteristic of the planet that supports life. (One interesting consequence is that Earth is the only planet in the solar system where one can ski.)

Orbiting the Sun at just over 100 Sun diameters is not as close as it may sound; a good analogy is to view a basketball placed just past first base while standing at home plate on a baseball diamond. For sunlight, the Sun-to-Earth trip takes 499 s or 8.32 min. Earth’s semimajor axis, a₃ = 1.4960 × 10¹¹m = 1 AU (astronomical unit), and orbital period, τ₃ = 365.26 days = 1 year, where the subscript 3 denotes the third planet out from the Sun, are used as convenient measures of distance and time. When the orbital period of a body encircling the Sun, τ, is expressed in years, and its semimajor axis, a, is expressed in AU, then Kepler’s third law is simply τ = a^3/2, with a proportionality constant of unity.

Length of Day

The Earth’s rotation has an enormous effect on the motions of its fluid envelope that accounts for the circular patterns of large storms like hurricanes, the formation of western boundary currents like the Gulf Stream, the intensity of jet streams, the extent of the Hadley cell, and the nature of fluid instabilities. All of these processes are thoroughly discussed in Sections 2-5. Interestingly, the reverse is also true: The Earth’s atmosphere and oceans have a measurable effect on the planet’s rotation rate. For all applications but the most demanding, the time Earth takes to turn once on its axis, the length of its day, is adequately represented by a constant value equal to 24 hours or 1440 minutes or 86,400 seconds. The standard second is the Système International (SI) second, which is precisely 9,192,631,770 periods of the radiation corresponding to the transition between two hyperfine levels of the ground state of the ¹³³Cs atom. When the length of day is measured with high precision, it is found that Earth’s rotation is not constant. The same is likely to hold for any dynamically active planet. Information can be obtained about the interior of a planet, and how its atmosphere couples with its surface, from precise length-of-day measurements. Earth is the only planet to date for which we have achieved such accuracy, although we also have high-precision measurements of the rotation rate of pulsars, the spinning neutron stars often seen at the center of supernova explosions.

The most stable pulsars lose only a few seconds every million years and are the best-known timekeepers, even better than atomic clocks. In contrast, the rotating Earth is not an accurate clock. Seen from the ground, the positions as a function of time of all objects in the sky are affected by Earth’s variable rotation. Because the Moon moves across the sky relatively rapidly and its position can be determined with precision, the fact that Earth’s rotation is variable was first realized when a series of theories that should have predicted the motion of the Moon failed to achieve their expected accuracy. In the 1920s and 1930s, it was established that errors in the position of the Moon were similar to errors in the positions of the inner planets, and by 1939, clocks were accurate enough to reveal that Earth’s rotation rate has both irregular and seasonal variations.

The quantity of interest is the planet’s three-dimensional angular velocity vector as a function of time, Ω(t). Since the 1970s, time series of all three components of Ω(t) have been generated by using very long baseline interferometry (VLBI) to accurately determine the positions of quasars and laser ranging to accurately determine the positions of man-made satellites and the Moon, the latter with corner reflectors placed on the Moon by the Apollo astronauts.

The theory of Earth’s variable rotation combines ideas from geophysics, meteorology, oceanography, and astronomy. The physical causes fall into two categories: those that change the planet’s moment of inertia (like a spinning skater pulling in her arms) and those that torque the planet by applying stresses (like dragging a finger on a spinning globe). Earth’s moment of inertia is changed periodically by tides raised by the Moon and the Sun, which distort the solid planet’s shape. Nonperiodic changes in the solid planet’s shape occur because of fluctuating loads from the fluid components of the planet, namely, the atmosphere, the oceans, and, deep inside the planet, the liquid iron-nickel core. In addition, shifts of mass from earthquakes and melting ice cause nonperiodic changes. Over long timescales, plate tectonics and mantle convection significantly alter the moment of inertia and hence the length of day.

An important and persistent torque that acts on Earth is the gravitational pull of the Moon and the Sun on the solid planet’s tidal bulge, which, because of friction, does not line up exactly with the combined instantaneous tidal stresses. This torque results in a steady lengthening of the day at the rate of about 1.4 ms per century and a steady outward drift of the Moon at the rate of 3.7 ± 0.2 cm/year, as confirmed by lunar laser ranging. On the top of this steady torque, it has been suggested that observed 5 ms variations that have timescales of decades are caused by stronger, irregular torques from motions in Earth’s liquid core. Calculations suggest that viscous coupling between the liquid core and the solid mantle is weak, but that electromagnetic and topographic coupling can explain the observations. Mountains on the core-mantle boundary with heights around 0.5 km are sufficient to produce the coupling and are consistent with seismic tomography studies, but not much is known about the detailed topography of the core-mantle boundary. Detailed model calculations take into account the time variation of Earth’s external magnetic field, which is extrapolated downward to the core-mantle boundary. New improvements to the determination of the magnetic field at the surface are enhancing the accuracy of the downward extrapolations.

Earth’s atmosphere causes the strongest torques of all. The global atmosphere rotates faster than the solid planet by about 10 ms⁻¹ on average. Changes in the global circulation cause changes in the pressure forces that act on mountain ranges and changes in the frictional forces between the wind and the surface. Fluctuations on the order of 1 ms in the length of day, and movements of the pole by several meters, are caused by these meteorological effects, which occur over seasonal and interannual timescales. General circulation models (GCMs) of the atmosphere routinely calculate the global atmospheric angular momentum, which allows the meteorological and nonmeteorological components of the length of day to be separated. All the variations in the length of day over weekly and daily timescales can be attributed to exchanges of angular momentum between Earth’s atmosphere and the solid planet, and this is likely to hold for timescales of several months as well. Episodic reconfigurations of the coupled atmosphere-ocean system, such as the El Niño-Southern Oscillation (ENSO), cause detectable variations in the length of day, as do changes in the stratospheric jet streams.

Vertical Structure of the Atmosphere

Earth may differ in many ways from the other planets, but not in the basic structure of its atmosphere. Planetary exploration has revealed that essentially every atmosphere starts at the bottom with a troposphere, where temperature decreases with height at a nearly constant rate up to a level called the tropopause, and then has a stratosphere, where temperature usually increases with height or, in the case of Venus and Mars, decreases much less quickly than in the troposphere. It is interesting to note that atmospheres are warm both at their bottoms and their tops, but do not get arbitrarily cold in their interiors. For example, on Jupiter and Saturn there is significant methane gas throughout their atmospheres, but nowhere does it get cold enough for methane clouds to form, whereas in the much colder atmospheres of Uranus and Neptune, methane clouds do form. Details vary in the middle-atmosphere regions from one planet to another, where photochemistry is important, but each atmosphere is topped off by a high-temperature, low-density thermosphere that is sensitive to solar activity and an exobase, the official top of an atmosphere, where molecules float off into space when they achieve escape velocity.

Interestingly, the top of the troposphere occurs at about the same pressure, about 0.1-0.3 bar, on most planets. This similarity is not coincidental but instead results from the pressure dependence of the atmospheric opacity to solar and especially infrared radiation. In the high-pressure regime of tropospheres, the gas is relatively opaque at infrared wavelengths, which inhibits heat loss by radiation from the deep levels and hence promotes a profile where temperature decreases strongly with altitude. In the low-pressure regime of stratospheres, the gas becomes relatively transparent at infrared wavelengths, which allows the temperature to become more constant—or in some cases even increase—with altitude. This transition from opaque to transparent tends to occur at pressures of 0.1-0.3 bar for the compositions of most planetary atmospheres in our solar system.

In the first 0.1 km of a terrestrial atmosphere, the effects of daily surface heating and cooling, surface friction, and topography produce a turbulent region called the planetary boundary layer, or PBL. Right at the surface, molecular viscosity forces the “no slip” boundary condition and the wind reduces to zero, such that even a weak breeze results in a strong vertical wind shear that can become turbulent near the surface. However, only a few millimeters above the surface, molecular viscosity ceases to play a direct role in the dynamics, except as a sink for the smallest eddies. The mixing caused by turbulent eddies is often represented as a viscosity with a strength that is a million times or more greater than the molecular viscosity.

Up to altitudes of about 80 km, Earth’s atmosphere is composed of 78% N₂, 21% O₂, 0.9% Ar, and 0.002% Ne by volume, with trace amounts of CO₂, CH₄, and numerous other compounds. Diffusion, chemistry, and other effects substantially alter the composition at altitudes above ∼90 km.

Troposphere

The troposphere is the lowest layer of the atmosphere, characterized by a temperature that decreases with altitude. The top of the troposphere is called the tropopause, which occurs at an altitude of 18 km at the equator but only 8 km at the poles (the cruising altitude of commercial airliners is typically 10 km). Gravity, combined with the compressibility of air, causes the density of an atmosphere to fall off exponentially with height, such that Earth’s troposphere contains 80% of the mass and most of the water vapor in the atmosphere, and consequently most of the clouds and stormy weather. Vertical mixing is an important process in the troposphere. Temperature falls off with height at a predictable rate because the air near the surface is heated and becomes light, and the air higher up cools to space and becomes heavy, leading to an unstable configuration and convection. The process of convection relaxes the temperature profile toward the neutrally stable configuration, called the adiabatic temperature lapse rate, for which the decrease of temperature with decreasing pressure (and hence increasing height) matches the drop-off of temperature that would occur inside a balloon that conserves its heat as it moves, that is, moves adiabatically.

In the troposphere, water vapor, which accounts for up to ∼1% of air, varies spatially and decreases rapidly with altitude. The water vapor mixing ratio in the stratosphere and above is almost 4 orders of magnitude smaller than that in the tropical lower troposphere.

Stratosphere

The nearly adiabatic falloff of temperature with height in Earth’s troposphere gives way above the tropopause to an increase of temperature with height. This results in a rarified, stable layer called the stratosphere. Observations of persistent, thin layers of aerosol and of long residence times for radioactive trace elements from nuclear explosions are direct evidence of the lack of mixing in the stratosphere. The temperature continues to rise with altitude in Earth’s stratosphere until one reaches the stratopause at about 50 km. The source of heating in Earth’s stratosphere is the photochemistry of ozone, which peaks at about 25 km. Ozone absorbs ultraviolet (UV) light, and below about 75 km nearly all this radiation gets converted into thermal energy. The Sun’s UV radiation causes stratospheres to form in other atmospheres, but instead of the absorber being ozone, which is plentiful on Earth because of the high concentrations of O₂ maintained by the biosphere, other gases absorb the UV radiation. On the giant planets, methane, hazes, and aerosols do the job.

The chemistry of Earth’s stratosphere is complicated. Ozone is produced mostly over the equator, but its largest concentrations are found over the poles, meaning that dynamics is as important as chemistry to the ozone budget. Mars also tends to have ozone concentrated over its poles, particularly over the winter pole. The dry martian atmosphere has relatively few hydroxyl radicals to destroy the ozone. Some of the most important chemical reactions in Earth’s stratosphere are those that involve only oxygen. Photodissociation by solar UV radiation involves the reactions O₂ + hv → O + O and O₃ + hv → O + O₂, where hv indicates the UV radiation. Three-body collisions, where a third molecule, M, is required to satisfy conservation of momentum and energy, include O + O + M → O₂ + M and O + O₂ + M → O₃ + M, but the former reaction proceeds slowly and may be neglected in the stratosphere. Reactions that either destroy or create “odd” oxygen, O or O₃, proceed at much slower rates than reactions that convert between odd oxygen. The equilibrium between O and O₃ is controlled by fast reactions that have rates and concentrations that are altitude-dependent. Other reactions that are important to the creation and destruction of ozone involve minor constituents such as NO, NO₂, H, OH, HO₂, and Cl. An important destruction mechanism is the catalytic cycle X + O₃ → XO + O₂ followed by XO + O → X + O₂, which results in the net effect O + O₃ → 2O₂. On Earth, human activity has led to sharp increases in the catalysts X = Cl and NO and subsequent sharp decreases in stratospheric ozone, particularly over the polar regions. The Montreal Protocol is an international treaty signed in 1987 that is designed to stop and eventually reverse the damage to the stratospheric ozone layer; regular meetings of the parties, involving some 175 countries, continually update the protocol.

Mesosphere

Above Earth’s stratopause, temperature again falls off with height, although at a slower rate than in the troposphere. This region is called the mesosphere. Earth’s stratosphere and mesosphere are often referred to collectively as the middle atmosphere. Temperatures fall off in the mesosphere because there is less heating by ozone and emission to space by carbon dioxide is an efficient cooling mechanism. The mesopause occurs at an altitude of about 80 km, marking the location of a temperature minimum of about 130 K.

Thermosphere

As is the case for ozone in Earth’s stratosphere, above the mesopause, atomic and molecular oxygen strongly absorb solar UV radiation and heat the atmosphere. This region is called the thermosphere, and temperatures rise with altitude to a peak that varies between about 500 and 2000 K depending on solar activity. Just as in the stratosphere, the thermosphere is stable to vertical mixing. At about 120 km, molecular diffusion becomes more important than turbulent mixing, and this altitude is called the homopause (or turbopause). Rocket trails clearly mark the homopause—they are rapidly mixed below this altitude but linger relatively undisturbed above it. Molecular diffusion is mass-dependent and each species falls off exponentially with its own scale height, leading to elemental fractionation that enriches the abundance of the lighter species at the top of the atmosphere.

For comparison with Earth, the structure of the thermospheres of the giant planets has been determined from Voyager spacecraft observations, and the principal absorbers of UV light are H₂, CH₄, C₂H₂, and C₂H₆. The thermospheric temperatures of Jupiter, Saturn, and Uranus are about 1000, 420, and 800 K, respectively. The high temperature and low gravity on Uranus allow its upper atmosphere to extend out appreciably to its rings.

Exosphere and Ionosphere

At an altitude of about 500 km on Earth, the mean free path between molecules grows to be comparable to the density scale height (the distance over which density falls off by a factor of e ≈ 2.7128). This defines the exobase and the start of the exosphere. At these high altitudes, sunlight can remove electrons from atmospheric constituents and form a supply of ions. These ions interact with a planet’s magnetic field and with the solar wind to form an ionosphere. On Earth, most of the ions come from molecular oxygen and nitrogen, where as on Mars and Venus most of the ions come from carbon dioxide. Because of the chemistry, however, ionized oxygen atoms and molecules are the most abundant ion for all three atmospheres.

Mechanisms of atmospheric escape fall into two categories, thermal and nonthermal. Both processes provide the kinetic energy necessary for molecules to attain escape velocity. When escape velocity is achieved at or above the exobase, such that further collisions are unlikely, molecules escape the planet. In the thermal escape process, some fraction of the high-velocity wing of the Maxwellian distribution of velocities for a given temperature always has escape velocity; the number increases with increasing temperature. An important nonthermal escape process is dissociation, both chemical and photochemical. The energy for chemical dissociation is the excess energy of reaction, and for photochemical dissociation, it is the excess energy of the bombarding photon or electron, either of which is converted into kinetic energy in the dissociated atoms. A common effect of electrical discharges of a kilovolt or more is “sputtering,” where several atoms can be ejected from the spark region at high velocities. If an ion is formed very high in the atmosphere, it can be swept out of a planet’s atmosphere by the solar wind. Similarly at Io, ions are swept away by Jupiter’s magnetic field. Other nonthermal escape mechanisms involve charged particles. Charged particles get trapped by magnetic fields and therefore do not readily escape. However, a fast proton can collide with a slow hydrogen atom and take the electron from the hydrogen atom. This charge-exchange process changes the fast proton into a fast, hydrogen atom that is electrically neutral and hence can escape.

Nonthermal processes account for most of the present-day escape flux from Earth, and the same is likely to be true for Venus. They are also invoked to explain the 62 ± 16% enrichment of the ¹⁵N/¹⁴N ratio in the martian atmosphere. If the current total escape flux from thermal and nonthermal processes is applied over the age of the solar system, the loss of hydrogen from Earth is equivalent to only a few meters of liquid water, which means that Earth’s sea level has not been affected much by this process. However, the flux could have been much higher in the past, since it is sensitive to the structure of the atmosphere.

Atmospheric Circulation

Processes Driving the Circulation

The atmospheric circulation on Earth, as on any planet, involves a wealth of phenomena ranging from global weather patterns to turbulent eddies only centimeters across and varies over periods of seconds to millions of years. All this activity is driven by absorbed sunlight and loss of infrared (heat) energy to space. Of the sunlight absorbed by Earth, most (∼70%) penetrates through the atmosphere and is absorbed at the surface; in contrast, the radiative cooling to space occurs not primarily from the surface but from the upper troposphere at an altitude of 5-10 km. This mismatch in the altitudes of heating and cooling means that, in the absence of air motions, the surface temperature would increase while the upper tropospheric temperature would decrease. However, such a trend produces an unstable density stratification, forcing the troposphere to overturn. The hot air rises, the cold air sinks, and thermal energy is thus transferred from the surface to the upper troposphere. This energy transfer by air motions closes the “energy loop,” allowing the development of a quasi-steady state where surface and atmospheric temperatures remain roughly steady in time. This vertical mixing process is fundamentally responsible for near-surface convection, turbulence, cumulus clouds, thunderstorms, hurricanes, dust devils, and a range of other small-scale weather phenomena.

At global scales, much of Earth’s weather results not simply from vertical mixing but from the atmosphere’s response to horizontal temperature differences. Earth absorbs most of the sunlight at the equator, yet it loses heat to space everywhere over the surface. This mismatch makes the near-surface equatorial air hot and the polar air cold. This configuration is gravitationally unstable—the hot equatorial air has low density and the cold polar air has high density. Just as the cold air from an open refrigerator slides across your feet, the cold polar air slides under the hot equatorial air, lifting the hotter air upward and poleward while pushing the colder air downward and equatorward. This overturning process transfers energy between the equator and the poles and leads to a much milder equator-to-pole temperature difference (about 30 K at the surface) than would exist in the absence of such motions. On average, the equatorial regions gain more energy from sunlight than they lose as radiated heat, while the reverse holds for the poles; the difference is transported between equator and pole by the air and ocean. The resulting atmospheric overturning causes many of Earth’s global-scale weather patterns, such as the 1000 km long fronts that cause much midlatitude weather and the organization of thunderstorms into clusters and bands. Horizontal temperature and density contrasts can drive weather at regional scales too; examples include air-sea breezes and monsoons.

Influence of Rotation

The horizontal pressure differences associated with horizontal temperature differences cause a force (the “pressure-gradient force”) that drives most air motion at large scales. However, how an atmosphere responds to this force depends strongly on whether the planet is rotating. On a nonrotating planet, the air tends to directly flow from high to low pressure, following the “nature abhors a vacuum” dictum. If the primary temperature difference occurs between equator and pole, this would lead to a simple overturning circulation between the equator and pole. On the other hand, planetary rotation (when described in a noninertial reference frame rotating with the solid planet) introduces new forces into the equations of motion: the centrifugal force and the Coriolis force. The centrifugal force naturally combines with the gravitational force and the resultant force is usually referred to as simply the gravity. For rapidly rotating planets, the Coriolis force is the dominant term that balances the horizontal pressure-gradient force in large-scale circulations (a balance called geostrophy). Because the Coriolis force acts perpendicular to the air motion, this leads to a fascinating effect—the horizontal airflow is perpendicular to the horizontal pressure gradient. A north-south pressure gradient (resulting from a hot equator and a cold pole, for example) leads primarily not to north-south air motions but to east-west air motions! This is one reason why east-west winds dominate the circulation on most planets, including Earth. For an Earth-sized planet with Earth-like wind speeds, rotation dominates the large-scale dynamics as long as the planet rotates at least once every 10 days.

Physically, the Coriolis force acts in the following way. Air moving eastward (i.e., in the same direction as the planet’s rotation, but faster) experiences a force that moves it away from the rotation axis—namely, equatorward—just as a child experiences an outward force on a spinning merry-go-round. Conversely, air moving westward (in the same direction as the planet’s rotation, but slower) would experience a poleward force. And, just as an ice skater spins faster as she pulls in her arms, air that moves toward the planetary rotation axis—namely, poleward—spins faster, which is equivalent to saying that it deflects eastward. Conversely, air that moves away from the planetary rotation axis (equatorward) deflects westward. If one pays attention to the directions of the force in each of these cases, one sees that, in the northern hemisphere, this rotationally induced force is always to the right of the air motion, while in the southern hemisphere, it is always to the left of the air motion.

Two other important effects of rapid rotation are the suppression of motions in the direction parallel to the rotation axis, called the Taylor-Proudman effect, and the coupling of horizontal temperature gradients with vertical wind shear, a 3-dimensional relationship described by the thermal-wind equation.

Observed Global-Scale Circulation

As described earlier, the atmospheric circulation organizes primarily into a pattern of east-west winds, and perhaps the most notable feature is the eastward-blowing jet streams in the midlatitudes of each hemisphere. In a longitudinal and seasonal average, the winter hemisphere wind maximum reaches 40 m s⁻¹ at 30° latitude, and the summer hemisphere wind maximum reaches 20-30 m s⁻¹ at 40-50^° latitude. In between these eastward wind maxima, from latitude 20°N to 20°S the tropospheric winds blow weakly westward. The jet streams are broadly distributed in height, with peak speeds at about 12-km altitude. Although the longitudinally and seasonally averaged winds exhibit only a single tropospheric eastward-wind maximum in each hemisphere, instantaneous 3-dimensional snapshots of the atmosphere illustrate that there often exist two distinct jet streams, the subtropical jet at ∼30° latitude and the polar jet at ∼50° latitude. These jets are relatively narrow—a few hundred km in latitudinal extent—and can reach speeds up to 100 m s⁻¹. However, the intense jet cores are usually less than a few thousand kilometers in longitudinal extent (often residing over continental areas such as eastern Asia and eastern North America), and the jets typically exhibit wide, time-variable wavelike fluctuations in position. When averaged over longitude and time, these variations in the individual jet streams smear into the single eastward maximum evident in each hemisphere.

Although the east-west winds dominate the time-averaged circulation, weaker vertical and latitudinal motion are required to transport energy from the equator to the poles. Broadly speaking, this transport occurs in two distinct modes. In the tropics exists a direct thermal overturning circulation called the Hadley cell, where, on average, air rises near the equator, moves poleward, and descends. This is an extremely efficient means of transporting heat and contributes to the horizontally homogenized temperatures that exist in the tropics. However, planetary rotation prevents the Hadley cell from extending all the way to the poles (to conserve angular momentum about the rotation axis, equatorial air would accelerate eastward to extreme speeds as it approached the pole, a phenomenon that is dynamically inhibited). On Earth, the Hadley cell extends to latitudes of ∼30°. Poleward of ∼30°, the surface temperatures decrease rapidly toward the pole; this is the location of the subtropical jet. Although planetary rotation inhibits the Hadley cell in this region, north-south motions still occur via a complex 3-dimensional process called baroclinic instability. Meanders on the jet stream grow, pushing cold high-latitude air under warm low-latitude air in confined regions ∼1000-5000 km across. These instabilities grow, mature, and decay over ∼5 day periods; new ones form as old ones disappear. These structures evolve to form regions of sharp thermal gradient called fronts, as well as 1000-5000 km long arc-shaped clouds and precipitation that dominate much of the winter weather in the United States, Europe, and other midlatitude regions.

Water vapor in Earth’s troposphere greatly accentuates convective activity because latent heat is liberated when moist air is raised above its lifting-condensation level, and this further increases the buoyancy of the rising air, leading to moist convection. Towering thunderstorms get their energy from this process, and hurricanes are the most dramatic and best-organized examples of moist convection. Hurricanes occur only on Earth because only Earth provides the necessary combination of high humidity and surface friction. Surface friction is required to cause air to spiral into the center of the hurricane, where it is then forced upward past its lifting-condensation level.

The Hadley cell exerts a strong control over weather in the tropics. The upward transport in the ascending branch of the Hadley circulation occurs almost entirely in localized thunderstorms whose convective towers cover only a small fraction (perhaps ∼1%) of the total horizontal area of the tropics. Because this ascending branch resides near the equator, equatorial regions receive abundant rainfall, allowing the development of tropical rainforests in Southeast Asia/Indonesia, Brazil, and central Africa.

On the other hand, this condensation and rainout of water dehydrates the air, so the descending branch of the Hadley cell, which occurs in the subtropics at ∼20-30° latitude, is relatively dry. Because of the descending motion and dry conditions, little precipitation falls in these regions, which explains the abundance of arid biomes at 20-30° latitude, including the deserts of the African Sahara, southern Africa, Australia, central Asia, and the southwestern United States. However, the simple Hadley cell is to some degree a theoretical idealization, and many regional 3-dimensional time-variable phenomena—including monsoons, equatorial waves, El Niño, and longitudinal overturning circulations associated with continent-ocean and sea-surface-temperature contrasts—affect the locations of tropical thunderstorm formation and hence the climatic rainfall patterns.

Satellite images dramatically illustrate the signature of the Hadley cell and midlatitude baroclinic instabilities as manifested in clouds. The east-west band of clouds stretching across the disk of Earth just north of the equator corresponds to the rising branch of the Hadley cell (this cloud band is often called the intertropical convergence zone). These clouds are primarily the tops of thunderstorm anvils. In the midlatitude regions of both hemispheres (30-70° latitude), several arc-shaped clouds up to 3000-5000 km long can be seen. These are associated with baroclinic instabilities. These clouds, which can often dominate midlatitude winter precipitation, form when large regions of warm air are forced upward over colder air masses during growing baroclinic instabilities. In many cases, the forced ascent associated with these instabilities produces predominantly sheet-like stratus clouds and steady rainfall lasting for several days, although sometimes the forced ascent can trigger local convection events (e.g., thunderstorms).

What causes the jet stream? This is a subtle question. At the crudest level, poleward-moving equatorial air deflects eastward due to the Coriolis acceleration (or, equivalently, due to the air’s desire to conserve angular momentum about the planetary rotation axis), so the formation of eastward winds in the midlatitudes is a natural response to the Hadley circulation. These strong eastward winds in mid-latitudes are also consistent with the large latitudinal thermal gradients in midlatitudes via the thermal-wind equation mentioned in Section 3.2. However, these processes alone would tend to produce a relatively broad zone of eastward flow rather than a narrow jet. Nonlinear turbulent motions, in part associated with baroclinic instabilities, pump momentum up-gradient into this eastward-flowing zone and help to produce the narrow jet streams.

Although the Earth’s equator is hotter than the poles at the surface, it is noteworthy that, in the upper troposphere and lower stratosphere (∼18 km altitude), the reverse is true. This seems odd because sunlight heats the equator much more strongly than the poles. In reality, the cold equatorial upper troposphere results from a dynamical effect: Large-scale ascent in the tropics causes air to expand and cool (a result of decreased pressure as the air rises), leading to the low temperatures despite the abundant sunlight. Descent at higher latitudes causes compression and heating, leading to warmer temperatures. Interestingly, this means that, in the lower stratosphere, the ascending air is actually denser than the descending air. Such a circulation, called a thermally indirect circulation, is driven by the absorption of atmospheric waves that are generated in the troposphere and propagate upward into the stratosphere. There is a strong planetary connection because all four giant planets—Jupiter, Saturn, Uranus, and Neptune—are also thought to have thermally indirect circulations in their stratospheres driven by analogous processes.

Insights from other Atmospheres

Planetary exploration has revealed that atmospheric circulations come in many varieties. Perhaps ironically, Earth is observed to have the most unpredictable weather of all. The goal of planetary meteorology is to understand what shapes and maintains these diverse circulations. The Voyager spacecraft provided close-up images of the atmospheres of Jupiter, Saturn, Uranus, and Neptune and detailed information on the three satellites that have atmospheres thick enough to sport weather—Io, Titan, and Triton. The atmospheres of Venus and Mars have been sampled by entry probes, landers, orbiting spacecraft, and telescopic studies. Basic questions like why Venus’ atmosphere rotates up to 60 times faster than does the planet, or why Jupiter and Saturn have superrotating equatorial jets, do not have completely satisfactory explanations. However, by comparing and contrasting each planet’s weather, a general picture has begun to emerge.

Theoretical studies and comparative planetology show that planetary rotation rate and size exert a major control over the type of global atmospheric circulations that occur. When the rotation rate is small, Hadley cells are unconfined and stretch from the equator to pole. Venus, with a rotation period of 243 days, seems to reside in such a state. Titan rotates in 16 days and, according to circulation models, its Hadley cell extends to at least ∼60° latitude, a transitional regime between Venus and Earth. On the other hand, fast rotation confines the Hadley cell to a narrow range of latitudes (0-30° on Earth) and forces baroclinic instabilities to take over much of the heat transport between low latitudes and the poles. Increasing the rotation rate still further—or making the planet larger—causes the midlatitudes to break into series of narrow latitudinal bands, each with their own east-west jet streams and baroclinic instabilities. The faster the rotation rate, the straighter and narrower are the bands and jets. This process helps explain the fact that Jupiter and Saturn, which are large and rapidly rotating, have ∼30 and 20 jet streams, respectively (as compared to only a few jet streams for Earth). Fast rotation also contributes to smaller structures because it inhibits free movement of air toward or away from pressure lows and highs, instead causing the organization of vortices around such structures. Thus, a planet identical to Earth but with a faster or slower rotation rate would exhibit different circulations, equatorial and polar temperatures, rainfall patterns, and cloud patterns, and hence would exhibit a different distribution of deserts, rainforests, and other biomes.

The giant planets Jupiter and Saturn exhibit numerous oval-shaped windstorms that superficially resemble terrestrial hurricanes. However, hurricanes can generate abundant rainfall because friction allows near-surface air to spiral inward toward the low-pressure center, providing a source of moist air that then ascends inside thunderstorms; in turn, these thunderstorms release energy that maintains the hurricane’s strength against the frictional energy losses. In contrast, windstorms like Jupiter’s Great Red Spot and the hundreds of smaller ovals seen on Jupiter, as well as the dozens seen on Saturn and the couple seen on Neptune, do not directly require moist convection to drive them and hence are not hurricanes. Instead, they are simpler systems that are closely related to three types of long-lasting, high-pressure “storms,” or coherent vortices, seen on Earth: blocking highs in the atmosphere and Gulf Stream rings and Mediterranean salt lenses (“meddies”) in the ocean. Blocking highs are high-pressure centers that stubbornly settle over continents, particularly in the United States and Russia, thereby diverting rain from its usual path for months at a time. For example, the serious 1988 drought in the U.S. Midwest was exacerbated by a blocking high. Gulf Stream rings are compact circulations in the Atlantic that break off from the meandering Gulf Stream, which is a river inside the Atlantic Ocean that runs northward along the eastern coast of the United States and separates from the coast at North Carolina, where it then jets into the Atlantic in an unsteady manner. Seen in three dimensions, the Gulf Stream has the appearance of a writhing snake. Similar western boundary currents occur in other ocean basins, for example, the Kuroshio Current off the coast of Japan and the Agulhas Current off the coast of South Africa. Jet streams in the atmosphere are a related phenomenon. When Gulf Stream rings form, they trap phytoplankton and zooplankton inside them, which are carried large distances. Over the course of a few months, the rings dissipate at sea, are reabsorbed into the Gulf Stream, or run into the coast, depending on which side of the Gulf Stream they formed. The ocean plays host to another class of long-lived vortices, Mediterranean salt lenses, which are organized high-pressure circulations that float under the surface of the Atlantic. They form when the extra-salty water that slips into the Atlantic from the shallow Mediterranean Sea breaks off into vortices. After a few years, these meddies eventually wear down as they slowly mix with the surrounding water. The mathematical description of these long-lasting vortices on Earth is the same as that used to describe the ovals seen on Jupiter, Saturn, and Neptune.

Given that we know that atmospheric motions are fundamentally driven by sunlight, and we know that the problem is governed by Newton’s laws of motion, why then are atmospheric circulations difficult to understand? Several factors contribute to the complexity of observed weather patterns. In the first place, fluids move in an intrinsically nonlinear fashion that makes paper-and-pencil analysis formidable and often intractable. Laboratory experiments and numerical experiments performed on high-speed computers are often the only means for making progress on problems in geophysical fluid dynamics. In the second place, meteorology involves the intricacies of moist thermodynamics and precipitation, and we are only beginning to understand and accurately model the microphysics of these processes. And for the terrestrial planets, a third complexity arises from the complicated boundary conditions that the solid surface presents to the problem, especially when mountain ranges block the natural tendency for winds to organize into steady east-west jet streams. For oceanographers, even more restrictive boundary conditions apply, namely, the ocean basins, which strongly affect how currents behave. The giant planets are free of this boundary problem because they are completely fluid down to their small rocky cores. However, the scarcity of data for the giant planets, especially with respect to their vertical structure beneath the cloud tops, provides its own set of difficulties.

Ironic Unpredictability—an Anecdote

The fickleness of Earth’s weather compared to that of the other planets provides many fascinating scientific problems for meteorologists. Trying to live on such a planet presents Earth’s inhabitants with practical problems as well. On the lighthearted side, there are common bromides such as “If you don’t like the weather, wait 15 minutes,” and “Everybody complains about the weather, but nobody does anything about it.” On the serious side, lightning storms and tornados wreak havoc every year, and before the advent of weather satellites, hurricanes once struck populated coastlines without warning, causing terrible loss of life.

Even now, the tracks of hurricanes are notoriously difficult to predict. The point is best made with an example, and the following is a lighthearted anecdote from the first author’s personal experience: Perhaps he should have known better than to leave the windows of his apartment open on such a warm, breezy morning in the summer of 1991, but the apartment needed airing out, and the author was preoccupied with a desire to come up with a good way to illustrate to a group of distinguished terrestrial meteorologists that the weather on Jupiter is more predictable than the weather on Earth. And so, he left the windows open, locked the door, and headed out to Boston’s Logan Airport to begin a 10 day trip to a symposium on “Vortex Dynamics in the Atmosphere and Ocean,” which was being held in Vienna. His preoccupation was not helped by the use of the singular “atmosphere” in the symposium’s title, which, one could argue, should have been written with the plural “atmospheres.” To be sure, Earth has its great vortices, like Gulf Stream rings, Mediterranean salt lenses, and atmospheric blocking highs, and even more powerful storms, like hurricanes, which are driven by moist thermodynamics (in fact, Hurricane Bob was at that moment slowly heading toward the Carolina coast). Yet Jupiter’s Great Red Spot and the hundreds of other long-lived vortices found on the gas giant planets are in many ways simpler systems to study, and we have excellent observations of them from spacecraft like Voyager, Galileo, and Cassini.

After arriving at the conference, the author decided to make his case by pointing out that a Voyager-style mission to track hurricanes on Earth would most likely end in failure. This is because the Voyager cameras had to be choreographed 30 days in advance of each encounter to give the flight engineers time to sort through the conflicting requests of the various scientists and time to program the onboard computer. For the atmospheric working group, this constraint meant that success or failure depended on the accuracy of 30-day weather forecasts for the precise locations of the drifting Great Red Spot and other targeted features. On Earth, storms rarely last 30 days, and much less do they end up where they are predicted to be going a month in advance. The fact that the Voyager missions to Jupiter were a complete success, as were the subsequent Saturn, Uranus, and Neptune missions, illustrates in a practical way the remarkable predictability of the weather on the gas giant planets relative to on Earth.

Having made his point, on the road back to the Vienna Airport after the conference the author was getting accustomed to the fact that taxis in German-speaking countries are Mercedes, when the driver explained that the announcer on the radio was saying that there had been a coup in Moscow. This left him worried about his Russian colleagues, several of whom he had just met in the preceding week. On the flight back across the Atlantic, he was thinking about this when the Lufthansa pilot announced, with resignation in his voice, that because of thunderstorms the plane could not land in Boston and was being redirected to Montreal. After about 2 hours in Montreal, where the plane was nestled between several other waylaid international planes that were littered across the tarmac, the go-ahead was given to finish the trip to Boston. The landing was bumpy, and the skyline was disturbingly dark, but there was a beautiful sunset that was framed with orange, red, and black clouds. It was only after getting off the plane that the author first learned that Hurricane Bob had just hit Boston. Boston? Wasn’t Bob supposed to hit the Carolina coast? It was difficult not to take this egregious forecasting error personally. On returning home, jet-lagged, the author discovered that his apartment was dark, the electricity was out, the windows were of course still open, the curtains, carpet, and furniture were soaked, and wall hangings and broken glass were strewn about the floor. The irony of the situation is not hard to grasp. Voyager would have returned beautiful, fair weather images of North Carolina and South Carolina, and would have completely missed the hurricane, which ended up passing through this author’s apartment 1000 km north of the previous week’s prediction.

Oceans

Earth is the only planet in the solar system with a global ocean at the surface. The oceans have an average thickness of 3.7 km and cover 71% of Earth’s area; the greatest thickness is 10.9 km, which occurs at the Marianas Trench. The total oceanic mass—1.4 × 10²¹ kg—exceeds the atmosphere mass of 5 × 10¹⁸ kg by nearly a factor of 300, implying that the oceans dominate Earth’s surface inventory of volatiles. (One way of visualizing this fact is to realize that, if Earth’s entire atmosphere condensed as ices on the surface, it would form a layer only ∼10 m thick.) The Earth therefore sports a greater abundance of fluid volatiles at its surface than any other solid body in the solar system. Even Venus’ 90 bar CO₂ atmosphere contains only one third the mass of Earth’s oceans. On the other hand, Earth’s oceans constitute only 0.02% of Earth’s total mass; the mean oceanic thickness of 3.7 km pales in comparison to Earth’s 6400 km radius, implying that the oceans span only 0.06% of Earth’s width. The Earth is thus a relatively dry planet, and the oceans truly are only skin deep.

It is possible that Earth’s solid mantle contains a mass of dissolved water (stored as individual water molecules inside and between the rock grains) equivalent to several oceans’ worth of water. Taken together, however, the total water in Earth probably constitutes less than 1% of Earth’s mass. In comparison, most icy satellites and comets in the outer solar system contain ∼40-60%H₂O by mass, mostly in solid form. This lack of water on Earth in comparison to outer solar-system bodies reflects the relatively dry conditions in the inner solar system when the terrestrial planets formed; indeed, the plethora of water on Earth compared to Venus and Mars has raised the question of whether even the paltry amount of water on Earth must have been delivered from an outer-solar-system source such as impact of comets onto the forming Earth.

The modern oceans can be subdivided into the Pacific, Atlantic, Indian, and Arctic Oceans, but these four oceans are all connected, and this contiguous body of water is often simply referred to as the global ocean.

Oceanic Structure

The top meter of ocean water absorbs more than half of the sunlight entering the oceans; even in the sediment-free open ocean, only 20% of the sunlight reaches a depth of 10 m and only ∼1% penetrates to a depth of 100 m (depending on the angle of the Sun from vertical). Photosynthetic single-celled organisms, which are extremely abundant near the surface, can thus only survive above depths of ∼100 m; this layer is called the photic zone. The much thicker aphotic zone, which has too little light for photosynthetic production to exceed respiration, extends from ∼100 m to the bottom of the ocean. Despite the impossibility of photosynthesis at these depths, the deep oceans nevertheless exhibit a wide variety of life fueled in part by dead organic matter that slowly sediments down from the photic zone.

From a dynamical point of view, the ocean can be subdivided into several layers. Turbulence caused by wind and waves homogenizes the top 20-200 m of the ocean (depending on weather conditions), leading to profiles of density, temperature, salinity, and composition that vary little across this layer, which is therefore called the mixed layer. Below the mixed layer lies the thermocline, where the temperature generally decreases with depth down to ∼0.5-1 km. The salinity also often varies with depth between ∼100-1000 m, a layer called the halocline. For example, regions of abundant precipitation but lesser evaporation, such as the North Pacific, have relatively fresh surface waters, so the salinity increases with depth below the mixed layer in those regions. The variation of temperature and salinity between ∼100-1000 m implies that density varies with depth across this layer too; this is referred to as the pycnocline. Below the thermocline, halocline, and pycnocline lies the deep ocean, where temperatures are usually relatively constant with depth at a chilly 0-4°C.

The temperature at the ocean surface varies strongly with latitude, with only secondary variations in longitude. Surface temperatures reach 25-30°C near the equator, where abundant sunlight falls, but plummet to 0°C near the poles. In contrast, the deep oceans (>1 km) are generally more homogeneous and have temperatures between 0-4°C all over the world. (When enjoying the bathtub-temperature water and coral reefs during a summer vacation to a tropical island, it is sobering to think that if one could only scuba dive deep enough, the temperature would approach freezing.) This latitude-dependent upper-ocean structure implies that the thermocline and pycnocline depths decrease with latitude: They are about ∼1 km near the equator and reach zero near the poles.

Because warmer water is less dense than colder water, the existence of a thermocline over most of the ocean implies that the top ∼1 km of the ocean is less dense than the underlying deep ocean. The implication is that, except for localized regions near the poles, the ocean is stable to vertical convective overturning.

Ocean Circulation

Ocean circulation differs in important ways from atmospheric circulation, despite the fact that the two are governed by the same dynamical laws. First, the confinement of oceans to discrete basins separated by continents prevents the oceanic circulation from assuming the common east-west flow patterns adopted by most atmospheres. (Topography can cause substantial north-south deflections in an atmospheric flow, which may help explain why Earth’s atmospheric circulation involves more latitudinal excursions than that of the topography-free giant planets; nevertheless, air’s ability to flow over topography means that atmospheres, unlike oceans, are still fundamentally unbounded in the east-west direction.) The only oceanic region unhindered in the east-west direction is the Southern Ocean surrounding Antarctica, and, as might be expected, a strong east-west current, which encircles Antarctica, has formed in this region.

Second, the atmosphere is heated from below, but the ocean is heated from above. Because air is relatively transparent to sunlight, sunlight penetrates through the atmosphere and is absorbed primarily at the surface, where it heats the near-surface air at the bottom of the atmosphere. In contrast, liquid water absorbs sunlight extremely well, so that 99% of the sunlight is absorbed in the top 3% of the ocean. This means, for example, that atmospheric convection—thunderstorms—predominate at low latitudes (where abundant sunlight falls) but are rare near the poles; in contrast, convection in the oceans is totally inhibited at low latitudes and instead can occur only near the poles.

Third, much of the large-scale ocean circulation is driven not by horizontal density contrasts, as in the atmosphere (although these do play a role in the ocean), but by the frictional force of wind blowing over the ocean surface. In fact, the first simple models of ocean circulation developed by Sverdrup, Stommel, and Munk in the 1940s and 1950s, which were based solely on forcing caused by wind stress, did a reasonably good job of capturing the large-scale horizontal circulations in the ocean basins.

As in the atmosphere, the Earth’s rotation dominates the large-scale dynamics of the ocean. Horizontal Coriolis forces nearly balance pressure-gradient forces, leading to geostrophy. As in the atmosphere, this means that ocean currents flow perpendicular to horizontal pressure gradients. Rotation also means that wind stress induces currents in a rather unintuitive fashion. Because of the existence of the Coriolis force, currents do not simply form in the direction of the wind stress; instead, the three-way balance between Coriolis, pressure-gradient, and friction forces can induce currents that flow in directions distinct from the wind direction.

Averaged over time, the surface waters in most mid-latitude ocean basins exhibit a circulation consisting of a basin-filling gyre that rotates clockwise in the northern hemisphere and counterclockwise in the southern hemisphere. This circulation direction implies that the water in the western portion of the basin flows from the equator toward the pole, while the water in the eastern portion of the basin flows from the pole toward the equator. However, the flow is extremely asymmetric: The equatorward flow comprises a broad, slow motion that fills the eastern 90% of the ocean basin; in contrast, the poleward flow becomes concentrated into a narrow current (called a western boundary current) along the western edge of the ocean basin. The northward-flowing Gulf Stream off the U.S. eastern seaboard and the Kuroshio Current off Japan are two examples; these currents reach speeds up to ∼1 m s⁻¹ in a narrow zone 50-100 km wide. This extraordinary asymmetry in the ocean circulation results from the increasing strength of the Coriolis force with latitude; theoretical models show that in a hypothetical ocean where Coriolis forces are independent of latitude, the gyre circulations do not exhibit western intensification. These gyres play an important role in Earth’s climate by transporting heat from the equator toward the poles. Their clockwise (counterclockwise) rotation in the northern (southern) hemisphere helps explain why the water temperatures tend to be colder along continental west coasts than continental east coasts.

In addition to the gyres, which transport water primarily horizontally, the ocean also experiences vertical overturning. Only near the poles does the water temperature become cold enough for the surface density to exceed the deeper density. Formation of sea ice helps this process, because sea ice contains relatively little salt, so when it forms, the remaining surface water is saltier (hence denser) than average. Thus, vertical convection between the surface and deep ocean occurs only in polar regions, in particular in the Labrador Sea and near parts of Antarctica. On average, very gradual ascending motion must occur elsewhere in the ocean for mass balance to be achieved. This overturning circulation, which transports water from the surface to the deep ocean and back over ∼1000 year timescales, is called the thermohaline circulation.

The thermohaline circulation helps explain why deep-ocean waters have near-freezing temperatures worldwide: All deep-ocean water, even that in the equatorial oceans, originated at the poles and thus retains the signature of polar temperatures. Given the solar warming of low-latitude surface waters, the existence of a thermocline is thus naturally explained. However, the detailed dynamics that control the horizontal structure and depth of the thermocline are subtle and have led to major research efforts in physical oceanography over the past 4 decades.

Despite the importance of the basin-filling gyre and thermohaline circulations, much of the ocean’s kinetic energy resides in small eddy structures only 10-100 km across. The predominance of this kinetic energy at small scales results largely from the natural interaction of buoyancy forces and rotation. Fluid flows away from pressure highs toward pressure lows, but Coriolis forces short-circuit this process by deflecting the motion so that fluid flows perpendicular to the horizontal pressure gradient. The stronger the influence of rotation relative to buoyancy, the better this process is short-circuited, and hence the smaller are the resulting eddy structures. In the atmosphere, this natural length scale (called the deformation radius) is 1000-2000 km, but in the oceans it is only 10-100 km. The rings and meddies described earlier provide striking examples of oceanic eddies in this size range.

Salinity

When one swims in the ocean, the leading impression is of saltiness. The ocean’s global-mean salinity is 3.5% by mass but varies between 3.3 and 3.8% in the open oceans and can reach 4% in the Red Sea and Persian Gulf; values lower than 3.3% can occur on continental shelves near river deltas. The ocean’s salt would form a global layer 150 m thick if precipitated into solid form. Sea salt is composed of 55% chlorine, 30% sodium, 8% sulfate, 4% magnesium, and 1% calcium by mass. The ∼15% variability in the salinity of open-ocean waters occurs because evaporation and precipitation add or remove freshwater, which dilutes or concentrates the local salt abundance. However, this process cannot influence the relative proportions of elements in sea salt, which therefore remain almost constant everywhere in the oceans.

In contrast to seawater, most river and lake water is relatively fresh; for example, the salinity of Lake Michigan is ∼200 times less than that of seawater. However, freshwater lakes always have both inlets and outlets. In contrast, lakes that lack outlets—the Great Salt Lake, the Dead Sea, the Caspian Sea—are always salty. This provides a clue about processes determining saltiness.

Why is the ocean salty? When rain falls on continents, enters rivers, and flows into the oceans, many elements leach into the water from the continental rock. These elements have an extremely low abundance in the continental water, but because the ocean has no outlet (unlike a freshwater lake), these dissolved trace components can build up over time in the ocean. Ocean-seafloor chemical interactions (especially after volcanic eruptions) can also introduce dissolved ions into the oceans. However, the composition of typical river water differs drastically from that of sea salt—typical river salt contains ∼9% chlorine, 7% sodium, 12% sulfate, 5% magnesium, and 17% calcium by mass. Although sodium and chlorine comprise ∼85% of sea salt, they make up only ∼16% of typical river salt. The ratio of chlorine to calcium is 0.5 in river salt but 46 in sea salt. Furthermore, the abundance of sulfate and silica is much greater in river salt than in sea salt. These differences result largely from the fact that processes act to remove salt ions from ocean water, but the efficiency of these processes depends on the ion. For example, many forms of sea life construct shells of calcium carbonate or silica, so these biological processes remove calcium and silica from ocean water. Much magnesium and sulfate seems to be removed in ocean water-seafloor interactions. The relative inefficiency of such removal processes for sodium and chlorine apparently leads to the dominance of these ions in sea salt despite their lower proportion in river salt.

Circumstantial evidence suggests that ocean salinity has not changed substantially over the past billion years. This implies that the ocean is near a quasi-steady state where salt removal balances salt addition via rivers and seafloor-ocean interactions. These removal processes include biological sequestration in shell material, abiological seafloor-ocean water chemical interactions, and physical processes such as formation of evaporate deposits when shallow seas dry up, which has the net effect of returning the water to the world ocean while leaving salt behind on land.

Atmosphere-Ocean Interactions

Many weather and climate phenomena result from a coupled interaction between the atmosphere and ocean and would not occur if either component were removed. Two major examples are hurricanes and El Niño.

Hurricanes are strong vortices, 100-1000 km across, with warm cores and winds often up to ∼70 m s⁻¹; the temperature difference between the vortex and the surrounding air produces the pressure differences that allow strong vortex winds to form. In turn, the strong winds lead to increased evaporation off the ocean surface, which provides an enhanced supply of water vapor to fuel the thunderstorms that maintain the warm core. This enhanced evaporation from the ocean must continue throughout the hurricane’s lifetime because the thermal effects of condensation in thunderstorms inside the hurricane provides the energy that maintains the vortex against frictional losses. Thus, both the ocean and atmosphere play crucial roles. When the ocean component is removed—say, when the hurricane moves over land—the hurricane rapidly decays.

El Niño corresponds to an enhancement of ocean temperatures in the eastern equatorial Pacific at the expense of those in the western equatorial Pacific; increased rainfall in western North and South America result, and drought conditions often overtake Southeast Asia. El Niño events occur every few years and have global effects. At the crudest level, “normal” (non-El Niño) conditions correspond to westward-blowing equatorial winds that cause a thickening of the thermocline (hence producing warmer sea-surface temperatures) in the western equatorial Pacific; these warm temperatures promote evaporation, thunderstorms, and upwelling there, drawing near-surface air in from the east and thus helping to maintain the circulation. On the other hand, during El Niño, the westward-blowing trade winds break down, allowing the thicker thermocline to relax eastward toward South America, hence helping to move the warmer water eastward. Thunderstorm activity thus becomes enhanced in the eastern Pacific and reduced in the western Pacific compared to non-El Niño conditions, again helping to maintain the winds that allow those sea-surface temperatures. Although El Niño differs from a hurricane in being a hemispheric-scale long-period fluctuation rather than a local vortex, El Niño shares with hurricanes the fact that it could not exist were either the atmosphere or the ocean component prevented from interacting with the other. To successfully capture these phenomena, climate models need accurate representations of the ocean and the atmosphere and their interaction, which continues to be a challenge.

Oceans on other Worlds

The Galileo spacecraft provided evidence that subsurface liquid-water oceans exist inside the icy moons Callisto, Europa, and possibly Ganymede. The recent detection of a jet of water molecules and ice grains from the south pole of Enceladus raises the question of whether that moon has a subsurface reservoir of liquid water. Theoretical models suggest that internal oceans could exist on a wide range of other bodies, including Titan, the smaller moons of Saturn and Uranus, Pluto, and possibly even some larger Kuiper Belt objects. These oceans of course differ from Earth’s ocean in that they are ice-covered; another difference is that they must transport the geothermal heat flux of those bodies and hence are probably convective throughout. Barring exotic chemical or fluid dynamical effects, then, one expects that such oceans lack thermoclines. In many cases, these oceans may be substantially thicker than Earth’s oceans; estimates suggest that Europa’s ocean thickness lies between 50 and 150 km.

The abundant life that occurs near deep-sea vents (“black smokers”) in Earth’s oceans has led to suggestions that similar volcanic vents may help power life in Europa’s ocean. (In contrast to Europa, any oceans in Callisto and Ganymede would be underlain by high-pressure polymorphs of ice rather than silicate rock, so such silicate-water interactions would be weaker.) However, much of the biological richness of terrestrial deep-sea vents results from the fact that Earth’s oceans are relatively oxygenated; when this oxidant-rich water meets the reducing water discharged from black smokers, sharp chemical gradients result, and the resulting disequilibrium provides a rich energy source for life. Thus, despite the lack of sunlight at Earth’s ocean floor, the biological productivity of deep-sea vents results in large part from the fact that the oceans are communicating with an oxygen-rich atmosphere. If Europa’s ocean is more reducing than Earth’s ocean, then the energy source available from chemical disequilibrium may be smaller. Nevertheless, a range of possible disequilibrium reactions exist that could provide energy to drive a modest microbial biosphere on Europa.

Climate

Earth’s climate results from a wealth of interacting physical, chemical, and biological effects, and an understanding of current and ancient climates has required a multidecadal research effort by atmospheric physicists, atmospheric chemists, oceanographers, glaciologists, astronomers, geologists, and biologists. The complexity of the climate system and the interdisciplinary nature of the problem have made progress difficult, and even today many aspects remain poorly understood. “Climate” can be defined as the mean conditions of the atmosphere/ocean system—temperature, pressure, winds/currents, cloudiness, atmospheric humidity, oceanic salinity, and atmosphere/ocean chemistry in three dimensions—when time-averaged over intervals longer than that of typical weather patterns. It also refers to the distribution of sea ice, glaciers, continental lakes and streams, coastlines, and the spatial distribution of ecosystems that result.

Basic Processes—Greenhouse Effect

Earth as a whole radiates with an effective temperature of 255 K, and therefore its flux peaks in the thermal infrared part of the spectrum. This effective temperature is 30 K colder than the average temperature on the surface, and quite chilly by human standards.

What ensures a warm surface is the wavelength-dependent optical properties of the troposphere. In particular, infrared light does not pass through the troposphere as readily as visible light. The Sun radiates with an effective temperature of 5800 K and therefore its peak flux is in the visible part of the spectrum (or stated more correctly in reverse, we have evolved such that the part of the spectrum that is visible to us is centered on the peak flux from the Sun). The atmosphere reflects about 31% of this sunlight directly back to space, and the rest is absorbed or transmitted to the ground. The sunlight that reaches the ground is absorbed and then reradiated at infrared wavelengths. Water vapor (H₂O) and carbon dioxide (CO₂), the two primary greenhouse gases, absorb some of this upward infrared radiation and then emit it in both the upward and downward directions, leading to an increase in the surface temperature to achieve balance. This is the greenhouse effect. Contrary to popular claims, the elevation of surface temperature by the greenhouse effect is not a situation where “the heat cannot get out.” Instead, the heat must get out, and to do so in the presence of the blanketing effect of greenhouse gases requires an elevation of surface temperatures.

The greenhouse effect plays an enormous role in the climate system. A planet without a greenhouse effect, but otherwise identical to Earth, would have a global-mean surface temperature 17°C below freezing. The oceans would be mostly or completely frozen, and it is doubtful whether life would exist on Earth. We owe thanks to the greenhouse effect for Earth’s temperate climate, liquid oceans, and abundant life.

Water vapor accounts for between one third and two thirds of the greenhouse effect on Earth (depending on how the accounting is performed), with the balance resulting from CO₂, methane, and other trace gases. Steady increases in carbon dioxide due to human activity seem to be causing the well-documented increase in global surface temperature over the past ∼100 years. On Mars, the primary atmospheric constituent is CO₂, which together with atmospheric dust causes a modest 5 K greenhouse effect. Venus has a much denser CO₂ atmosphere, which, along with atmospheric sulfuric acid and sulfur dioxide, absorbs essentially all the infrared radiation emitted by the surface, causing an impressive 500 K rise in the surface temperature. Interestingly, if all the carbon held in Earth’s carbonate rocks were liberated into the atmosphere, Earth’s greenhouse effect would approach that on Venus.

Basic Processes—Feedbacks

The Earth’s climate evolves in response to volcanic eruptions, solar variability, oscillations in Earth’s orbit, and changes in internal conditions such as the concentration of greenhouse gases. The Earth’s response to these perturbations is highly nonlinear and is determined by feedbacks in the climate system. Positive feedbacks amplify a perturbation and, under some circumstances, can induce a runaway process where the climate shifts abruptly to a completely different state. In contrast, negative feedbacks reduce the effect of a perturbation and thereby help maintain the climate in its current state. Some of the more important feedbacks are as follows.

Thermal feedback: Increases in the upper tropospheric temperature lead to enhanced radiation to space, tending to cool the Earth. Decreases in the upper tropospheric temperature cause decreased radiation to space, causing warming. This is a negative feedback.

Ice-albedo feedback: Ice caps and glaciers reflect visible light easily, so the Earth’s brightness (albedo) increases with an increasing distribution of ice and snow. Thus, a more ice-rich Earth absorbs less sunlight, promoting colder conditions and growth of even more ice. Conversely, melting of glaciers causes Earth to absorb more sunlight, promoting warmer conditions and even less ice. This is a positive feedback.

Water-vapor feedback: Warmer surface temperatures allow increased evaporation of water vapor from the ocean surface, increasing the atmosphere’s absolute humidity. Because water vapor is a greenhouse gas, it promotes an increase in the strength of the greenhouse effect and hence even warmer conditions. Cooler conditions inhibit evaporation, lessen the greenhouse effect, and cause additional cooling. This is a positive feedback.

Cloud feedback: Changes in climate can cause changes in the spatial distribution, heights, and properties of clouds. Greater cloud coverage means a brighter Earth (higher albedo), leading to less sunlight absorption. Higher altitude clouds have colder tops that radiate heat to space less well, promoting a warmer Earth. For a given mass of condensed water in a cloud, clouds with smaller particles reflect light better, promoting a cooler Earth. Unfortunately, for a specified climate perturbation (e.g., increasing the CO₂ concentration), the extent to which the coverage, heights, and properties of clouds will change remains unclear. Thus, not only the magnitude but even the sign (positive or negative) of this feedback remains unknown.

The sum of these and other feedbacks determine how Earth’s climate evolved during past epochs and how Earth will respond to current human activities such as emissions of CO₂. Much of the uncertainty in current climate projections results from uncertainty in these feedbacks. A related concept is that of thresholds, where the climate undergoes an abrupt shift in response to a gradual change. For example, Europe enjoys temperate conditions despite its high latitude in part because of heat transported poleward by the Gulf Stream. Some climate models have suggested that increases in CO₂ due to human activities could suddenly shift the ocean circulation in the North Atlantic into a regime that transports heat less efficiently, which could cause widespread cooling in Europe (although this might be overwhelmed by the expected global warming that will occur over the next century). The rapidity with which ice ages ended also suggests that major reorganizations of the ocean/atmosphere circulation occurred during those times. Although thresholds play a crucial role in past and possibly future climate change, they are notoriously difficult to predict because they involve subtle nonlinear interactions.

Recent Times

A wide range of evidence demonstrates that Earth’s global-mean surface temperature rose by about 0.6°C between 1900 and 2000. Since the mid-1970s, the global-mean rate of temperature increase has been ∼0.17°C per decade (with a greater rate of warming over land than ocean). As of 2006, 20 of the hottest years measured since good instrumental records started in ∼1860 have occurred within the past 25 years, and the past 25 years has been the warmest 25 year period of the past 1000 years. There is widespread consensus among climate experts that the observed warming since ∼1950 has been caused primarily by the release of CO₂ due to human activities, primarily the burning of oil, coal, natural gas, and forests: The greater CO₂ concentration has increased the strength of the greenhouse effect, modified by the feedbacks discussed in Section 5.2. Before the Industrial Revolution, the CO₂ concentration was ∼280 ppm (i.e., a mole fraction of 2.8 × 10⁻⁴), and in 2006 the CO₂ concentration was 380 ppm—a 36% increase. Interestingly, only half of the CO₂ released by human activities each year remains in the atmosphere; the remainder is currently absorbed by the biosphere and especially the oceans. The increase in mean surface temperature has been accompanied by numerous other climate changes, including retreat of glaciers worldwide, thawing of polar permafrost, early arrivals of spring, late arrivals of autumn, changes in the Arctic sea-ice thickness, approximately 0.1-0.2 m of sea-level increase since 1900, and various effects on natural ecosystems. These changes are expected to accelerate in the 21st century.

Ice Ages

The repeated occurrence of ice ages, separated by warmer interglacial periods, dominates Earth’s climatic record of the past 2 million years. During an ice age, multi-kilometer-thick ice sheets grow to cover much of the high-latitude land area, particularly in North America and Europe; most or all of these ice sheets melt during the interglacial periods (however, ice sheets on Antarctica and Greenland have resisted melting during most interglacials, and these two ice sheets still exist today). The sea level varies by up to 120 m between glacial and interglacial periods, causing migration of coastlines by hundreds of kilometers in some regions. The time history of temperature, ice volume, and other variables can be studied using stable isotopes of carbon, hydrogen, and oxygen as recorded in glacial ice, deep-sea sediments, and land-based records such as cave calcite and organic material. This record shows that glacial/interglacial cycles over the past 800,000 years have a predominant period of ∼100,000 years. During this cycle, glaciers gradually increase in volume (and air temperature gradually decreases) over most of the 100,000 year period; the glaciers then melt, and the temperature increases over a relatively short ∼5000 year interval. The cycle is thus extremely asymmetric and resembles a saw-tooth curve rather than a sinusoid. The last ice age peaked 18,000 years ago and ended by 10,000 years ago; the modern climate corresponds to an interglacial period. Analysis of ancient air trapped in air bubbles inside the Antarctic and Greenland ice sheets shows that the atmospheric CO₂ concentration is low during ice ages—typically about 200 ppm—and rises to ∼280 ppm during the intervening interglacial periods.

Ice ages seem to result from changes in the strength of sunlight caused by periodic variations in Earth’s orbit, magnified by several of the feedbacks discussed in Section 5.2. A power spectrum of the time series shows that temperature, ice volume, and CO₂ vary predominantly on periods of 100, 41, 23, and 19 thousand years (ka; the summation of sinusoids at each of these periods leads to the saw-tooth patterns). Interestingly, these periods match the periods over which northern hemisphere sunlight varies due to orbital oscillations. The Earth’s orbital eccentricity oscillates on periods of 100 ka, the orbital obliquity (the tilt of Earth’s rotation axis) oscillates on a period of 41 ka, and the Earth’s rotation axis precesses on periods of 19 and 23 ka. These variables affect the difference in sunlight received at Earth between winter and summer and between the equator and pole. In turn, these sunlight variations determine the extent to which snowpack accumulates in high northern latitudes during winter, and the extent to which this snowpack resists melting during summer; glaciers build up when snow that falls during winter cannot melt the following summer. The idea that these orbital variations cause ice ages has become known as the Milankovitch theory of ice ages.

By themselves, however, orbital variations are only part of the story. Sunlight variations due to the 100 ka eccentricity variations are much weaker than sunlight variations due to the 41, 23, and 19 ka obliquity and precession variations. Thus, if the orbit-induced sunlight variations translated directly into temperature and ice variations, ice ages would be dominated by the 41, 23, and 19 ka periods, but instead, the 100 ka period dominates. This means that some nonlinearity in the climate system amplifies the climatic response at 100 ka much better than at the shorter periods. Furthermore, the observed oscillations in CO₂ between glacial and interglacial periods indicates that ice ages are able to occur partly because the greenhouse effect is weak during ice ages but strong during interglacial periods. Most likely, atmospheric CO₂ becomes dissolved in ocean water during ice ages, allowing the atmospheric CO₂ levels to decrease; the ocean then rejects this CO₂ at the end of the ice age, increasing its atmospheric concentration. Recent analyses of Antarctic ice cores show that, at the end of an ice age, temperature rise precedes CO₂ rise in Antarctica by about 800 years, indicating that CO₂ variation is an amplifier rather than a trigger of ice-age termination. Interestingly, however, both of these events precede the initiation of deglaciation in the northern hemisphere. These observations suggest that the end of an ice age is first triggered by a warming event in the Antarctic region; this initiates the process of CO₂ rejection from the oceans to the atmosphere, and the resulting increase in the greenhouse effect, which is global, then allows deglaciation to commence across the rest of the planet. The ice-albedo and water-vapor feedbacks (Section 5.2) help amplify the transition. However, many details, including the exact mechanism that allows CO₂ to oscillate between the ocean and atmosphere, remain to be worked out.

Figure 4 shows how the increase in CO₂ caused by human activities compares to the natural variability in the past. The saw-toothed variations in CO₂ between 200 and 280 ppm over 100,000 year periods indicate the ice-age/interglacial cycles, and the vertical spike in CO₂ (from 280 to 380 ppm) shows the human-induced increase. The current CO₂ concentration far exceeds that at any previous time over the past 420,000 years, and is probably the greatest CO₂ level the Earth has seen since 20 million years ago. The fact that CO₂ rises by 30-40% at the end of an ice age indicates that very large magnitude climate changes can accompany modest CO₂ variations; it is noteworthy that human activities have so far increased CO₂ by an additional 36% beyond preindustrial values. The relationship between CO₂ and global temperature during ice ages may differ from the relationship these quantities will take over the next century of global warming; however, it is virtually certain that additional CO₂ will cause global temperature increases and widespread climate changes. Current economic and climate projections indicate that, because of continued fossil fuel burning, the atmospheric CO₂ will reach 500-1000 ppm by the year 2100 unless drastic measures are adopted to reduce fossil fuel use.

Volatile Inventories of Terrestrial Planets

Venus, Earth, and Mars have present-day atmospheres that are intriguingly different. The atmospheres of Venus and Mars are both primarily CO₂, but they represent two extreme fates in atmospheric evolution: Venus has a dense and hot atmosphere, whereas Mars has a thin and cold atmosphere. It is reasonable to ask whether Earth is ultimately headed toward one or the other of these fates, and whether these three atmospheres have always been so different.

The history of volatiles on the terrestrial planets includes their origin, their interactions with refractory (nonvolatile) material, and their rates of escape into space. During the initial accretion and formation of the terrestrial planets, it is thought that most or all of the original water reacted strongly with the iron to form iron oxides and hydrogen gas, with the hydrogen gas subsequently escaping to space. Until the iron cores in the planets were completely formed and this mechanism was shut down, the outflow of hydrogen probably took much of the other solar-abundance volatile material with it. Thus, one likely possibility is that the present-day atmospheres of Venus, Earth, and Mars are not primordial, but have been formed by outgassing and by cometary impacts that have taken place since the end of core formation.

The initial inventory of water that each terrestrial planet had at its formation is a debated question. One school of thought is that Venus formed in an unusually dry state compared with Earth and Mars; another is that each terrestrial planet must have started out with about the same amount of water per unit mass. The argument for an initially dry Venus is that water-bearing minerals would not condense in the high-temperature regions of the protoplanetary nebula inside of about 1 AU. Proponents of the second school of thought argue that gravitational scattering caused the terrestrial planets to form out of materials that originated over the whole range of terrestrial-planet orbits, and therefore that the original water inventories for Venus, Earth, and Mars should be similar.

An important observable that bears on the question of original water is the enrichment of deuterium (D) relative to hydrogen. A measurement of the D/H ratio yields a constraint on the amount of hydrogen that has escaped from a planet. For the D/H ratio to be useful, one needs to estimate the relative importance of the different hydrogen escape mechanisms and the original D/H ratio for the planet. In addition, one needs an idea of the hydrogen sources available to a planet after its formation, such as cometary impacts. The initial value of D/H for a planet is not an easy quantity to determine. A value of 0.2 × 10⁻⁴ has been put forward for the protoplanetary nebula, which is within a factor of 2 or so for the present-day values of D/H inferred for Jupiter and Saturn. However, the D/H ratio in Standard Mean Ocean Water (SMOW, a standard reference for isotopic analysis) on Earth is 1.6 × 10⁻⁴, which is also about the D/H ratio in hydrated minerals in meteorites, and is larger by a factor of 8 over the previously mentioned value. At the extreme end, some organic molecules in carbonaceous chondrites have shown D/H ratios as high as 20 × 10⁻⁴. The enrichment found in terrestrial planets and most meteorites over the protoplanetary nebula value could be the result of exotic high-D/H material deposited on the terrestrial planets, or it could be the result of massive hydrogen escape from the planets early in their lifetimes through the hydro-dynamic blow off mechanism (which is the same mechanism that currently drives the solar wind off the Sun).

Life in the Atmosphere-Ocean System

Interplanetary Spacecraft Evidence for Life

An ambitious but ever-present goal in astronomy is to detect or rule out life in other solar systems, and in planetary science that goal is to detect or rule out life in our own solar system apart from Earth. Water in its liquid phase is one of the few requirements shared by all life on Earth, and so the hunt for life is focused on the search for liquid water. We know that Mars had running water on its surface at some point in its history because we can see fluvial channels in high-resolution images, and because the Mars rovers Spirit and Opportunity have discovered aqueous geochemistry on the ground; there is even some evidence suggesting present-day seepage in recent orbiter images. Farther out in the solar system, we know that Europa, a satellite of Jupiter, has a smooth icy surface with cracks and flow features that resemble Earth’s polar ice fields and suggest a liquid-water interior, while its larger sibling, Ganymede, exhibits a conductive reaction to Jupiter’s magnetic field that is most easily explained by a salty liquid-water interior.

However, to date we have no direct evidence for extraterrestrial life. This includes data from landers on Venus, Mars, and the Moon, and flyby encounters with 8 planets, a handful of asteroids, a comet (Halley in 1986), and over 60 moons. Are the interplanetary spacecraft we have sent out capable of fulfilling the goal of detecting life? This question has been tested by analyzing data from the Galileo spacecraft’s two flyby encounters with Earth, which, along with a flyby encounter with Venus, were used by the spacecraft’s navigation team to provide gravity assists to send Galileo to Jupiter. The idea was to compare ground-truth information to what we can learn solely from Galileo.

Galileo‘s first Earth encounter occurred on December 8, 1990, with closest approach 960 km above the Caribbean Sea; its second Earth encounter occurred on December 8, 1992, with closest approach 302 km above the South Atlantic. A total of almost 6000 images were taken of Earth by Galileo‘s camera system. Figure 5 shows the Earth-Moon system as seen by Galileo. Notice that the Moon is significantly darker than Earth. The spacecraft’s instruments were designed and optimized for Jupiter; nevertheless, they made several important observations that point to life on Earth. These strengthen the null results encountered elsewhere in the solar system. The evidence for life on Earth includes complex radio emissions, nonmineral surface pigmentation, disequilibrium atmospheric chemistry, and large oceans.

Radio Emissions

The only clear evidence obtained by Galileo for intelligent life on Earth was unusual radio emissions. Several natural radio emissions were detected, none of which were unusual, including solar radio bursts, auroral kilometric radiation, and narrowband electrostatic oscillations excited by thermal fluctuations in Earth’s ionospheric plasma. The first unusual radio emissions were detected at 1800 UT and extended through 2025 UT, just before closest approach. These were detected by the plasma wave spectrometer (PWS) on the nightside, in-bound pass, but not on the day side, out-bound pass. The signal strength increased rapidly as Earth was approached, implying that Earth itself was the source of the emissions. The fact that the signals died off on the day side suggests that they were cut off by the day side ionosphere, which means we can place the source below the ionosphere.

The unusual signals were narrowband emissions that occurred in only a few distinct channels and had average frequencies that remained stable for hours. Naturally occurring radio emissions nearly always drift in frequency, but these emissions were steady. The individual components had complicated modulations in their amplitude that have never been detected in naturally occurring emissions. The simplest explanation is that these signals were transmitting information, which implies that there is advanced technological life on Earth. In fact, the radio, radar, and television transmissions that have been emanating from Earth over the last century result in a nonthermal radio emission spectrum that broadcasts our presence out to interstellar distances.

Surface Features

During its first encounter with Earth, the highest-resolution mapping of the surface by Galileo‘s Solid-State Imaging System (SSI) covered Australia and Antarctica with 1-2 km resolution. No usable images were obtained from Earth’s night side on the first encounter. The second encounter netted the highest resolution images overall of Earth by Galileo, 0.3-0.5 km per pixel, covering parts of Chile, Peru, and Bolivia. The map of Australia from the first encounter includes 2.3% of Earth’s total surface area, but shows no geometric patterns that might indicate an advanced civilization. In the second encounter, both the cities of Melbourne and Adelaide were photographed, and yet no geometric evidence is visible because the image resolution is only 2 km. The map of Antarctica, 4% of Earth’s surface, reveals nearly complete ice cover and no signs of life. Only one image, taken of southeastern Australia during the second encounter, shows east-west and north-south markings that would raise suspicions of intelligent activity. The markings in fact were caused by boundaries between wilderness areas, grazing lands, and the border between South Australia and Victoria. Studies have shown that it takes nearly complete mapping of the surface at 0.1-km resolution to obtain convincing photographic evidence of an advanced civilization on Earth, such as roads, buildings, and evidence of agriculture.

On the other hand, many features are visible in the Galileo images that have not been seen on any other body in the solar system. The SSI camera took images in six different wavelength channels. A natural-color view of Earth was constructed using the red, green, and violet filters, which correspond to wavelengths of 0.670, 0.558, and 0.407 μm, respectively. The images reveal that Earth’s surface is covered by enormous blue expanses that specularly reflect sunlight, and end in distinct coastlines, which are both easiest to explain if the surface is liquid. This implies that much of the planet is covered with oceans. The land surfaces show strong color contrasts that range from light brown to dark green.

The SSI camera has particular narrowband infrared filters that have never been used to photograph Earth before, and so they yielded new information for geological, biological, and meteorological investigations. The infrared filters allow the discrimination of H₂O in its solid, liquid, and gaseous forms; for example, clouds and surface snow can be distinguished spectroscopically with the 1 μm filter. False-color images made by combining the 1 μm channel with the red and green channels reveal that Antarctica strongly absorbs 1 μm light, establishing that it is covered by water ice. In contrast, large regions of land strongly reflect 1 μm without strongly reflecting visible colors, which conflicts with our experience from other planetary surfaces and is not typical of igneous or sedimentary rocks or soil. Spectra made with the 0.73 and 0.76 μm channels reveal several land areas that strongly absorb red light, which again is not consistent with rocks or soil. The simplest explanation is that some nonmineral pigment that efficiently absorbs red light has proliferated over the planet’s surface. It is hard to say for certain if an interstellar explorer would realize that this is a biological mechanism for gathering energy from sunlight, probably so, but certainly we would recognize it on another planet as the signature of plant life. We know from ground truth that these unusual observations are caused by the green pigments chlorophyll a (C₅₅H₇₂MgN₄O₅) and chlorophyll b (C₅₅H₇₀MgN₄O₆), which are used by plants for photosynthesis. No other body in the solar system has the green and blue colorations seen on Earth.

Oxygen and Methane

Galileo’s Near-Infrared Mapping Spectrometer (NIMS) detected the presence of molecular oxygen (O₂) in Earth’s atmosphere with a volume mixing ratio of 0.19 ± 0.05. Therefore, we know that the atmosphere is strongly oxidizing. (It is interesting to note that Earth is the only planet in the solar system where one can light a fire.) In light of this, it is significant that NIMS also detected methane (CH₄) with a volume mixing ratio of 3 ± 1.5 × 10⁻⁶. Because CH₄ oxidizes rapidly into H₂O and CO₂, if thermodynamical equilibrium holds, then there should be no detectable CH₄ in Earth’s atmosphere. The discrepancy between observations and the thermodynamic equilibrium hypothesis, which works well on other planets (e.g., Venus), is an extreme 140 orders of magnitude. This fact provides evidence that Earth has biological activity and that it is based on organic chemistry. We know from ground truth that Earth’s atmospheric methane is biological in origin, with about half of it coming from nonhuman activity like methane bacteria and the other half coming from human activity like growing rice, burning fossil fuels, and keeping livestock. NIMS also detected a large excess of nitrous oxide (N₂O) that is most easily explained by biological activity, which we know from ground truth comes from nitrogen-fixing bacteria and algae.

The conclusion is that the interplanetary spacecraft we have sent out to explore our solar system are capable of detecting life on planets or satellites, both the intelligent and primitive varieties, if it exists in abundance on the surface. On the other hand, if there is life on a planet or satellite that does not have a strong signature on the surface, as would probably be the case if Europa or Ganymede harbor life, then a flyby mission may not be adequate to decide the question. With regard to abundant surface life, we have a positive result for Earth and a negative result for every other body in the solar system.

Conclusions

Viewing Earth as a planet is the most important change of consciousness that has emerged from the space age. Detailed exploration of the solar system has revealed its beauty, but it has also shown that the home planet has no special immunity to the powerful forces that continue to shape the solar system. The ability to remotely sense Earth’s dynamic atmosphere, oceans, biosphere, and geology has grown up alongside our ever-expanding ability to explore distant planetary bodies. Everything we have learned about other planets influences how we view Earth. Comparative planetology has proven in practice to be a powerful tool for studying Earth’s atmosphere and oceans. The lion’s share of understanding still awaits us, and in its quest we continue to be pulled outward.