Tema

 5  
Estructura   de   la   Tierra.   Origen   y   evolución   de   la  
Hidrosfera.   Distribución   del   agua   en   la   Hidrosfera:   el  
ciclo   hidrológico   externo.   Dinámica   de   la   Hidrosfera.  
Los  ciclos  hidrológicos  endógenos.  Origen  del  agua  de  
la   Tierra.   El   agua   en   el   Universo.   El   agua   y   el   Cambio  
ClimáCco  (calentamiento  global).  
    Procedencia  del  agua  de  la  Tierra  
      Ciclo  hidrológico  externo  

•  El ciclo hidrológico externo es la síntesis de la

circulación del agua disponible sobre la superficie de la
Tierra
•  Los procesos implicados en el ciclo hidrológico son
•  Precipitación
•  Evaporación
•  Infiltración
•  Escorrentía
•  Transpiración
      El  ciclo  hidrológico  
    Procedencia  del  agua  de  la  Tierra  
      el  agua  de  escorren6a  
•  Comienza la arroyada en manto
•  La capacidad de infiltración está controlada por
– Intensidad y duración de la lluvia
– Condiciones de humedad previa del suelo
– Textura del suelo
– Inclinación de la vertiente
– Tipo de cubierta vegetal
•  La arroyada en manto comienza con pequeños
canales llamados rills o regueros  
 El  agua  de  escorren6a  
•  Cursos de agua
•  Los dos tipos de cursos de agua están determinados
principalmente por la velocidad
–  Flujo laminar
–  Flujo turbulento
•  Factores que determinan la velocidad
–  Gradiente, o pendiente
–  Características del canal lo que incluye dimensiones, forma y
rugosidad.
 El  agua  de  escorren6a  
•  Cursos fluviales
•  Factores que determinan la velocidad
–  Caudal – es el volumen de agua que pasa por un punto dado en un
intervalo de tiempo determinado
– Velocidad
– Descarga
– Dimensiones del canal

•  Cambios desde nacimiento a desembocadura

•  Perfil
–  Sección longitudinal de un curso fluvial
–  Desde el nacimiento a la desembocadura.
      Perfil  longitudinal  de  un  curso  
fluvial  
 El  agua  de  escorren6a  
•  Cambios de nacimiento a desembocadura
• Factores que disminuyen aguas abajo
– Gradiente
– Rugosidad del canal
•  Nivel de base y cursos fluviales graduales
• Nivel de base es el punto mas bajo hasta el
que el curso fluvial puede erosionar.
 El  agua  de  escorren6a  
•  Nivel de base y cursos de agua gradados
•  Dos grandes tipos de nivel de base
– Final (nivel del mar) varía con clima, tectónica, etc.
– Local o temporal
•  Los cambios en las condiciones producen un reajuste
en el comportamiento del curso de agua.
– El ascenso del nivel de base produce deposición
– El descenso del nivel de base produce erosión
Ajuste  del  nivel  de  base  al  cambio  de  
condiciones  
Copyright  ©  Keith  Richards  2002  
Copyright  ©  Ann  Di8mer  2002  
Copyright  ©  Norm  Ca8o  2002  
Erosión  y  deposición  a  lo  largo  de  un  curso  meandriforme  
    Codo  de  meandro  en  el  Río    
   Colorado  
 El  agua  de  escorren6a  
•  Redes de drenaje
• La superficie donde se colecta el agua con los
canales por la que ésta circula hasta llegar al
rio principal es la cuenca de drenaje
• La línea imaginaria (la derivada de valor
cero de la curva señalada por el relieve) que
separa una cuenca de otra es la divisoria
Cuenca  de  drenaje  
 El  agua  de  escorren6a  
•  Tipos de red de drenaje
• Disposición interconectada de la red de
drenaje en una zona concreta
• Tipos mas frecuentes de red de drenaje
– Dendrítica
– Radial
– Rectangular
– Emparrillado
Tipos  de  redes  de  drenaje  

Figure 16.32
 El  agua  de  escorren6a  
•  Erosión remontante y captura fluvial
•  Un curso de agua puede aumentar su longitud por:
– Construcción de un delta
– Erosión remontante
•  La erosión remontante puede dar lugar a capturas –
que produce el desvío del drenaje de un río a otro.
Con ello también se incrementa la longitud del río
que captura al otro.
Una  cascada  es  un  ejemplo  de  nivel  de  
base  local  
El  caso  del  rio  Xallas  en  Ézaro  
 Volumen  de  agua  de  la  Hidrosfera  

•  Toda  el  agua  contenida  en  los  océanos  de  la  

Tierra
 cabe
 en
 una
 esfera
 de
 138
 km
 de
 diámetro.
 
•  En
 las
 zonas
 conCnentales
 aun
 hay
 menos.
 
 NEWS & ANALYSIS
W AT E R R E S O U R C E S
Kenyan Find Heralds New
Era in Water Prospecting
Dozens of meters below the parched, tree-
less landscape of northwestern Kenya lies a
hidden treasure: vast supplies of water. Last
week, the leaders of a new hydrogeologi-
cal survey, part of a U.N. program aimed at
quenching the thirst of drought-stricken
countries, announced that they have discov-
ered five deep aquifer systems. They hold
enough water to boost Kenya’s known water
reserves by 17%.
Properly managed, the find could be a
game-changer for Kenyans, who have suf-
fered through a series of severe droughts in
recent decades. But it also marks a triumph
for prospectors who are melding traditional
geology and cutting-edge digital tools to
reveal hidden water supplies in seemingly
unpromising places. Such approaches rep-
resent “a revolution” in groundwater detec-
tion, says geologist Alain Gachet, head of
Radar Technologies International (RTI), a
French company that assisted with the work
in Kenya. “We’ve just started to initiate the
discovery of these deep aquifers.”
The hunt for Kenya’s water began last
year, after seasonal rains failed in the Horn
of Africa. Crops and livestock died, prompt-
ing a food crisis that affected millions in
Kenya, Somalia, South Sudan, and Ethio-
pia. In response, the United Nations Edu-
cational, Scientific and Cultural Organiza-
tion (UNESCO) launched the Groundwater
Resources Investigation for Drought Miti-
gation in Africa Programme to find ground-
water supplies that could offer solutions.
In the past, such searches have been some-
what haphazard affairs—a dash of savvy sci-
ence mixed with educated guesses and brute
force exploratory drilling. But RTI and other
groups are now using powerful computer pro-
grams to merge refined geological knowledge
with a flood of data from ground- and space-
based sensors. In recent years, they’ve found
signs of groundwater reserves in Sudan’s
Darfur region and elsewhere.
In Kenya, UNESCO hired RTI to apply
the new tools to a survey of Turkana County,
which covers 36,000 square kilometers in
the country’s arid northwestern corner. The
CREDIT: RTI
county is bordered to the east by Lake Tur-
kana, a water-filled depression that is part of
www.sciencemag.org
NEWS & ANALYSIS
the East African Rift, a region of stretched
and fractured crust. Between the fractures,
sunken blocks of land called grabens can
form sediment-filled troughs as much as
4 kilometers deep.
The troughs were a prime target, Gachet
says, because they are potential traps for
ground water. To identify them, RTI collected
Bull’s-eye! Geologist Alain Gachet (right) celebrates striking water in Kenya’s Lotikipi Basin Aquifer, one of
the five deep aquifers he helped discover.
traditional geologic maps, hydrological data,
satellite images, and gravity and seismic sur-
vey data. Then they added commercial sat-
ellite radar images that can detect soil mos-
iture. However, surface objects—such as
rocks and villages—can obscure the images.
So RTI developed its proprietary image-
processing technology, WATEX, which
“erases” the obstacles, revealing the hints
of moisture that can betray a hidden aquifer.
“Like a radio tuner, you can fine-tune the final
result into a WATEX map,” Gachet says.
The data suggested “that there was plenty
of water under the desert,” Gachet says. In
the end, WATEX helped identify five likely
aquifer systems, each more than 100 meters
below the surface, containing water reserves
totaling at least 250 billion cubic meters. So
far, drilling has confirmed two of the aquifers,
known as the Lotikipi and Lodwar basins.
Lotikipi, the largest, is roughly the size of
Rhode Island.
SCIENCE VOL 341 20 SEPTEMBER 2013
Published by AAAS
The discoveries are a “critical scientific
breakthrough,” Judi Wakhungu, Kenya’s
Cabinet secretary of the Ministry of Environ-
ment, Water and Natural Resources, said at
a UNESCO meeting in Nairobi on 11 Sep-
tember. “This newly found wealth … opens a
door to a more prosperous future for the peo-
ple of Turkana and the nation as a whole.”
The Kenyan find demonstrates the value
of maintaining a diverse fleet of Earth-
sensing satellites, which often face fund-
ing challenges, says hydroclimatologist
Benjamin Zaitchik of Johns Hopkins Univer-
sity in Baltimore, Maryland. It is also “a pow-
Downloaded from www.sciencemag.org on September 19, 2013
erful example of a private company making
use of government-supported satellite capa-
bilities to support U.N. humanitarian goals.”
Tapping the water is the next hurdle.
“Drilling equipment, to drill water below
80 meters, is a problem in Kenya,” Gachet
says. It will probably fall to foreign gov-
ernments and aid organizations to fund the
needed infrastructure.
Another looming challenge is managing
the resource. Studies suggest the aquifers are
replenished by rainfall in the highlands of
Kenya and Uganda. But the recharge rates are
slow, meaning the two nations may need to
cooperate to avoid overuse.
Deep groundwater resources may exist
in many other regions with similar geology,
Gachet says. “I have great certainty we’ll find
a deep buried aquifer in South Sudan,” he
says, where surface waters are often contami-
nated with parasites. “We can create change
all along the rift system, from Syria down to
Mozambique. Our playground is very big.”
–CAROLYN GRAMLING
1327
 Salida
 de
 agua
 en
 los
 hot
 spot
 

•  Los  hot  spot  transportan  materiales  desde  la  

base  de  la  mesosfera  hasta  la  superficie  de  la  
Tierra  
Fig.  7.13  
El  agua  en  el  Universo  
F. Maggi, C. Pallud / Journal of Hydrology 416–417 (2012) 12–18 15

Fig. 1. A conceptual representation of the components of the hydrological cycle of water in the universe.
Edición en galego Rss Edición Impresa: Servicios | Diario en PDF | Monográficos

lavoz.es tienda empresas

buscar

PORTADA GALICIA DEPORTES SOCIEDAD DINERO ESPAÑA MUNDO OPINIÓN PARTICIPA BLOGS OCIO Y CULTURA SERVICIOS CANALES TIENDA INMO MOTOR EMPL
Gente Tendencias Tecnología Canal Si La Guía TV Hoy mujer XLSemanal

ESTUDIO

Los océanos deben su origen a la

colisión de asteroides gigantes, según
científicos franceses
Efe 29/10/2009

Valoración (1 votos) Gracias.

Los océanos de la Tierra no deben su origen a la actividad volcánica de nuestro

Noticias + vistas Vídeos Álbumes
planeta, como se creía, sino a la colisión de asteroides gigantes hace millones
de años, según un estudio del Centro Nacional de la Investigación Científica de 1. Amy Winehouse sorprende con su nuevo
un patinazo
Francia (CNRS), recogido en la revista Nature. El científico Francis Albarède 2. Misa con muerto en Cedeira
sostiene que «asteroides gigantes cubiertos de hielo» chocaron contra la Tierra 3. La condesa de Fenosa cedió a la Fundació
entre 80 y 130 millones de años después de la formación del planeta. Según su todo su legado
4. El Real Madrid saca a relucir sus vergüen
tesis, la colisión de estos cuerpos helados extraterrestres provocó además un ridículo histórico
movimiento de las placas tectónicas que favoreció la creación de los 5. Libertad sin fianza pero con cargos para
de Castro de Rei
continentes y los océanos, condición probablemente necesaria para la aparición
de la vida. Este fenómeno sería también responsable de la formación de la Canal Compras - Ver Más
atmósfera, hasta ahora atribuida a los «vapores emitidos durante el amanecer
Botas y petos Botosvalverd
de nuestro planeta». Esos impactos pudieron también repetirse en Marte, pero
Compra desde tu casa, sin es
el agua se habría secado antes de llegar a penetrar en profundidad. Consultar

Compartir ¿Cómo compartir?

Sciphone G2 Wifi. Móviles d
LETTER doi:10.1038/nature10519

Ocean-like water in the Jupiter-family comet

103P/Hartley 2
Paul Hartogh1, Dariusz C. Lis2, Dominique Bockelée-Morvan3, Miguel de Val-Borro1, Nicolas Biver3, Michael Küppers4,
Martin Emprechtinger2, Edwin A. Bergin5, Jacques Crovisier3, Miriam Rengel1, Raphael Moreno3, Slawomira Szutowicz6
& Geoffrey A. Blake2

For decades, the source of Earth’s volatiles, especially water with a Our measured D/H value is substantially larger than that which
deuterium-to-hydrogen ratio (D/H) of (1.558 6 0.001) 3 1024, has characterized the young Sun (4.5 Gyr ago; the protosolar ratio),
been a subject of debate. The similarity of Earth’s bulk composition believed to be about 2.1 3 1025, which in turn is slightly higher than
to that of meteorites known as enstatite chondrites1 suggests a dry the value found in the local interstellar medium today (1.6 3 1025) and
proto-Earth2 with subsequent delivery of volatiles3 by local accre-
tion4 or impacts of asteroids or comets5,6. Previous measurements
in six comets from the Oort cloud yielded a mean D/H ratio of 15 a HDO (110–101)
(2.96 6 0.25) 3 1024. The D/H value in carbonaceous chondrites,
(1.4 6 0.1) 3 1024, together with dynamical simulations, led to
models in which asteroids were the main source of Earth’s water7, 10
with #10 per cent being delivered by comets. Here we report that
the D/H ratio in the Jupiter-family comet 103P/Hartley 2, which 5
originated in the Kuiper belt, is (1.61 6 0.24) 3 1024. This result

Main beam brightness temperature (mk)

substantially expands the reservoir of Earth ocean-like water to
include some comets, and is consistent with the emerging picture 0
of a complex dynamical evolution of the early Solar System8,9.
On 17 November 2010, using the Herschel Space Observatory, we
determined the D/H ratio in a comet from a reservoir other than the –5
Oort cloud—103P/Hartley 2. Such Jupiter-family comets are believed
to originate from the Kuiper belt, which exists beyond the orbits of the b H218O (110–101)
giant planets at radii between 30 and 50 astronomical units10 (1 AU is
the average Earth–Sun distance). In contrast, Oort-cloud comets are 100
theorized to have originated from radii near the gas giants and to have
been subsequently ejected to the Oort cloud (.5,000 AU)11. The
Herschel measurement therefore traces the water D/H ratio in a new
population of water-ice-rich bodies in the Solar System that are a 50
potential source of water on the Earth.
To obtain an accurate determination of the D/H ratio in water, we
carried out simultaneous observations of optically thin isotopic var- 0
iants of water, specifically HDO and H218O (Fig. 1), as part of our Solar
System observing programme12. This was critical for comet 103P/
Hartley 2, whose activity and water outgassing rates exhibited signifi- –5 0 5
cant short-term variations13. We used state-of-the-art excitation models Velocity (km s–1)
to determine the HDO and H218O beam integrated column densities
Figure 1 | Submillimetre water emission lines from comet 103P/Hartley 2.
and production rates from the measured line intensities. Observation
The time of the observations was 20 days after perihelion, when the comet was
and modelling details are given in Supplementary Information. A 1.095 AU from the Sun and 0.212 AU from Herschel. Because the H2O ground
critical point is that all observations sampled the same region of the state rotational lines in comets are optically thick29,30, observations of the rare
coma, about 6,500 km in diameter. oxygen isotopic counterpart, H218O, provide a more reliable reference for the
The retrieved gas column densities and production rates are sensitive D/H determination. The spectra of the 110–101 lines of HDO (a) and H218O
to collisional cross-sections, along with the density and temperature (b) at 509.292 and 547.676 GHz, respectively, were obtained with the
profiles of H2O and electrons, and we thus considered a range of model Heterodyne Instrument for the Far Infrared (HIFI) High Resolution
parameters (Table 1). Although the production rates determined for the Spectrometer (HRS) between 17.28 and 17.64 November 2010 UT. The line
various model parameters differ slightly, the value of the D/H ratio is intensities, expressed in the main-beam brightness temperature scale, are
0.011 6 0.001 and 0.117 6 0.002 K km s21, for HDO and H218O respectively,
estimated to be (1.61 6 0.24) 3 1024. In our analysis, we assumed an averaging the two instrument polarizations. The velocity scale is given relative
H216O/H218O ratio of 500 6 50, a range that encompasses the Earth to the velocity of the comet’s nucleus. The spectral resolution is 141 and
value and is consistent with previous measurements in cometary 132 m s21 for the HDO and H218O spectra, respectively. For details of the
water14 (see also Supplementary Information). The quoted 1s uncer- observational sequence and basic parameters of the data analysis, see
tainty in the D/H ratio includes a 5% uncertainty related to modelling. Supplementary information.
1
Max-Planck-Institut für Sonnensystemforschung, Max-Planck-Str. 2, 37191 Katlenburg-Lindau, Germany. 2California Institute of Technology, Pasadena, California 91125, USA. 3LESIA-Observatoire de
Paris, CNRS, UPMC, Université Paris-Diderot, 5 place Jules Janssen, 92195 Meudon, France. 4Rosetta Science Operations Centre, European Space Astronomy Centre, 28691 Villanueva de la Cañada,
Madrid, Spain. 5Astronomy Department, University of Michigan, Ann Arbor, Michigan 48109, USA. 6Space Research Centre, Polish Academy of Sciences, 00-716 Warsaw, Poland.

0 0 M O N T H 2 0 1 1 | VO L 0 0 0 | N AT U R E | 1
©2011 Macmillan Publishers Limited. All rights reserved
 RESEARCH LETTER

Table 1 | Calculating the D/H ratio in water in comet 103P/Hartley 2 a result, small Solar System bodies are expected to exhibit different
Model Tgas xne ,N(HDO). Q(HDO) ,N(H218O). Q(H218O) D/H D/H ratios in their water ice depending on the distance from the Sun at
(K) (1010 cm22) (1024 s21) (1011 cm22) (1025 s21) which they were formed.
(1) 50 0. 4.9 3.1 3.5 2.1 1.49 3 1024 In the context of this simple nebular model, the D/H ratio of
50 0.2 3.6 2.3 2.5 1.5 1.55 3 1024 (1.61 6 0.24) 3 1024 in comet 103P/Hartley 2—a factor of two lower
70 0.2 3.7 2.4 2.5 1.5 1.60 3 1024
Law 0.1 5.7 3.6 3.7 2.2 1.63 3 1024
than that measured in Oort-cloud comets (Fig. 2) and, within un
(2) 50 0.2 4.3 2.7 2.9 1.8 1.54 3 1024 certainties, consistent with that of the Earth’s oceans (for which the
Law 0.1 4.8 3.1 3.2 1.9 1.58 3 1024 Vienna Standard Mean Ocean Water (VSMOW) value is
The parameter xne, scaling the electron density profile in the models, is constrained by mapping (1.558 6 0.001) 3 1024)—is therefore surprising, and compatible
observations. ,N. and Q are respectively beam integrated column density and production rate, with two different schemes: (1) either this comet did not form in a
determined using different parameters in the excitation models14,29. Production rates were computed
assuming isotropic outflow of water from the nucleus, with a velocity of 0.6 km s21, consistent with the region that was further from the Sun than the assembly zone of the
width of the H218O line. We accounted for the 1099 offset between the centre of the beam and the Oort-cloud comets, or (2) the dependence of the water D/H ratio with
position of the peak of the H2O distribution. Values of 50 and 70 K for the gas kinetic temperature, Tgas,
are consistent with multi-transition measurements of gaseous species in the millimetre and near-
distance from the Sun is not as expected on the basis of current models.
infrared range, respectively. The gas kinetic temperature is expected to decrease with increasing Concerning the first possibility, dynamical models indeed suggest that
distance owing to quasi-adiabatic expansion of the escaping gases: the temperature law assumes that a fraction of the Jupiter-family comets originate in the Oort cloud22.
Tgas 5 80 K for r , 270 km, Tgas 5 12 K for r . 630 km, with a linear decrease between 270 and 630 km,
where r is the distance from the nucleus. Collision cross-sections involving water molecules and Still, even if comet 103P/Hartley 2 stems from the Oort cloud, this
electrons are modelled differently in models (1) and (2). Both models use an electron density profile would not explain why its D/H ratio is different from that seen in other
based on in situ measurements in comet 1P/Halley scaled to the activity of 103P/Hartley 2 (ref. 29). The
D/H ratio is equal to 0.5 3 Q(HDO)/Q(H2O), with Q(H2O) 5 500 3 Q(H218O). See Supplementary
Oort-cloud comets. Models also suggest that a fraction of the Jupiter-
information for details of the models and model parameters. family comets may have originated from the Trojan asteroid swarms
sharing the orbit of Jupiter23. The Trojans are generally thought to have
comparable to the primordial D/H ratio in the Universe after the Big resided at their current location since the formation of the Solar
Bang (Fig. 2)15. Protosolar water, on the other hand, is believed to be System. Therefore, Jupiter family comets originating in the Trojan
highly enriched (D/H < 1 3 1023)16 due to the low-temperature region could, in theory, display deuterium enrichment values lower
(,10–30 K) non-equilibrium chemistry that characterizes the dense than those for bodies originating in the Kuiper belt, if they indeed
interstellar medium17, either via gas-phase isotopic exchange reactions formed in the vicinity of Jupiter. However, the most probable scenario
involving ions and radicals, or grain-surface processes. Consequently, is that 103P/Hartley 2 originated in the Kuiper belt.
the resulting D/H ratio in water ice is very sensitive to the physical It is difficult to explain the low D/H ratio in 103P/Hartley 2 (com-
conditions, in particular the kinetic temperature of the medium. After pared to that of previous measurements in comets) with the formation
the protosolar cloud collapsed to form the solar nebula, isotopic regions of comets, thus models of the gradient of D/H in the Solar
exchange reactions between molecular hydrogen and HDO molecules System—predictions not yet directly confirmed by observations,
would have led to a gradual reduction of D/H in water18, as compared owing to scarcity of accurate isotopic measurements—may need to
to the initial interstellar medium value. Because the efficiency of these be revisited. In fact, one recent model has suggested that the D/H ratio
reactions and the turbulent mixing within the solar nebula is correlated of water vapour can be locally enhanced24. However, the vapour must
with the gas density and temperature, the deuterium enhancement in then be implanted into cometary ices. Moreover, until the measure-
water has been predicted to increase with the heliocentric distance19–21. ment of 103P/Hartley 2 there was no observational confirmation of
Ices, captured by planetesimals and cometesimals, would have then variations in the D/H ratio. One possible solution is that there was
preserved the deuterium enrichment in water from this early epoch. As large-scale movement of material between the inner and outer Solar

Hale–Bopp
Enceladus

Hyakutake
10–3

Tuttle

Ikeya–Zhang
2002T7
Halley

Hartley 2
D/H ratio

Earth

10–4

Protosolar
ISM

10–5
CI J S U N Oort cloud Jupiter family

Figure 2 | D/H ratios in the Solar System. Orange squares, values measured
for water in the Oort-cloud comets 1P/Halley, C/1996 B2 (Hyakutake), C/1995
O1 (Hale–Bopp), C/2002 T7 (LINEAR) and 8P/Tuttle. Arrow (for 153P/Ikeya–
Zhang), upper limit. Purple square, present measurement in the water of 103P/
Hartley 2. Black symbols, D/H ratio in H2 in the atmosphere of the giant
planets—Jupiter (J), Saturn (S), Uranus (U) and Neptune (N). Light blue and
green symbols, D/H values for water in the plume of Saturn’s moon Enceladus
and in CI carbonaceous chondrites, respectively. Error bars, 1s. The D/H
determinations in comets originating from the Oort cloud are twice the value
for the Earth’s ocean (blue line) and about a factor of ten larger than the
protosolar value in H2 (broad yellow line), the latter being comparable to the
value in atomic hydrogen found in the local interstellar medium (ISM, red
horizontal line). The D/H ratio in the Jupiter-family comet 103P/Hartley 2 is
the same as the Earth’s ocean value and the chondritic CI value. Uranus and
Neptune have been enriched in deuterium by the mixing of their atmospheres
with D-rich protoplanetary ices. For further details, see Supplementary Table 1.

2 | N AT U R E | VO L 0 0 0 | 0 0 M O N T H 2 0 1 1
©2011 Macmillan Publishers Limited. All rights reserved
Vol 461j29 October 2009jdoi:10.1038/nature08477

PROGRESS
Volatile accretion history of the terrestrial
planets and dynamic implications
Francis Albarède1

Accretion left the terrestrial planets depleted in volatile components. Here I examine evidence for the hypothesis that the
Moon and the Earth were essentially dry immediately after the formation of the Moon—by a giant impact on the proto-Earth—
and only much later gained volatiles through accretion of wet material delivered from beyond the asteroid belt. This view is
supported by U–Pb and I–Xe chronologies, which show that water delivery peaked ,100 million years after the isolation of the
Solar System. Introduction of water into the terrestrial mantle triggered plate tectonics, which may have been crucial for the
emergence of life. This mechanism may also have worked for the young Venus, but seems to have failed for Mars.

E
arth, Mars and Venus are three planets with very different
histories. Active plate tectonics, the presence of a liquid
ocean, and teeming life characterize the Earth. Mars has a
tenuous atmosphere, shows no incontrovertible evidence of
plate tectonics, lacks an abundant hydrosphere, and has so far evaded
all attempts to reveal the presence of life. Venus is a dead inferno of
carbon dioxide with clouds of sulphuric acid, and its solid surface
seems to have been violently resurfaced some 600–900 million years
(Myr) ago. At a time when so many satellites and telescopes are
probing the Universe in search of habitable planets, it is worth ques-
tioning what made three planets of our Solar System so different from
each other that only one of them harbours life.
0.60
Ivuna
0.50
Orgueil
0.40 Tagish Murchison
Lake Cold
Bokkeveld
δ66Zn (‰)
When life is the topic, the presence of liquid water immediately
becomes an issue. The origin of water and its fate have also been central
concerns of the planetary sciences, because water changes the rheolo-
gical properties of planetary mantles and seems to be a prerequisite for
plate tectonics to operate. A commonly held view is that the terrestrial
ocean results from the outgassing of the Earth’s mantle1. Recent
dynamic calculations, however, emphasize the hot temperatures next
to the nascent Sun, and suggest that the planetary embryos from the
inner Solar System remained essentially dry. It is only late in accretion
history that perturbations of asteroid orbits in the inner Solar System
by giant planets started to launch ice-rich material that delivered water
to the planets in that part of the Solar System. A simplistic but efficient
description of the debate on planetary water is that of ‘dry’ versus ‘wet’
accretion.
This Review will first show that the depletion of the volatile content
of the Earth and other terrestrial planets is not due to loss by out-
gassing but to incomplete stepwise accretion of nebular material; this
accretion was interrupted by energetic electromagnetic radiation that
was emitted by the young Sun during its T Tauri phase and swept
through the disk. The timing of volatile accretion, water in particular,
with respect to the giant lunar impact, the formation of the terrestrial

0.30 Allende Efremovka core, and its implications for the contrasting geodynamic regimes of
Lance Earth
the Earth, Mars, and Venus will be discussed.
Vigarano
Ornans
0.20 Water in the deep Earth
The Earth contains water in different forms: a liquid ocean at the
surface, various forms of ground water and, in the mantle, dense
0.10
hydrous mineral phases and water dissolved in nominally anhydrous
minerals such as olivine. Evidence for how much water the solid
0.00 Earth contains is robust. In the absence of a gaseous phase, that is,
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 at depths in excess of about 90 km, water behaves as an incompatible
Mg/Zn (g g–1)
element with properties remarkably similar to those of cerium2: the
Figure 1 | Fractionation of Zn isotopes is incompatible with volatilization. The H2O/Ce ratio is about 200 in the source of both mid-ocean ridge
data87 represent the d66Zn value, the relative deviation of the 66Zn/64Zn ratio in basalts and ocean island basalts, which are widely taken as melts from
per mil with respect to a terrestrial standard, of each sample. Data for terrestrial upper- and lower-mantle material, respectively. Given this strong
samples are shown in pale grey; data for the different classes of carbonaceous constraint and reasonable assumptions on the degrees of melting
chondrites are shown in dark blue (CI), green (CM), red (CV) and orange (CO).
prevailing during the formation of these melts, the amount of water
In the case of volatilization, preferential loss of the light 64Zn isotope in the
residue is expected to accompany the increase of the Mg/Zn ratio, as magnesium held by the mantle is approximately equivalent to an oceanic mass,
(T50 5 1,336 K) is much more refractory than zinc (T50 5 726 K). Because the which translates into 150–350 p.p.m. H2O in the mantle and a bulk
opposite is observed, fractionation processes other than volatilization must be terrestrial water concentration of 300–550 p.p.m. (refs 3, 4). The
sought. T50, at which 50% of the element is condensed13. water content of the Archaean mantle is open to interpretation4,
1
Ecole Normale Supérieure de Lyon, Université Claude-Bernard Lyon 1, and CNRS, 46 allée d’Italie, 69007 Lyon, France.
1227
©2009 Macmillan Publishers Limited. All rights reserved
PROGRESS
900
800
700
Atomization enthalpy (kJ mol–1)
Alkali and
600 high-temperature
chalcophile elements
500
Low-temperature
chalcophile elements
400
and halogens
Ge
C Sn
300 Ga
In
200 Pb
Tl
Zn Bi Na
Cd Rb
100
Te
N O Hg I S Sb
0 BrSe Cs
0 500
F 1,000
50% condensation temperature, T50 (K)
Figure 2 | Stepwise accretion of the elements on cooling of the solar
nebula, shown as atomization enthalpy versus T50. The two axes represent
two different volatility scales, an intrinsic scale given by the atomization
enthalpy29, which is essentially equivalent to the mean bonding energy per
atom in solids, and a scale dependent on the nebular chemistry, T50 (see Fig. 1
legend). Black letters, major elements; grey letters, minor and trace elements.
The platinum group elements, plus Al, Ti, Zr and W, and most rare-earth
elements and actinides, which condense first above 1,600 K, are
preferentially found in refractory inclusions. The group of mildly refractory
but seems to have been lower than that of the modern mantle. As it is
particularly difficult to ascertain whether the measured water con-
tents of lavas reflect their contents at the time of eruption, the Ce
contents of melt inclusions in olivine from the 2.7-Gyr-old koma-
tiites from Belingwe, Zimbabwe5, and Alexo, Canada6, can be used as
a proxy; they show that if the H2O/Ce ratio has not significantly
evolved since that time, the mantle at 2.7 Gyr ago contained
,150 p.p.m. H2O for 50% melting and ,60 p.p.m. for 20% melting.
It has been suggested that much of the mantle’s water may accu-
mulate in the transition zone in waddsleyite7,8 and, therefore, that the
water content of the mantle may not be homogeneous. This is true
only if water is held back in the transition zone on subduction and not
dragged down with its carrier minerals. Fractionation across the
phase-change boundary requires that when wet material is trans-
ported in and out of the transition zone, the advective flux is com-
pensated for by diffusion of hydrogen: at the temperatures prevailing
in the sinking plates (600–1,200 uC), the diffusion coefficients of
hydrogen in the major minerals are very small9 and the wet boundary
layer appearing across the reaction zone very thin. Accumulation of
water in waddsleyite is therefore of little significance to the overall
budget of water in the mantle.
Volatile depletion in the terrestrial planets
Planets formed by accretion of solid material condensed from the
solar nebula. The strong electromagnetic winds emitted by the early
Sun swept away the nebular gas and condensation stopped well
before the debris disk was cleared, a process that lasted less than a
few Myr after the isolation of the Solar System10. Most of the material
that eventually formed the Earth accreted well within the ‘snow line’,
which is the virtual line beyond which water freezes out11 ,12. A widely
used scale of volatility is the temperature T50 at which 50% of the
nebular inventory has been accreted to the solid phase. Let us con-
sider two elements with widely different values of T50, namely, the
refractory element uranium (T50 5 1,610 K), and the more volatile
element potassium (T50 5 1,006 K)13. The K/U ratio is a measure of
the relative depletion of planetary volatile elements with respect to
1228
©2009 Macmillan Publishers Limited. All rights reserved
NATUREjVol 461j29 October 2009
Refractory siderophile
W
and lithophile elements
Ta Os
Re
Nb
Ir Hf Zr
Ru Mo
Pt Th
Rh
B U
V Ce
Si Co Ti
Ni La YGd
Pd Pr Lu
Au P Fe Sc
Cu Cr
Al
Mn Be Nd Er
Ag Dy
Li Eu Sr Tm
Ca Sm
K
Mg Ba Yb
Cl As
1,500 2,000
lithophile elements, which comprise the major elements Si, Mg, Fe and Ca,
accrete down to 1,300 K as metal, olivine and pyroxenes, and make up the
bulk of the planetary mantle and core. These refractory elements are
separated from the high-T chalcophile elements (As, Ga, Ge, Cu, Ag),
chlorine, and the alkali elements (Li, Na, K, Rb and Cs), which condense
between 1,150 and 850 K. From 750 to 530 K, the low-T chalcophile elements
(Pb, Bi, Sn, Zn, Cd, S, Te) precipitate, followed by the truly volatile elements
(N, C, H) and Hg. Each group appears to be separated from its neighbours by
a temperature gap (highlighted in white).
refractory elements. Carbonaceous chondrites and the solar pho-
tosphere provide a coherent estimate of the initial K/U ratio of
60,000 (ref. 14). The terrestrial value of ,10,000 (refs 15, 16) indi-
cates that the Earth lacks 85% of the nebular K inventory, while the
deficit for the Moon (K/U < 3,000) is 95%. Martian meteorites that
have been witnessed falling and therefore evaded long periods of
terrestrial weathering show K/U ratios ,20,000 (ref. 17). Depletion
of 92–98% is also inferred for Zn, Ag, As, Sb, Sn, Pb and, most
importantly, S (refs 18, 19). Planets that lack such large fractions of
moderately volatile elements cannot have been endowed with large
amounts of water.
As shown by isotopic abundances of different elements, the volatile
deficit in planetary material is not due to volatilization but to incom-
plete accretion20. Loss of volatile elements due to impact or volati-
lization during accretion, whether it reflects planetary escape
velocities or the molecular stampede of hydrogen escape known as
hydrodynamic entrainment21, is expected to be mass-dependent, and
the light isotopes should preferentially be enriched in the vapour
phase. The homogenous stable isotope composition of K in planetary
samples is inconsistent with preferential evaporation of the light
isotopes from initially condensed (CI chondrite) material. Even more
conspicuously, the Earth is depleted in the heavy isotopes of Zn by
about 0.1% per atomic mass unit with respect to CI chondrites, as are
the CV and CO chondrites, which are enriched in refractory elements
(Fig. 1). This is exactly the opposite of what is expected from pref-
erential loss of the light 64Zn during volatilization. These observa-
tions contrast with evidence from lunar soils, which on volatilization
by impacts become remarkably enriched in the heavy 66Zn and 68Zn
isotopes by 3–4% per atomic mass unit22. The processes leading to
reversed mass fractionation are currently not well understood, and
may involve kinetic effects23 or electromagnetic sorting (Hall effect)
of the ionized fraction of the nebular gas24. Incomplete accretion is
also inferred from the terrestrial abundances of alkali elements,
which have comparable volatilities and chemical properties over a
broad range of atomic mass M. The depletions of K (M 5 39,
T50 5 1,006 K), Rb (M 5 85, T50 5 800 K)25 and Cs (M 5 133,
NATUREjVol 461j29 October 2009
T50 5 800 K)26 are of comparable magnitude, which also argues
against loss by volatilization in a gravity field.
A robust consequence is that the Earth, like the other terrestrial
planets, is depleted in volatile elements, not because these were lost by
impact or vaporization, but because they did not accrete along with
the more refractory elements in the first place.
The stepwise accretion of planetary material
In order to explain the volatile deficit, the idea being pursued here is
that the nebular gas was blown off by the energetic radiation of the
young Sun before temperature decreased enough for volatile ele-
ments to condense on the terrestrial planets. Thermodynamic calcu-
lations27,28 successfully predict the mineralogy and composition of
planetary objects. Most elements condense within a narrow temper-
ature interval. As a consequence, accretion takes place as a stepwise
process and bulk accretion goes through plateaus when shifting from
one group of elements to another with decreasing temperature.
The narrow temperature range of condensation of most elements is
a common feature of thermodynamics calculations13,27,28 and reflects
relatively simple properties of solid–gas equilibria. The temperature
range DT over which elements condense is approximately RT50 2
=DHati ,
where R is the gas constant and DHati is the atomization enthalpy29.
Energy of mineral formation from solid elements and energy of melting
can be safely neglected with respect to the enormous energies involved
in vaporization. For most elements, DT corresponds to a few tens of
degrees, except for the alkali elements, Eu and Yb (150–200 K).
Intrinsic volatility is not the sole control of planetary abundances.
Figure 2 is a plot of atomization enthalpy versus T50. The highly
refractory elements—that is, the platinum group elements, Al, Ti,
Zr, W, and most rare-earth elements and actinides, which are all
particularly abundant in chondrite refractory inclusions—condense
above 1,600 K. They are followed by the dominant group of refractory
lithophile elements (Si, P, and the alkali-earth and transition elements
typical of chondrules), which condense down to 1,300 K. These
refractory elements are separated from a group comprising the
high-T chalcophile elements (As, Ga, Ge, Cu and Ag), chlorine, and
the alkali elements (Li, Na, K, Rb and Cs), which accrete between 1,150
and 850 K. At lower temperatures (750 to 530 K), the group of low-T
1,800
1,600
1,400
Temperature (K)
Plagioclase, clinopyroxene
1,200
1,000
800 Metal, olivine, orthopyroxene
600
0.0 0.2 0.4
Fraction condensed
Figure 3 | Stepwise accretion of the elements upon cooling of the solar
nebula, shown as fraction condensed versus temperature. Plotted is the
fraction of the original condensable matter remaining in the solar nebula at a
given temperature (redrawn with changes from ref. 30). The kinks in the
accretion rate correspond to the temperature gaps in Fig. 2. After formation of
the most abundant refractory phases (plagioclase and clinopyroxene), the
bulk of the metallic core and the mantle silicates (olivine and orthopyroxene)
is quickly removed from the nebular gas (purple). Alkali elements and high-T
chalcophile elements precipitate next (green), but do not constitute a large
fraction of the planetary material. Low-T chalcophile elements, sulphur and
halogens come after (red), followed by volatile elements. Most elements
accrete over a few tens of degrees, but over 150–250 K for the alkali elements,
and Eu and Yb. A planet with a substantial deficit of K or Zn with respect to
chondrites therefore cannot have accreted much, if any, water.
PROGRESS
chalcophile elements (Pb, Bi, Sn, Zn, Cd, S and Te) and the rest of the
halogens leave the vapour phase. A last group consists of the three
most volatile elements (N, C and H) and Hg. Each group appears to be
separated from its neighbouring groups by a temperature gap. The
stepwise accretion of the planetary material is well illustrated in refs 30
and 31, which show that, at 1,200 K, 50–60% of the nebular inventory
is still in the gas phase (Fig. 3). This reflects the combination of
the elements into stoichiometric solid phases such as oxides, olivine,
pyroxenes, feldspars and metal13, and, to a lesser extent, the intrinsic
volatility of the elements themselves.
The energetic electromagnetic radiation of the young Sun swept
away the nebular gas and stopped accretion while the nebular tem-
perature in the neighbourhood of the terrestrial planets was still in the
temperature range of alkali element condensation (800–1,000 K)30,32,33
(Fig. 4). This is the main cause of volatile element depletion in the
inner Solar System. In contrast, in material accreted at the snow line
and beyond, water and other volatile elements were plentiful. The ‘late
veneer’ hypothesis34, which holds that small amounts of chondritic
material from the asteroidal belt or beyond were added to the Earth at
an unspecified late stage, was devised to account for the present-day
excess of highly siderophile elements, such as the platinum group
elements, in the terrestrial mantle. The depletion of the most volatile
elements discussed above places an upper limit on the proportion of
low-temperature CI-like material of the order of 2–5%, which is
still somewhat higher than the 0.3% proportion of carbonaceous
chondrite-type (with ,10% water) material that is deemed4 necessary
to account for terrestrial water.
Making planets and adding water
One model35 of planetary growth and development may be summarized
into three stages: (1) the settling of dust onto the equatorial disk of the
planetary nebula and formation of kilometre-sized planetesimals, (2)
the runaway growth of planetesimals into Mars-sized planetary
embryos, and (3) collision of planetary embryos to form the planets
with more or less their modern masses. How fast the nebular gas was
‘Snow line’
1,000
Mid-plane temperature (K)
Gas
100
Sulphides
Jupiter
Venus
Earth
Mars
0.6 0.8 1.0
10
0.1 1 10
Distance to Sun (AU)
Figure 4 | The thermal structure of the planetary solar nebula:
temperature at mid-plane of the nebular disk. Data are taken from ref. 88.
Two models of temperature distribution are shown for two different values
of the viscous dissipation parameter, a (viscosity 3 sound velocity at mid-
plane 3 mean elevation above mid-plane) for a rate of solar accretion of 1%
per Myr: a 5 1023 (solid line) and a 5 1021 (dashed line). The position of
the wiggly ‘snow line’, which separates two domains, ice-free inwards and
frosty outwards, nearly coincides with that of the asteroid belt12. Colour
coded areas correspond to the four groups of elements identified by their T50
values in Fig. 2. The depletion of volatiles in the inner Solar System is caused
by the strong electromagnetic winds emitted by the young Sun, which swept
away the nebular gas before accretion was complete.
1229
©2009 Macmillan Publishers Limited. All rights reserved
PROGRESS
blown off by energetic solar radiation is unclear: the delay is constrained
by the short lifetime (,3 Myr) of the debris disk10 and by dynamical
models, which require a reduction of planetary eccentricity by dynamic
friction between planets, asteroids and planetary embryos36. It is likely,
however, that the planets may have accreted to a substantial fraction of
their final mass before the disk was blown off.
Water delivery by comets37 with a D/H ratio of 3 3 1024 (refs 38,
39) seems incompatible with the terrestrial D/H ratio of the terrestrial
ocean (1.5 3 1024)40. Carbonaceous chondrites represent an alterna-
tive source of terrestrial water1,41. Orbits of planetary embryos across
the Solar System during accretion have been computed42 and these
authors concluded that the bulk of the water at present on Earth was
carried by only a few of these objects, originally formed in the outer
asteroid belt and accreted to the Earth during the waning stages of
its formation. As long as the nebular gas was present and was absorb-
ing electromagnetic radiation, the inner Solar System was too hot to
allow for significant water condensation along with metal and sili-
cates: only during the late stages of accretion had temperature declined
enough for water delivery to be efficient. Dynamical models43 suggest
that early accretion of asteroid and planetary embryos was local, and
therefore that the accreted material was largely dry up to the snow line
and ice-rich beyond. It is also found that radial mixing and inward
migration of wet high-eccentricity objects from the outer Solar System
increased dramatically in the later phases of accretion. Water absorp-
tion in the hot inner Solar System has been proposed44 but, as two-
dimensional liquid films cannot form below the critical temperature
of water, this process is unlikely to be significant.
The mechanism of water accretion depends on the relative timing
of three important events—the segregation of the terrestrial core, the
giant lunar impact, and the arrival of the late veneer (Fig. 5). The
identical 182W excesses found in the silicate fractions of the Earth and
the Moon require that the giant impact took place at the earliest 60
Myr after the formation of calcium-aluminium-rich inclusions45
unless the Hf/W ratios of the two planetary bodies also are similar,
in which case this time can be reduced to ,30 Myr (refs 46, 47). A
Hf–W
Runaway
growth
T-Tauri Lunar
phase giant
U–Pb Late veneer
impact
I–Xe
U–Pb
0 20 40 60 80 100 120 140 160
Time after Solar System isolation (Myr)
Figure 5 | A tentative chronology of the Earth’s accretion. Chronometers
shown in brown. Accretion of planetary material was interrupted by
energetic electromagnetic radiation (T Tauri phase) sweeping across the disk
within a few Myr of the isolation of the solar nebula. Runaway growth of
planetesimals produces Mars-sized planetary embryos, which, collision after
collision, form the planets with their modern masses. The last of these ‘giant’
collisions left material orbiting the Earth that later reassembled to form the
Moon. The 182Hf–182W chronometer dates metal–silicate separation. The
identical abundance of radiogenic 182W between the Earth and the Moon
indicates that either the Moon formed after all the short-lived 182Hf had
disappeared (.60 Myr) or, rather, the Moon-forming impact and terrestrial
core segregation took place simultaneously 30 Myr after isolation of the solar
nebula. Addition of a late veneer of chondritic material coming from beyond
2.5 AU provides a strong explanation for the modern abundances of
siderophile and volatile elements in the terrestrial mantle. This material also
contained water and other volatile elements, which account for the origin of
the terrestrial ocean. Such a model indicates that most of the terrestrial Pb
and Xe was delivered by the asteroids that constituted the late veneer, and
therefore that the young Pb–Pb and I–Xe ages of the Earth date, not the
Earth, but events that affected the asteroids. It is suggested here that these
events are those of the accretion to the Earth of the late veneer.
1230
©2009 Macmillan Publishers Limited. All rights reserved
NATUREjVol 461j29 October 2009
lunar impact with a wet Earth, that is, after substantial amounts of
late veneer had been accreted, is difficult to reconcile with the
remarkable depletion of volatiles in the lunar mantle30,48 with respect
to the terrestrial mantle. The contrasting abundances of the stable
(unradiogenic) Ne isotopes 20 and 22 (ref. 49) and of the stable Xe
isotopes 124, 128 and 130 (ref. 50) between the terrestrial mantle,
represented by basalts and CO2 wells, and the terrestrial atmosphere
also require that a substantial fraction of the volatiles was accreted
after the last large-scale melting of the Earth. The invariant propor-
tions of the stable Ar isotopes 36 and 38 in all terrestrial solids and
gases51,52 indicates that isotope fractionation of Ne and Xe was not
caused by hydrodynamic escape. Although the case has been made
recently that the lunar mantle may locally hold traces of water53, the
depletion of the Moon in volatile elements, regardless of their atomic
weights and therefore of gravitational loss, remains unexplained.
However, the occasional closeness of planetary bodies with very dif-
ferent degrees of hydration and, because of lateral transfer, the highly
variable water contents of impactors are strong features of dynamic
simulations43.
Because Pb is volatile, with a T50 value (,725 K; ref. 13) similar to
that of Zn, the arrival of the late veneer can be dated by Pb–Pb
chronology. Old feldspars and galenas indicate Pb–Pb ages of the
Earth younger by 50–160 Myr relative to the age of the Solar
System54 ,55 (Fig. 6). The delayed ingrowth of radiogenic Pb therefore
is best explained by impacts of asteroids with very low, chondrite-like
m (238U/204Pb) ratios56 onto a volatile-depleted, high-m proto-Earth.
The m value of the lunar mantle attested to by Pb isotope systematics
of Apollo samples is 200–600 times the ratio of the modern terrestrial
mantle and 1000–10,000 times that of CI chondrites57. Simple mass
balance using the m values suggests that well over 99% of terrestrial Pb
was added by the late veneer. The young Pb–Pb age of the Earth
therefore appears to be the mean age of the arrival of the late veneer
material and does not represent a date in the history of primordial
terrestrial Pb. Some U/Pb fractionation due to volatilization or to
incomplete mergers (‘hit-and-run’ collisions) during the encounter
of these asteroids with the Earth may account for the young Pb–Pb
ages. Likewise, the apparent I–Xe and Pu–Xe ‘ages of the Earth’,
which are bracketed between 50 and 110 Myr (refs 58–60), date the
arrival on Earth of I and Xe, two highly volatile elements, and are
Age, T (Gyr)
1.0 2.0 3.0 4.0 5.0
20
18
16
206Pb/204Pb
14
12
10 Canyon Diablo
4.40 ± 0.04 Gyr
8
0.0 0.2 0.4 0.6 0.8 1.0 1.2
exp(λ238UT) – 1
Figure 6 | Galena data and the young age of the Earth. Data are taken from
ref. 54. The evolution of 206Pb/204Pb in conformable Pb deposits is linear in
exp(l238U T) 2 1, where l238U is the decay constation of 238U, which
indicates that the mantle source of galenas behaved as a closed system since
the early geological ages. The primordial Pb of the Solar System, represented
by the 4.56-Gyr-old Canyon Diablo troilite89, is not on the galena trend,
which suggests that terrestrial Pb is younger than that of this meteorite. This
condition reflects that either Pb segregated into the core at a late stage89 or, as
suggested here, that terrestrial Pb was largely inherited from the late
accretion of volatile-rich material (late veneer) to the Earth.
NATUREjVol 461j29 October 2009
reasonably consistent with the Pb–Pb age. Hence, arrival of the late
veneer about 100 6 50 Myr after the collapse of the nebula is sup-
ported by independent sets of geochemical observations.
The nature of the wet, late-accreting icy bodies from beyond the
snow line, either identified meteoritic or so-far unsampled material, is
not understood1,41. It is likely that any source of planetary material that
significantly contributed to planetary accretion during the first 50 Myr
of the Solar System is now exhausted. This makes the quest for a
particular type of modern chondritic contributor intrinsically elusive.
Hydrating the mantle and setting off planetary dynamics
If water is a late addition to terrestrial accretion, the question arises of
when and how modern mantle water was introduced into the solid
Earth4. Solution in the primordial magma ocean (ingassing)61
requires a very high atmospheric temperature (.800 uC) to prevent
mantle hydration. Foundering the hydrous crust of serpentine and
other hydrous minerals62 of the magma ocean is an alternative mech-
anism for early deep Earth hydration. In both cases, however, crys-
tallization of the magma ocean is expected to concentrate water, a
very incompatible molecule, in the residual melt and therefore acts to
remove most of the water initially present in the upper mantle.
Water and all the other incompatible elements are particularly
depleted in the upper mantle for two main reasons. First, orogenic
volcanism at converging plate boundaries removes most of the volatile
and incompatible lithophile elements from the hydrous lithospheric
plates. Later, mid-ocean-ridge magmas remove from the upper mantle
most of what survived the subduction ‘filter’. In addition, it is a general
feature that magma extraction under mid-ocean ridges and ocean
islands is a rather shallow process (,100 km) that takes place during
decompression of the mantle: the deeper into the mantle water and
incompatible elements are buried, the less likely they are to become
involved in melting and extraction, which is apparent in convection
models dealing with geochemical tracers63. Water is continuously lost
to the deep mantle by subduction of a post-serpentine hydrous phase
known as phase D64–66. The presence of hydrous phases at the top of the
lower mantle may actually account for stronger seismic attenuation at
subduction zones than in the ambient mantle67.
Water decreases the viscosity of major mantle minerals, notably
olivine68. It reduces the strain before lithospheric failure and there-
fore affects mantle dynamics69. A strong reduction of mantle viscosity
contingent on subtle olivine hydration is suspected to shift the elastic
thickness of the lithosphere and to change mantle dynamics from a
stagnant-lid regime (strong dry mantle) to a plate tectonic regime
(soft wet mantle)70.
The varying fates of water on the different terrestrial planets provide
a relatively simple explanation of their contrasting evolution.
Admitting that water and Ce have similar residence times of
6–8 Gyr (ref. 71) in the terrestrial mantle, ,50% of the ocean must
have been subducted into the mantle after 4.5 Gyr, which is precisely
what the widely held view of ‘one ocean inside for one ocean outside’ is
predicting.
The status of water on Venus can be explained by conditions
opposite to those on the Earth: hot surface vapour quickly reacts
with the crust and is rapidly transported downwards because the
more hydrous the mantle, the faster its convection. The loss of water
from the surface is largely controlled by its reactivity and therefore by
surface temperature. On both planets, the depletion of the upper
mantle acts to hold water back in the lower mantle. The unique
hypsometric maximum on Venus contrasts, however, with the bimo-
dal distribution of elevations on Earth72 and demonstrates its lack of
true continents. Evidence of plate tectonics, if it existed in the past,
was erased by resurfacing ,600–900 Myr ago73,74. Although some
water is currently lost from the atmosphere of Venus75, the relative
abundances of atmospheric rare gases argues against hydrodynamic
escape in the past76. If Venus and the Earth initially had oceans of
similar sizes, an alternative may be found to atmospheric loss on
Venus: water may simply have been dragged into the mantle by early
©2009 Macmillan Publishers Limited. All rights reserved
PROGRESS
plate tectonics. Eventually, this may become the fate of the terrestrial
oceans and continents as well, once enough water has been intro-
duced into the mantle, making it softer than it is today.
It has been argued77 that Martian water was delivered by asteroids
and comets from beyond 2.5 AU. Also, it was suggested that Mars’
interior is dry because, although the planet acquired a late volatile-
rich veneer, it did not get folded into the mantle78. A dry mantle
explains the early demise of plate tectonics on Mars. In contrast to
the Earth and Venus, Mars lost its atmosphere to space by hydro-
dynamic escape during the first few hundred Myr after accretion79.
Fractionation of the atmospheric 38Ar/36Ar ratio attests to hydrody-
namic escape76. The weak gravity field due to the small planet radius
has a triple effect: (1) escape from the atmosphere is much easier on
Mars than on Earth; (2) because the pressure gradient is about three
times weaker on Mars, the phase changes giving rise to the viscosity
jump of the terrestrial transition zone are shifted to the bottom of the
Martian mantle, thereby preventing significant accumulation of
water at depth; and (3) the bending of plates and, hence, the onset
of plate tectonics is made more difficult. It has been suggested that the
mantle source of Martian shergottites could be wet80,81, but this is
only possible if these rocks are younger than a few hundred Myr, a
point that was recently challenged on the basis of Pb–Pb ages in excess
of 4 Gyr (ref. 82).
Contrasting dynamic regimes, and, therefore, the amount of water
present in planetary mantles, have diverse effects on the dynamo: a
plate tectonic regime allows heat to be evacuated more efficiently
from the core than a stagnant-lid regime83. The Earth, and possibly
early Venus, illustrates the former case with active plate tectonics. In
the latter case, the dynamo chokes, the magnetosphere vanishes, and
the atmosphere gets eroded by solar wind: this was probably the fate
of Mars, initially triggered by a water-poor mantle and the weak
gravity field of the planet, and may become the fate of Venus.
Future directions
Beyond the fascinating question of how the Earth formed and
acquired its hydrosphere lies the question of the origin of life.
Associating life on Earth with plate tectonics and the presence of
continents, all of them being contingent on the presence of water, is
tantalizing and justifies the ‘follow the water’ motto of space agencies.
Although the reducing environments created on reduction of water by
the magma ocean may have jump-started biological activity84, it is not
clear that a ‘water world’ can provide a steady source of nutrients and
sustain life85. Both the timing of water delivery (relevant to the onset of
plate tectonics) and the amount of water (for the persistence of a
shallow ocean with subaerial reliefs) need to be understood. The
detailed chemical reactions at work during the giant lunar impact86
and even its timing remain poorly understood. New dynamical
models of accretion will need to deal with radial transfer across the
nebula and with hit-and-run impacts that severely affect planetary
chemistry. The past decade witnessed major conceptual changes in
our understanding of the early history of the terrestrial planets, and
more surprises seem to be ahead of us.
1. Drake, M. J. & Righter, K. Determining the composition of the Earth. Nature 416,
39–44 (2002).
2. Michael, P. Regionally distinctive sources of depleted MORB: evidence from trace
elements and H2O. Earth Planet. Sci. Lett. 131, 301–320 (1995).
Brackets the geochemical behaviour of water during magma generation between
that of rare-earth elements lanthanum and neodymium.
3. Saal, A. E., Hauri, E. H., Langmuir, C. H. & Perfit, M. R. Vapour undersaturation in
primitive mid-ocean-ridge basalt and the volatile content of Earth’s upper mantle.
Nature 419, 451–455 (2002).
4. Marty, B. & Yokochi, R. in Water in Nominally Anhydrous Minerals (eds Keppler, H.
& Smyth, J. R) 421–450 (Rev. Mineral. Geochem. 62, Mineral. Soc. Am., 2006).
5. McDonough, W. F. & Ireland, T. R. Intraplate origin of komatiites inferred from
trace-elements in glass inclusions. Nature 365, 432–434 (1993).
6. Lahaye, Y., Barnes, S. J., Frick, L. R. & Lambert, D. D. Re-Os isotopic study of
komatiitic volcanism and magmatic sulfide formation in the southern Abitibi
greenstone belt, Ontario, Canada. Can. Mineral. 39, 473–490 (2001).
1231
•  El
 agua
 de
 los
 océanos
 de
 la
 Tierra
 fue
 traída
 por
 un
 cometa
 de
 hielo.
 En
 la
 
figura
 se
 presenta
 
 un
 cometa
 gigante
 estrellándose
 contra
 un
 planeta
 
rocoso
 aportando
 polvo
 de
 hielo
 rico
 en
 carbono
 hasta
 la
 superficie
 del
 
planeta.
 Un
 brillante
 destello
 rojo
 capta
 el
 momento
 del
 impacto
 en
 el
 
planeta.A
 la
 izquierda
 aparece
 en
 colores
 amarillo-­‐
 blanco
 la
 estrella
 Eta
 
Corvi
 con
 cometas
 moviéndose
 hacia
 ella.
 El
 telescopio
 espacial
 Spitzer
 
detectó
 espectros
 de
 agua
 helada,
 materia
 orgánica
 y
 rocas
 alrededor
 de
 
Eta
 Corvi
 -­‐
 componentes
 esenciales
 de
 los
 cometas.
 Esta
 es
 la
 primera
 vez
 
en
 que
 se
 ve
 una
 tormenta
 de
 cometas
 alrededor
 de
 otra
 estrella.
 Eta
 
Corvi
 Cene
 la
 edad
 adecuada,
 unos
 mil
 millones
 de
 años,
 para
 
experimentar
 un
 bombardeo
 de
 cometas
 similar
 a
 lo
 que
 ocurrió
 en
 
nuestro
 propio
 sistema
 solar
 de
 600
 a
 800
 millones
 de
 años
 de
 edad,
 
durante
 el
 llamado
 Late
 Heavy
 Bombardment
 
 provocado
 en
 nuestro
 
Sistema
 Solar
 por
 la
 migración
 de
 los
 planetas
 exteriores,
 que
 expulsaron
 
algunos
 cometas
 de
 hielo
 y
 enviaron
 otros
 hacia
 el
 interior
 del
 Sistema
 
Solar.
 Los
 cometas
 entrantes
 impactaron
 en
 nuestra
 Luna
 y
 también
 a
 
otros
 planetas
 interiores
 incluso
 en
 la
 Tierra
 ayudando
 al
 comienzo
 de
 la
 
vida
 actual.
 
 J.L. Bada / Earth and Planetary Science Letters 226 (2004) 1–15

mal accretionary phase of

n the period immediately
g impact event at 4.51 Ga,
ely covered with a liquid
. Any water at the surface
steam in the atmosphere.
ould have incinerated any
d from whatever sources.
he Earth was likely devoid
oir, especially in compar-
ere 20–30% of the surficial
matter.
of the Earth prior to ~3.9
isotopic analyses of the 4.3
icate that temperatures had
point that liquid water was
ce about 100–200 million
Earth’s water is thought to
om degassing of hydrated
the mantle and to a lesser
eroids and comets [20]. If
deed mainly come from
Fig. 1. The impact history of the early Earth and the various stages
the early Earth underwent
involved in the origin and evolution of life. Based on [2,22].
f the water on the surface
y released early in Earth’s Earth, which gives rise to the so-called faint young Sun
<em>Science</em>Shot: The First-Known Comet to Strike Earth http://news.sciencemag.org/print/earth/2013/10/scienceshot-first-known-comet-strike-earth

Published on Science/AAAS | News (http://news.sciencemag.org)

Home > ScienceShot: The First-Known Comet to Strike Earth

Article Title:
ScienceShot: The First-Known Comet to Strike Earth
October 11, 2013
Carolyn Gramling

A black, diamond-spackled pebble just a few centimeters across is the remainder of a comet that struck Earth almost 29 million years ago—making
it the first direct evidence of a comet exploding in our atmosphere, scientists say. The stone, which the scientists named “Hypatia” after an
Alexandrine mathematician and philosopher, was found in 1996 among tumbled bits of yellow sand glass (also known as the Libyan Desert Glass)
scattered across tens of kilometers in southwestern Egypt, near the border with Libya. The glass itself, one large polished piece of which has a
prominent place in a necklace that belonged to Egyptian pharaoh Tutankhamen, has been dated to 28.5 million years and has long been thought to
be the result of a meteorite impact or an airburst caused by a comet breaking up in Earthʼs atmosphere. To determine its origin, scientists
Terry Bakker performed a range of tests on the tiny pebble, examining its mineralogy, bulk chemistry, carbon isotope, and noble gas content. The stoneʼs noble
gas content supports an extraterrestrial origin, while the presence of tiny diamonds—larger than nanodiamonds found in a common kind of
meteorite called chondrites, but similar in size to diamond aggregates known to be formed by impacts—supports a cometary origin. The stone is also markedly carbon-rich,
more so than other known extraterrestrial material aside from comets, the researchers will report in an upcoming issue of Earth and Planetary Science Letters. All of this, they
say, points to a cometlike object entering Earthʼs atmosphere [1], where it exploded, cooking the desert sand below to 2000°C and forming the Libyan Desert Glass.

See more Science [2]Shots [2].

Links:
[1] http://www.sciencedirect.com/science/article/pii/S0012821X13004998
[2] http://news.sciencemag.org/scienceshot
 Published on Science/AAAS | News (http://news.sciencemag.org)
Home > Stellar Graveyard Shows Signs of Possible (Past) Life

Article Title:
Stellar Graveyard Shows Signs of Possible (Past) Life
October 10, 2013
Lizzie Wade

Mark A. Garlick

Ripped to shreds. The white dwarf GD 61 devoured a water-rich asteroid, pointing to the possibility that its solar system contained habitable exoplanets before its
parent star died.

The next time you gaze up at the night sky, remember that youʼre looking at a graveyard. The Milky Way is studded with dead stars, from black holes to neutron stars to dim
white dwarfs. Often, these stellar corpses reveal little about their past. But once in a while, they provide clues to how they—and any planets they once hosted—lived and died.
Now, scientists have discovered evidence that a white dwarf known as GD 61 was once orbited by a rocky, water-rich asteroid—just the kind of thing youʼd need to build a
habitable alien world.

Most stars (including, in about 4 billion years, our sun) end their lives as white dwarfs, after they have burned all their nuclear fuel. These super dense stellar embers exert
such strong gravity that any element heavier than helium will immediately sink to the dwarfʼs core. So imagine astronomersʼ surprise when they discovered that some white
dwarfs are cloaked in layers of “pollution” made up of silicon, oxygen, and other elements much higher up on the periodic table.

This pollution is made up of “pieces of planetary systems that are falling into their central stars,” explains Jay Farihi, an astronomer at the University of Cambridge in the United
Kingdom. By measuring the pollutionʼs constituent elements, scientists can peer back in time and discover what the original solar systemʼs asteroids, comets, and planets were
made of. “Itʼs a wonderful technique for doing planetary forensics,” says Michael Jura, a white dwarf expert at the University of California, Los Angeles, who wasnʼt involved in
the current research.

In GD 61ʼs pollution, Farihi and his colleagues noticed a curious abundance of oxygen. Their first thought was that the original asteroid must have been encrusted with carbon
dioxide in the form of dry ice. Trouble is, there was no carbon anywhere to be found around GD 61. So in order to account for the extra oxygen, “the only chemically viable
substance left is water,” Farihi says.

Online today in Science, the team proposes that GD 61 has “shredded” a rocky asteroid that was 26% to 28% water [1]. About the same size as Vesta in our solar systemʼs
asteroid belt, the asteroid likely orbited the white dwarfʼs precursor, an A-type star slightly bigger than the sun. After that star died, the white dwarfʼs stronger gravity probably
dragged in the asteroid and tore it apart.

Water-rich asteroids are considered to be important for the formation of habitable planets, crashing into them and supplying them with life-giving liquid water. Although “we
certainly canʼt rewind the clock completely” to discover what GD 61ʼs original solar system looked like, Farihi says, the discovery of the asteroid reveals “the building blocks of
Earth-like planets were there.” In the future, he hopes to look at the system with a powerful telescope like the Atacama Large Millimeter/submillimeter Array radio array in Chile
to see if any of the original planets survived the death of their star, or whether anything remains of the asteroid belt where the water-rich planetesimal was born.

Finding a water-rich asteroid near a white dwarf bolsters the long-shot idea that life might rise again around these dead stars, says John Debes, an astronomer at the Space
Telescope Science Institute in Baltimore, Maryland, who wasnʼt involved in the research. After their starsʼ violent deaths, white dwarfs remain remarkably stable for billions
Stellar Graveyard Shows Signs of Possible (Past) Life http://news.sciencemag.org/print/chemistry/2013/10/stellar-graveyard-shows-signs-possible-past-life

1 de 2 more years, which would mean “theyʼd actually be really good places to live,” Debes says. The problem, however, is that a planet would have to be extremely close to a dim 12/10/13 12:17
white dwarf to be warm enough to support life—as close as the devoured asteroid was to GD 61, in fact. Still, Debes says, if a rocky planet managed to settle into a stable orbit
that close to a white dwarf, Farihiʼs teamʼs discovery shows that “there might be some hope of getting water onto those planets” via asteroids—just what happened on Earth so
many billions of years ago.

Links:
[1] http://www.sciencemag.org/lookup/doi/10.1126/science.1239447
 Until now, no white dwarf has shown re- (i) The observed carbon deficiency indicates that dently conclude that the debris around GD 61
liable evidence for the accretion of water-rich, this element has no impact on the total oxygen originated in a water-rich parent body. Although
rocky planetary material. Unambiguous signa- budget, even if every atom is delivered as CO2. the lifetimes of disks at white dwarfs are not
tures of icy asteroids at white dwarfs should (ii) The elements Mg, Al, Si, and Ca are as- robustly constrained, the best estimates imply
include (i) atmospheric metal pollution rich in
refractory elements; (ii) trace oxygen in excess
of that expected for metal oxides; (iii) circum-
stellar debris from which these elements are ac-
creted; and, where applicable, (iv) trace hydrogen
(in a helium-dominated atmosphere) sufficient
to account for the excess oxygen as H2O. The
presence of a circumstellar disk signals that ac-
cretion is ongoing, identifies the source material,
and enables a confident quantitative assessment
of the accreted elemental abundances, which in

Table 1. Oxide and water mass fractions in

the planetary debris at GD 61. We adopt the
steady-state values, which assume accretion-diffusion
equilibrium.
Oxygen carrier Steady state Early phase
CO2 <0.002 <0.002
MgO 0.17 0.18
Al2O3 <0.02 <0.02
SiO2 0.32 0.27 Fig. 1. Oxygen budget in GD 61 and terrestrial bodies. The first two columns are the early phase
CaO 0.02 0.01 (EP) and steady-state (SS) fractions of oxygen carried by all the major rock-forming elements in GD 61,
FeO* 0.05 0.02 assuming that all iron is carried as FeO. Additional columns show the oxide compositions of the bulk
Excess 0.42 0.50 silicate (crust plus mantle) Earth, Moon, Mars, and Vesta (35). Their totals do not reach 1.0 because trace
H2O in debris 0.26 0.33 oxides have been omitted. The overall chemistry of GD 61 is consistent with a body composed almost
*All iron is assumed to be contained in FeO; some metallic Fe entirely of silicates, and thus appears relatively mantle-like but with substantial water. In contrast, Earth is
will modestly increase the excess oxygen. relatively water-poor and contains approximately 0.023% H2O (1.4 × 1024 g).

www.sciencemag.org SCIENCE VOL 342 11 OCTOBER 2013 219

~160  
•  Meteor Crater, Arizona, is one the
worlds most well known crater.
•  Less than 1 mile across, it was
created about 50,000 years ago.
•  Formed by an iron asteroid.
Lots of melted droplets and solid
pieces of an iron-nickel material have
been recovered in the area.  
 Crater diameter:
12.6 km
Buried under rocks
and sand for a long
time, it has been
uncovered again by
recent erosion.

Possible crater
Aorounga may be part
of a crater chain
 Structure diameter:
30 miles
Formed by volcanic
processes.

Not every circular feature on Earth is an impact crater!

It is necessary to visit the feature on the ground to observe its structural features and
obtain rock samples. Only then we can be sure of what it is.
 D<Dth

D>Dth

Dth= Threshold diameter for transition from simple to

complex craters (around 4 km on Earth)
      Ciclo  hidrológico  

•  El ciclo hidrológico es el resumen de la circulación del

agua disponible en la superficie de la Tierra
•  Cualquier interrupción en el ciclo hidrológico provoca
un aumento en el agua retenida en los océanos (produce
una transgresión marina) o en los continentes: recarga
de acuíferos, glaciares (produce una regresión marina)
con efectos inmediatos en el nivel del mar.
      El  ciclo  hidrológico