Climate Evolution on Venus and Mars

James KastingDepartment of Geosciences

Penn State University

Talk Outline

Part 1: Evolution of Venus’ atmosphere: runaway and moist greenhousesPart 2: Evolution of Mars’ atmosphere: the H2-CO2 greenhouse

Venus

• 93-bar, CO2-rich atmosphere• Practically no water (10-5

times Earth)• D/H ratio = 150 times that on Earth

What went wrong with it?

Possible answers:1) Venus never had any water to begin with

-- But this is unlikely for a variety of reasons, including the high D/H ratio

2) Venus’ climate got out of control because of positive feedback loops in the climate system-- The positive feedback loop involving water vapor can lead to a runaway greenhouse

Question:What went wrong with Venus?

Classical “runaway greenhouse”

Goody and Walker, Atmospheres (1972)After Rasool and deBergh, Nature (1970)

Assumptions:• Start from an airless planet• Outgas pure H2O or a mixture of H2O and CO2

• Solar luminosity remains fixed at present value• Calculate greenhouse effect with a gray atmosphere model

1 bar

Problems with the classical runaway greenhouse model

• Gray atmosphere approximation• No convection• No variation in solar luminosity• Planets acquire atmospheres during

accretion by impact degassing of incoming planetesimals

Alternative runaway greenhouse calculation

• Imagine a thought experiment in which you push the present Earth closer to the Sun

J. F. Kasting, Icarus, 1988

• Do this by gradually increasing the surface temperature in one’s climate model

Vertical temperature structure

• Lower atmosphere temperature structure should be approximately adiabatic• Get moist or dry adiabat near the surface, depending on whether liquid water is present

Ocean present No ocean

Calculated T and H2O profiles

Temperature Water vapor

• The troposphere expands as the surface temperature rises• Water vapor becomes a major constituent of the stratosphere at surface temperatures above ~340 K (Ingersoll, JAS, 1969)• Hydrogen can then escape rapidly to space because the diffusion limit is overcome

Tropopause cold trap

• Temperature decreases rapidly with height in the troposphere, then levels out (or increases) in the stratosphere

• The H2O vapor pressure decreases with height in the troposphere, then remains constant (or increases) in the stratosphere

• The H2O saturation mixing ratio

fsat = Psat/Pmust therefore go through a minimum at some height. We call that height the tropopause cold trap

Cold trap

Catling and Kasting, Atmospheric Evolution, Appendix B, in prep.

Alternative runaway greenhouse calculation

• Now, calculate radiative fluxes. Define

FIR = net outgoing IR fluxFS = net absorbed solar flux for the

present solar luminosity• Then

SEFF = FIR/Fs = solar flux (relative to today) needed to sustain that temperature

Runaway greenhouse: FIR and FS

J. F. Kasting, Icarus (1988)

• Outgoing IR flux levels out above ~360 K (90oC) because the atmosphere is now opaque at those wavelengths

Present Earth

Simpson-Nakajima limit

Planetary albedo vs. surface temperature

• The albedo decreases with increasing Ts initially because of increased absorption of solar near-IR radiation by H2O• At higher Ts, the albedo increases because of increased Rayleigh scattering by H2O

J. F. Kasting, Icarus (1988)

(Seff)

• Recall that Seff = FIR/FS

• The stratosphere becomes wet (and the oceans are thus lost) at Seff = 1.1. The corresponding orbital distance is 0.95 AU• Venus is at 0.72 AU

Moistgreen-house

Runawaygreenhouse

Update: New albedo calculations using the HITEMP database

Goldblatt model* Kasting (1988) model

• As first pointed out to us by Colin Goldblatt (U. Victoria), our old climate model may have seriously underestimated absorption of visible/near-IR radiation by H2O. New data are available from the HITEMP database

Goldblatt et al., Nature Geosci. (2013)

Runaway greenhouse thresholds: old and new

New model(Kopparapu et al., Ap.J., 2013)

Old model(Kasting et al., 1988)

• Our own calculations using updated absorption coefficients for both H2O and CO2 suggest that the runaway greenhouse threshold is much closer than previously believed (runaway: 0.97 AU, moist greenhouse: 0.99 AU)

3-D modeling of runaway greenhouse atmospheres

• Fortunately, new studies using 3-D climate models show that the runaway greenhouse threshold is increased by ~10% because the tropical Hadley cells act like radiator fins– This behavior was pointed

out 20 years ago by Ray Pierrehumbert (JAS, 1995) in a paper dealing with Earth’s tropics

• This puts the runaway greenhouse threshold back at 0.95 AU (Seff 1.1)

Leconte et al., Nature (2013)

Outgoing IR radiation

• Venus lost its water either during or after accretion because it developed either a runaway or moist greenhouse, allowing hydrogen to escape rapidly to space– Hamano et al. (Nature, 2013) (not discussed) argue

that the runaway greenhouse was triggered during accretion, so that water loss occurred early

– This model is likely correct, as it more easily explains the loss of Venus’ oxygen

• Once the water was gone, volcanic CO2 (and SO2) built up in Venus’ atmosphere, leading to its present, hellish state

Summary of Part 1: Evolution of Venus’ atmosphere

Part 2: Evolution of Mars’ atmosphere: the H2-CO2 greenhouse

Evolution of Mars’ climate and atmosphere

• Let’s switch now to Mars

• Present Mars is cold and dry, but early Mars looks like it was wet, and maybe warm

MARS PATHFINDERTwin peaks view

• Mars today is a frozen desert• But the surface of Mars is riddled with ancient fluvial features

NirgalVallis

(Viking)

From: J. K. Beattyet al., The New Solar System, 4th ed.

200 km

Terrestrial vs. martian valleys

Colorado RiverCanyon

Nanedi Vallis

10 km 10 km

Valleys on Mars looka lot like valleys (orcanyons) on Earth

Figure credit: C. Harman (unpublished)

How did the martian valleys form?

Two basic types of models:

1. Warm early Mars modelMars’ early climate was essentially Earth-like. The water needed to carve the valleys came from rainfall (Pollack et al., 1987)

2. Cold early Mars model(s)Mars was cold most of the time, but it warmed up sporadically due to impacts (Segura et al., Science, 2002; Segura et al., JGR, 2008; Segura et al., Icarus, 2012)

Nanedi Vallis(from Mars Global Surveyor)

~3 km

River channel

• The Grand Canyon on Earth required ~17 million years to form• During this time, roughly 5106 m of rain should have fallen on the Colorado plateau (assuming 30 cm/yr of rainfall)• By contrast, those who have argued that the martian valleys could have formed in a cold environment, following large impacts (Segura et al. Science, 2002) assume much lower volumes of precipitated water, 50-500 m• The terrestrial analogy suggests that these estimates are too low by a factor of 104-105!

How could early Mars have been kept warm?

• To answer this question, we first need to think about how Earth remained habitable early in its history– I talked about this in the previous lecture

• Low solar luminosity was a problem for the Earth, as well

• Earth has stabilizing feedbacks, though, most importantly the one between CO2 and climate, acting through the carbonate-silicate cycle…

The carbonate-silicate cycle

• This cycle regulates Earth’s atmospheric CO2 level over long time scales and has acted as a planetary thermostat during much of Earth’s history, because CO2 builds up as the climate cools• This same negative feedback could have operated on early Mars

• But, there is a problem with early Mars because a CO2-H2O greenhouse is not capable of warming the planet

SURFACE PRESSURE (bar)

0.001 0.01 0.1 1 10

SU

RF

AC

E T

EM

PE

RA

TU

RE

(K

)

180

220

260

300

MARS

0.7

0.8

0.9S/S0 = 1

J. F. Kasting, Icarus (1991)

Martian surface temperature vs. pCO2 and solar luminosity

• Previous calculations showed that greenhouse warming by CO2 (and H2O) could not have kept early Mars’ mean surface above freezing

S/S0 = 0.75 at 3.8. b.y. ago, when most of the valleys formed

Freezing point of water

Why is it difficult to warm early Mars?

• Two reasons:– Condensation of CO2

reduces the tropospheric lapse rate, thereby lowering the greenhouse effect

– CO2 is a good Rayleigh scatterer (2.5 times better than air), so as surface pressure increases, the increase in albedo outweighs the increase in the greenhouse effect

CO2

condensationregion

J. F. Kasting, Icarus (1991)

• Fortunately, it may be able to supplement the greenhouse effect on early Mars by adding additional greenhouse gases– CH4 doesn’t work. It’s main absorption band, at 7.7

m, is overlapped by CO2 when CO2 levels are high, plus it produces anti-greenhouse cooling from absorption of near-IR solar radiation in the stratosphere

– SO2 doesn’t work. It rains out when the climate is warm, and it forms sulfate aerosols when oxidized that reflect incoming solar radiation

– H2 works, however! H2 has no permanent electric dipole moment, but collision-induced absorption is able to excite its pure rotational spectrum

Science, Jan. 4, 2013

• As shown by Wordsworth and Pierrehumbert (2013), H2 can be an excellent greenhouse gas when present in sufficient concentrations• Could H2 have warmed early Mars?

Warming early Mars with H2 and CO2

• A combination of 1.3-4 bar of CO2 and 5-20% H2 by volume is enough to keep early Mars warm

• This amount of H2 could have been provided by volcanic outgassing on Mars if the planet was actively recycling volatiles, as Earth does, and if Mars’ mantle was more reduced (which it was, based on evidence from SNC meteorites)

Conclusions• Venus formed too close to the Sun to hold onto

its water, and hence lost that water by a runaway or moist greenhouse– The distance within which this occurs is ~0.95 AU

(i.e., Seff 1.1)

• Mars formed too far from the Sun to be warmed by a CO2-H2O greenhouse, but it could have been kept warm by the addition of 5-20% H2

– The distance beyond which CO2 and H2O cannot warm the planet—termed the ‘maximum greenhouse’ limit—is at ~1.7 AU for present solar luminosity (Seff 0.34)

• These calculations can be used to define the width of the habitable zone, which I will talk about tomorrow

• Backup slides

Sources of H2 on early Earth

• H2 could have come from volcanism

• On Earth, H2O and CO2 are the main gases emitted by volcanoes

• The H2:H2O ratio in volcanic gases is determined by the equilibrium reaction

H2O H2 + ½O2

pH2/pH2O = Keq/fO2½

• Mantle fO2 is near QFM (~108.5 atm at 1450 K), so pH2/pH2O 0.024

Mantle fO2 buffers

QFM• Quartz-fayalite-

magnetite

SiO2-FeSiO3-Fe3O4

• This is essentially the boundary between Fe+2 and F

• At 1200oC, fO2 is about 10-8.5 atm

IW• Iron-wüstite

Fe-FeO

• This is the boundary between Fe+2 and Fe0

• At 1200oC, fO2 is about 10-12.5 atm

Oxygen fugacities of SNCs

C. K. Shearer et al., Amer. Mineralogist (2006)

QFM

• But, Mars’ mantle appears to be more reduced than Earth’s, based on evidence from SNC meteorites• The reason may have to do with Mars being smaller, so that ferrous iron did not disproportionate in its lower mantle

• Finally, it looks now as if Venus’ runaway greenhouse occurred during accretion and never collapsed

• Other work considered what may have happened during planetary accretion

• Type I planets, like the Earth, develop steam atmospheres during accretion, but then these atmospheres collapse to form oceans within ~5 million years

Nature, 2013

Type I planet (Earth-like)

• Type II planets, like Venus, are too close to the Sun for the steam atmosphere to collapse, so the atmosphere remains in the runaway greenhouse phase until all the water is lost

• This model can more easily account for the loss of the oxygen left over from H escape, because it can react with the molten Earth as the escape happens

Type II planet (Venus-like)

Martian ‘Outflow’ Channel (Viking)

From: J. K. Beatty et al., The New Solar System, 4th ed

~ 200 km

Physics of H2 absorption• The energy levels of a (homonuclear) rigid rotator are spaced as:

Here, I is the moment of inertia of the molecule, R isthe intermolecular spacing, and m is the mass of eitheratom. The spacing between levels is readily shown to be

Physics of H2 absorption (cont.)

• If we express energies in terms of equivalent wavenumbers (1/), then

• So, remembering that

we see that we can absorb at 1000 cm-1 (10 m) at j = 7for H2, but that we would need j 250 for N2. This doesn’thappen because of non-rigid effects (bond stretching andbreaking).