Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Earth science

Geomagnetism under scrutiny

New calculations show that the electrical resistance of Earth's liquid-iron core is lower than had been thought. The results prompt a reassessment of how the planet's magnetic field has been generated and maintained over time. See Letter p.355

Fluid flow in Earth's liquid-iron core sustains the planet's magnetic field against persistent losses due to the electrical resistance of liquid iron. The battle between generation and loss of the magnetic field suggests that a decrease in electrical resistance would tilt the balance in favour of generation. Surprisingly, the opposite may be closer to the truth. On page 355 of this issue, Pozzo et al.1 use a mathematical method called density functional theory to predict the electrical resistance of liquid-iron alloys at the high pressures and temperatures found in Earth's core. Their values are only two to three times lower than previous estimates2, but this change is large enough to affect our understanding of the dynamics and evolution of Earth's interior.

A good electrical conductor is essential for generating a planetary magnetic field. The movement of conducting material induces electric currents, which reinforce the initial magnetic field. A self-sustaining process requires that the effect of fluid motion exceeds the loss due to the conductor's electrical resistance. The ratio of these two effects is often characterized by the magnetic Reynolds number, which must exceed a threshold value for a magnetic field to be sustained3. High fluid velocity and/or low electrical resistance promote a large Reynolds number. It follows that a low resistance should enhance field generation, but only if the velocity is maintained at the necessary level.

Metals are good thermal conductors because electrons are more effective than atomic vibrations in transporting heat. Pozzo and colleagues' simulations confirm that the thermal conductivity of liquid iron under the conditions in Earth's core is several times higher than previous estimates2. They predict a value of roughly 125 watts per metre per kelvin (W m−1 K−1) at the top of the core and more than 200 W m−1 K−1 at the boundary between the outer and inner parts of the core. Such large thermal conductivities allow a substantial amount of heat to be carried by conduction, leaving less heat to drive convection. Convection may even cease in parts of the core4.

To illustrate the situation, let us consider a representative temperature profile in the liquid core (Fig. 1). Increase of temperature with depth (or pressure) causes conduction of heat towards the top of the core. The depth dependence of temperature in a convecting fluid is well approximated by an adiabatic profile, which is based on the idea that rising and sinking parcels of fluid do not exchange heat. When Pozzo et al. applied their new estimate for the thermal conductivity to an adiabatic profile in the core, they obtained a conductive heat flow of 15 terawatts (1012 W) near the top of the core. This value may exceed the heat flow across the core–mantle boundary5. In that case, warm fluid would accumulate at the top of the core, creating a stably stratified layer. As a result, convection and magnetic-field generation would be largely confined to the region below the stratified layer.

Figure 1: Temperature profile of Earth's interior.

The liquid-iron outer core lies between the mantle and the solid inner core. The increase in temperature across the liquid outer core is well described by an adiabatic process, in which no heat transfer occurs between ascending and descending liquid-iron parcels. Thermal conduction carries heat towards the top of the core. Pozzo and colleagues' study1 indicates that heat conduction inside the core may exceed the flow of heat across the core–mantle boundary. As a result, the temperature in the boundary region departs from the adiabatic profile to match the boundary's heat flow.

Pozzo et al. assess the consequences of high thermal conductivity for magnetic-field generation by constructing thermal 'histories' for the core. They present a suite of histories that could sustain a magnetic field, but in each case a very thick stratified layer or an additional energy source due to decay of radioactive elements would be required in the core. Reasonable arguments can be made against both of these options6,7, but one or the other seems to be unavoidable for maintaining the magnetic field. If Pozzo and colleagues' calculations are correct, then some of our basic assumptions about the core must be wrong.

One might question the calculations that predict high thermal conductivities. However, similar results have been obtained in independent calculations8, and there is further experimental support9 (albeit at temperatures much lower than core conditions). Accepting high thermal conductivity means that heat loss through conduction would substantially weaken thermal convection. A modest (subadiabatic) heat flow from the core would confine convection to a small region below a thick, thermally stratified layer. Such a layer would suppress variations in the magnetic field with time, which is at odds with observations. Alternatively, the need for a stratified layer could be eliminated if the heat flow from the core exceeded the heat conducted along the adiabatic profile. However, it is unclear how this high heat flow could be maintained over geological time. Indeed, most studies suggest that the heat flow from the core was higher in the past10. Perhaps the answer involves an unknown energy source. For example, chemical interactions between the core and the mantle might draw on the planet's gravitational energy. However, lack of the necessary understanding of the relevant chemistry at high pressures and temperatures means that this possibility cannot be assessed.

A high thermal conductivity for the liquid-iron core also has implications for the dynamics of the solid inner core. Iron at inner-core conditions is under higher pressure, and probably has a lower concentration of impurities, than iron in the overlying liquid core. Both of these factors would increase the thermal conductivity, so the value in the inner core should exceed 200 W m−1 K−1. Such a high value would make convection in the inner core11, including its 'translational' form12,13, unlikely. Instead, the inner core should cool by conduction. In such conditions, strong thermal stratification would develop and radial motion would effectively be suppressed. Because radial motion in the inner core is often invoked to explain the directional dependence (anisotropy) of seismic-wave speed in the inner core14, the high values of thermal conductivity should force researchers to look elsewhere for the cause of the seismic anisotropy.

It is remarkable that a modest change in thermal conductivity can have such a dramatic affect on the dynamics of Earth's core. More broadly, the latest study reveals how the properties of liquid iron make the operation of magnetic dynamos in terrestrial planets even more precarious than was previously believed. We are left with the challenge of understanding how Earth has succeeded in maintaining its magnetic field over most of geological time.


  1. 1

    Pozzo, M., Davies, C., Gubbins, D. & Alfè, D. Nature 485, 355–358 (2012).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Stacey, F. D. & Anderson, O. L. Phys. Earth Planet. Inter. 124, 153–162 (2001).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Christensen, U. R. & Aubert, J. Geophys. J. Int. 166, 97–114 (2006).

    ADS  Article  Google Scholar 

  4. 4

    Gubbins, D., Thomson, C. J. & Whaler, K. A. Geophys. J. R. Astron. Soc. 68, 241–251 (1982).

    ADS  Article  Google Scholar 

  5. 5

    Lay, T., Hernlund, J. & Buffett, B. A. Nature Geosci. 1, 25–32 (2008).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Gillet, N., Schaeffer, N. & Jault, D. Phys. Earth Planet. Inter. 187, 380–390 (2011).

    ADS  Article  Google Scholar 

  7. 7

    Corgne, A., Keshav, S., Fei, Y. W. & McDonough, W. F. Earth Planet. Sci. Lett. 256, 567–576 (2007).

    ADS  CAS  Article  Google Scholar 

  8. 8

    de Koker, N., Steinle-Neumann, G. & Vlček, V. Proc. Natl Acad. Sci. USA 109, 4070–4073 (2012).

    ADS  CAS  Article  Google Scholar 

  9. 9

    Hirose, K. et al. Mineral. Mag. 75, 1027 (2011).

    Google Scholar 

  10. 10

    Nakagawa, T. & Tackley, P. J. Geochem. Geophys. Geosyst. 11, Q06001 (2010).

    ADS  Article  Google Scholar 

  11. 11

    Buffett, B. A. Geophys. J. Int. 179, 711–719 (2009).

    ADS  Article  Google Scholar 

  12. 12

    Monnereau, M. et al. Science 328, 1014–1017 (2010).

    ADS  CAS  Article  Google Scholar 

  13. 13

    Alboussière, T., Deguen, R. & Melzani, M. Nature 466, 744–747 (2010).

    ADS  Article  Google Scholar 

  14. 14

    Sun, X. & Song, X. Phys. Earth Planet. Inter. 167, 53–70 (2008).

    ADS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Bruce Buffett.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Buffett, B. Geomagnetism under scrutiny. Nature 485, 319–320 (2012).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing