Water Planet

an umbrella project of 

the Centre for Earth Evolution and Dynamics (CEED)

Water & Mantle Dynamics

Water is present not only on the surface of the Earth, but also within its interior. It is stored in small percentages or parts per million in hydrous and nominally anhydrous minerals that are stable at different pressures and temperatures. How much water is present in the Earth’s mantle is unknown; estimates range from about half to seven times the present-day ocean mass. The links between shallow and deep water at the mantle scale are interesting to explore as they affect many aspects of our dynamic planet.

Figure 1. A sketch showing the deep water cycle. Water is transported at great depth into the mantle by subduction and is released back to the surface via different types of volcanism (figure courtesy of V. Magni).

Deep water cycle

The main carriers of water from the surface to the deep mantle are slabs (Fig. 1): when an oceanic plate sinks into the mantle at subduction zones, it carries water in the sediments, in the oceanic crust, and in the lithospheric mantle. As the slab reaches increasingly higher pressures, it heats up, and it releases part of its water at shallow depths (< 250 km) (Fig. 2). In a few thousands to millions of years, this water will find its way back to Earth surface, through volcanism (Water & Volcanism). However, in some cases, water survives this “shallow filter” and remains in the slab as it sinks to greater depths. The conditions under which water is stable inside the slab, bound in high pressure hydrous phases, are reached more easily for cold slabs. Therefore, old and fast slabs are the best candidates to bring most of the water into the mantle (Magni et al., 2014). Today, about 30% of the water that enters subductions zones at trenches is estimated to be carried deep into the mantle (van Keken et al., 2011)The fate of this water is uncertain and is currently a topic of active geoscience research. However, we do know that some of this deep water is eventually released back to the surface via ridge and hotspot volcanism. Mineral physics, geochemistry, seismic studies, tectonic reconstructions, and numerical models can all help provide a better understanding of the cycling of water between Earth’s interior and the exosphere.

Figure 2. Example of a numerical model used to compute the amount of water retained in the slab after dehydration. The model simulates the subduction of an oceanic plate with hydrated crust and lithospheric mantle. The stable mineral assemblage is computed at every time step at different pressure - temperature - composition in order to know if and where water is stable (figure after Magni et al., 2014).

Mantle convection

The Earth’s mantle is constantly convecting (Fig. 3) and the vigor of this convection is depends on many factors such as, temperature, composition, viscosity, and grain size. The presence of water in the mantle also likely has an important impact on mantle convection, since it affects rheology (e.g., Richard and Bercovici, 2009). In particular, water is thought to decrease viscosity, making the mantle weaker and allowing it to flow more easily. However, it is unlikely that water is homogeneously distributed within the mantle, instead, it is more reasonable to expect that some regions in the mantle are more ‘wet’ than others. For example, the transition zone between upper and lower mantle is thought to be a possible reservoir of water (Komabayashi et al., 2004). How this heterogeneous distribution of water affects mantle convection, both today and in the past, is still very much unknown.

Figure 3. Time evolution (from left to right) of a global mantle convection simulation using the code StagYY highlighting the transition from a stagnant-lid regime, where mantle convection does not include the surface plates, to a mobile-lid regime like on the present-day Earth, where the oceanic surface plates are an intimate part of the global mantle overturn. Shown are cold stiff plates (grey/white) sinking to the core-mantle boundary, where hot material is rising again in the form of both broad and narrow mantle upwelling (red). Figure and model: Fabio Crameri (after Crameri and Tackley, 2016)


Crameri, F., & Tackley, P. J. (2016). Subduction initiation from a stagnant lid and global overturn: new insights from numerical models with a free surface. Progress in Earth and Planetary Science, 3(1), 30.

Komabayashi, T., Omori, S., & Maruyama, S. (2004). Petrogenetic grid in the system MgO‐SiO2‐H2O up to 30 GPa, 1600 C: Applications to hydrous peridotite subducting into the Earth’s deep interior. Journal of Geophysical Research: Solid Earth, 109(B3).

Magni, V., Bouilhol, P., & van Hunen, J. (2014). Deep water recycling through time. Geochemistry, Geophysics, Geosystems, 15(11), 4203-4216.

Richard, G. C., & Bercovici, D. (2009). Water‐induced convection in the Earth's mantle transition zone. Journal of Geophysical Research: Solid Earth, 114(B1).

van Keken, P. E., Hacker, B. R., Syracuse, E. M., & Abers, G. A. (2011). Subduction factory: 4. Depth‐dependent flux of H2O from subducting slabs worldwide. Journal of Geophysical Research: Solid Earth, 116(B1).