About me Teaching activities My publications Various documents and links Contact me


  My main interests are the long term evolution of terrestrial planets and their habitability. Comparing well-known terrestrial planets and studying how they evolved can improve knowledge of both how these complex systems work and what determines their habitability, in particular when it comes to the Earth.

  Habitability is normally defined as the existence of liquid water at or near the surface.The "fluid envelope" of a terrestrial planet (i.e. atmosphere and ocean, if present) evolves greatly with time due to coupling with planet's interior (mantle and crust) and escapes to space.

  Earth is habitable while sustaining a plate tectonics convection regime. On the other hand, Mars and Venus do not possess plate tectonics and are not currently habitable. The increased exchanges permitted by plate tectonics could be an important factor for habitability. From another point of view, it can be noted that in the solar system plate tectonics only exists on the planets where liquid water is found. Plate tectonics might therefore require liquid water to be present. This illustrates how habitability and surface conditions are clearly linked to the inner dynamics of the planets and the exchanges of volatiles between solid and fluid layers.


  Mars is a logical start to studies about habitability, even though liquid water is not stable on the surface at present-day. Many signs point at very different conditions earlier in the history of the planet. I investigated how surface conditions could have changed over the last 4-4.5 billion years of martian evolution.

Figure 1: Possible mechanisms contributing to atmosphere formation and erosion [from Gillmann, 2012].

  This work was supported by French PNP funding (National Planetology Programm). It led to three publications (Gillmann et al., 2009a; Gillmann et al., 2011; Leblanc et al., 2012). I used modelling of non-thermal escape (sputtering, dissociative recombination, ion pick up, ionospheric outflow) powered by Extreme UV flux. It was constrained by ASPERA-3 (Analyzer of Space Plasma and EneRgetic Atoms) measurements (i.e. Lundin et al., 2009) and numerical modelling (Fox, 2004; Leblanc and Johnson, 2002; Ma et al., 2004; Lammer et al., 2003).

  Degassing from the mantle replenished the volatile content of the atmosphere. It was estimated through three different sources. First published results from numerical models (O’Neill et al., 2007; Manga et al., 2006; Breuer and Spohn, 2006), secondly observation of the surface (Greeley and Schneid, 1991) and finally self-consistent modelling using the stagYY code (i.e. Tackley, 2008).  

Figure 2: (left) Evolution of CO² partial pressure during the history of Mars for different CO² contents of the lavas. (right) Corresponding part of the atmosphere coming from volcanism.


  I studied the evolution of water at the surface, CO2 in the atmosphere and used Carbon, Nitrogen and Argon isotopic ratios as constraints. Our approach uses the present-day as starting point and reconstructs the history of Mars backward in time, which had not been done quantitatively at the time. Our results show that the martian atmosphere is quite young (2 billion years old), as opposed to the remnants of a primordial atmosphere. It also mainly originates from volcanism. Additionally, we showed that the martian atmosphere was thin except during the first 500 Myr and that it was cold and dry with water under solid form or in the subsurface, Higher partial pressure and water content in the atmosphere could only be reconcilied with late evolution and present conditions if one takes into account different mechanisms, such as the effects of impact delivery of volatiles.

Figure 3: Global evolution of martian CO² pressure during the last 4 Gyr, including the effect of LHB.


  As we became aware of the huge influence of early history of the planets on their long term evolution, we started to study this phase in greater depth. Venus was a good place to start as its thick atmosphere retains traces of events that occured early on. We started with the modelling of the first 500 Myr of the evolution of its atmosphere.

  It also led to a publication (Gillmann et al., 2009b). We modelled the intense linked hydrodynamic escape of hydrogen and oxygen and its effect on noble gas isotopic ratios (Zahnle and Kasting 1986; Kasting and Pollack, 1983; Hunten et al., 1987; Chassefière, 1996a,b).

Figure 4: Shematic view of the hydrodynamic escape process, leading to an important depletion in light species, but also affecting noble gases and other heavy components (water, CO²).

self consistent scenario of the evolution of Venus detailing the interactions and coupling between atmosphere, hydrodynamic escape and the “solid” planet.

By analyzing the different parameters governing the escape, we were able to reproduce the present-day isotopic ratios for Neon and Argon with limited hydrodynamic escape for Venus and an Earth-like situation. We were also able to propose a self-consistent scenario for the early evolution of Venus and its differences from the Earth. The study predicts that Venus should have received less water than the Earth during its accretion and that the hydrodynamic escape controlled the rate of the solidification of the magma ocean (which froze after 70-100 Ma) by pumping volatiles out of it (fig. 6). The intense escape would also have led to a very dry state by efficiently removing most of the primitive water from the atmosphere and magma ocean. On the contrary, Earth would have received more water early on (Morbidelli et al., 2000; Raymond et al., 2006) and would have remained wet until after the freezing of the magma ocean, leaving volatiles both in the mantle and at the surface. In that case, one of the differences between Earth and Venus could have occurred as early as the first few hundred million years.

Figure 5: self consistent scenario of the evolution of Venus detailing the interactions and coupling between atmosphere, hydrodynamic escape and the “solid” planet.



   We then wanted to couple this early evolution to the long term history of the planet, thus we investigated the coupling between mantle convection and surface onditions of Venus.

Figure 6: Mechanisms and feedbacks between layers in a terrestrial planet: current state of the model.

  It was funded by an ETH Fellowship in Zürich and led to a publication (Gillmann and Tackley, 2014). We investigated the coupled evolution of the atmosphere and mantle on Venus, focusing on mechanisms that deplete or replenish the atmosphere: atmospheric escape and volcanic degassing of the mantle. These processes were linked to obtain a coupled model of mantle convection and atmospheric evolution, including feedback of the atmosphere on the mantle via the surface temperature. During early atmospheric evolution hydrodynamic escape is dominant, while for later evolution we focused on non-thermal escape. The atmosphere is replenished by degassing, using mantle convection simulations (Armann and Tackley, 2012), and includes episodic lithospheric overturn. The evolving surface temperature was calculated from the amount of CO2 and water in the atmosphere using a gray radiative-convective atmosphere model. This surface temperature in turn acts as a boundary condition for the mantle convection model. We obtained a Venus-like behavior for the solid planet and an atmospheric evolution leading to the present conditions. CO2 pressure is unlikely to vary much over the history of the planet; a late massive build-up of the atmosphere over several resurfacing events seems unlikely. In contrast, water pressure is strongly sensitive to volcanic activity and varies rapidly, leading to variations in surface temperatures of up to 200 K, which have an effect on volcanic activity and mantle convection. Low surface temperatures trigger a mobile lid regime that stops once surface temperatures rise again, making way to stagnant lid convection that insulates the mantle.

Figure 7: Comparative evolution of volcanic production rate, surface temperature and volume averaged mantle temperature with time for the reference case. Also indicated are the different convection regimes. The transition from mobile lid tio stagnant lid is progressive. Early evolution (before 400 Myr) follows an episodic, but mostly stagnant, lid pattern.



  As we already observed in the case of Mars, other mechanisms could have influence on the evolution of terrestrial planets. Large impacts by meteorites, for example, are a prime suspect for large scale modification of surface and interior modifications. Impacts can indeed erode the atmposphere of the planet, but also bring volatiles to its surface. Additionally, those collisions can deposit a large amount of energy in the solid part of the planet. We investigated each of these effects.

  Impact erosion of the atmosphere has been incorporated by using results of numerical simulations by the SOVA hydrocode developed by Shuvalov et al. (2009). We compared those results to the simple tangent plane model as a constraint. Both models agree that large impacts are quite inefficient at removing volatiles from Venus' atmosphere. Large impacts are not numerous enough to substantially erode Venus' atmosphere. Single impacts don't have enough eroding power. Swarms of small bodies (<50km radius) might be a better candidate for this process. The amount of volatiles brought by large ordinary chondrite impactors is superior to losses and comparable to the degassing caused by the impact. Carbonaceous chondrite impactors are unlikely: they release too many volatiles, causing surface temperature to stay above 900K up to present-day.

  Mantle dynamics can also be modified by the heating caused by impacts. Heated material propagates by spreading across the upper mantle due to its buoyancy. Old crust destroyed or remixed in the mantle.

Figure 8: Evolution of Venusian mantle temperature and mantle composition with a 800km radius impact occuring at 150 Ma. For composition, 0 stands for a depleted mantle and 1 for basaltic composition. Atmosphere coupling is taken into account.

  A large part of the upper mantle melts, leading to its depletion and degassing. The timing of the impact has a small influence on the final surface temperature of the planet. However, it has a strong influence on the evolution of the surface temperature between 3.5 and 1.5 Ga, thus affecting surface conditions AND mantle (feedback). A large impact could lead to bypassing completely the low surface temperature period and thus make a strong difference for mantle convection evolution. Large impacts can affect long term evolution of terrestrial planets through their influence on key periods of the history of the body and trigger changes dur to the atmosphers/mantle feedback.

  With enough evently distributed high energy impacts, the mantle can be depleted by more than 90% of its volatiles during Late Veneer. This drastically cuts down degassing in the late history of the planet and leads to lower present-day surface temperatures. Total depletion of the mantle seems unlikely, meaning either few large impacts (1 to 4) or low energy (slow, grazing...) collisions. Combined with the lack of plate tectonics and volatile recycling in the interior of Venus, Late Veneer collisions could help explain why Venus seems dry today.

Figure 9: Long term effects of large impacts depending on their timing. Reference case without impact is in black line.