Episodic warming of early Mars by punctuated volcanism

The widespread evidence for liquid water on the surface of early Mars is di cult to reconcile with a dimmer early Sun. Many geomorphological features suggestive of aqueous activity, such as valley networks and open-basin lakes, date to approximately 3.7 billion years ago1–5, coincident with a period of high volcanic activity5,6. This suggests that volcanic emissions of greenhouse gases could have sustained a warmer and wetter climate on early Mars. However, models that consider only CO2 and H2O emissions fail to produce such climates7,8, and the net climatic e ect of the sulphur-bearing gases SO2 and H2S is debated9–11. Here we investigate the atmospheric response to brief and strong volcaniceruptions, includingsulphuremissionsandanevolving population of H2SO4-bearing aerosols, using a microphysical aerosol model. In our simulations, strong greenhouse warming by SO2 is accompanied by modest cooling by sulphate aerosol formation in a presumably dusty early Martian atmosphere. The simulated net positive radiative e ect in an otherwise cold climate temporarily increases surface temperatures to permit above-freezing peak daily temperatures at low latitudes. We conclude that punctuated volcanic activity can repeatedly lead to warm climatic conditions that may have persisted for decades to centuries on Mars, consistent with evidence for transient liquid water on the Martian surface. Geomorphological evidence for the presence of liquid water on the surface of early Mars includes valley networks1–3, openbasin, occasionally interconnected lakes4,5, meandering channels12, deltaic features13, and relatively rapid erosion rates14. Furthermore, hydrated minerals, especially various phyllosilicates and sulphates, are ubiquitous on ancient terrains15. Although many of the mineralogical observations do not strictly require a warmer climate, the requirement for surface runoff to explain the geomorphological features suggests that liquid water was present, at least transiently, on the early Martian surface. By inference, it has been widely suggested that the climate was warmer16, but the latest generation of high-order climate models cannot sustain near-melting mean annual temperatures anywhere on the planet even with a multi-bar CO2 atmosphere7,8. A clustering of valley networks and open-basin lakes in the Late Noachian and Early Hesperian1–5 (∼3.7Ga) overlaps with a long maximum in volcanic activity5,6. Most of the lava erupted during this time occurs as basaltic plains (‘Hesperian ridged plains’ and related units)6,17,18, with thicknesses up to hundreds of metres, and occasionally one to two kilometres, and an estimated area greater than 30% of the planet’s surface17. Two independent approaches suggest that the basaltic plains were emplaced by a series of brief eruptions, characterized by extreme effusion rates, and separated by millennial-scale hiatuses in eruptive activity. First, the basaltic plains seem to have effused predominantly fromwide fissures rather than central edifices, and the observed dimensions of the fissures imply effusion rates of 105–106 m3 s−1 (ref. 19). Dividing estimates of the total volume of the basaltic plains (3.3× 107 km3; ref. 17) by the lower end of this range of effusion rates gives an upper limit of ∼10,000 years on cumulative eruptive duration, which is only ∼0.01% of the total emplacement time of the basaltic plains (100–200Myr; ref. 18). Second, the thick, laterally extensive and topographically flat morphology of the basaltic plains, as well as the scarcity of central edifices, suggest an analogy to terrestrial flood basalts that may guide estimates of eruption rates. Radiometric ages and volumes of flow packages in the Columbia River Flood Basalts and the Deccan Traps indicate that individual eruptive episodes lasted one to ten years, and were separated by quiescent periods lasting up to 10,000 years20,21. Again, active eruption only ∼0.01% of the time is implied. The effusion rates implied by both approaches (105–106 m3 s−1) are more than 1,000 times higher than those estimated for earlier volcanic activity associated with the emplacement of the Tharsis rise (100–300m3 s−1; ref. 22). With typical terrestrial basaltic volatile content, such brief, high-rate (‘punctuated’) effusion implies instantaneous per-area outgassing rates up to several hundred times the terrestrial global average rate. The higher sulphur content of Martian magmas23, together with the large area and volume of the ridged plains relative to terrestrial flood basalts (Fig. 1), means that transient sulphur emission rates as high as a few thousand times the terrestrial global average were probable during plains emplacement. Thus, the average volcanic emission rate was probably composed of episodes of punctuated eruption separated by long quiescent periods. During eruptive phases, large amounts of SO2 were rapidly injected into the atmosphere, converted to H2SO4 with e-folding times of several centuries24, and incorporated into aerosols. The net climatic effect is debated. Strong greenhouse warming by SO2 (ref. 10) may be countered by comparably strong cooling due to scattering of solar radiation by aerosols11. If a CO2-dominated atmosphere were unable to sustain warm temperatures7,8, early Mars is expected to have been cold. The spatial distribution of ice under these conditions7 leavesmuch of the planet’s surface exposed and desiccated—and, therefore, a potential source of atmospheric dust. A modern-day dust lofting efficiency and atmospheric loading, augmented by fine-grained volcanic ash during explosive eruptions, carries important implications for Mars’ early climate. First, in addition to nucleating new, highly reflective H2SO4–H2O aerosols11, H2SO4 would condense onto pre-existing dust and ash particles. Unless the coating makes up much of the particle’s mass, its optical properties resemble those of dust, which is less reflective than homogeneous H2SO4–H2O (Supplementary Fig. 1). Second, the H2SO4 coatings, as well as any