The irreversible conversion of methane into higher hydrocarbons in Titan's stratosphere implies a surface or subsurface methane reservoir. Recent measurements from the cameras aboard the Cassini orbiter fail to see a global reservoir, but the methane and smog in Titan's atmosphere impedes the search for hydrocarbons on the surface. Here we report spectra and high-resolution images obtained by the Huygens Probe Descent Imager/Spectral Radiometer instrument in Titan's atmosphere. Although these images do not show liquid hydrocarbon pools on the surface, they do reveal the traces of once flowing liquid. Surprisingly like Earth, the brighter highland regions show complex systems draining into flat, dark lowlands. Images taken after landing are of a dry riverbed. The infrared reflectance spectrum measured for the surface is unlike any other in the Solar System; there is a red slope in the optical range that is consistent with an organic material such as tholins, and absorption from water ice is seen. However, a blue slope in the near-infrared suggests another, unknown constituent. The number density of haze particles increases by a factor of just a few from an altitude of 150 km to the surface, with no clear space below the tropopause. The methane relative humidity near the surface is 50 per cent.

Abstract. High resolution Mars Orbiter Camera (MOC)images have revealed the existence of clusters of small cones in the Cerberus plains, Matte Valles, and Amazonis Planitia, Mars. These cones are similar in both morphology and planar dimensions to the larger of Icelandic rootless cones, which form due to explosive interactions between surficial lavas and near-surface groundwater. Impact crater sizefrequency relationships indicate that surfaces upon which the cones sit are no older than 10 Ma. If martian cones form in the same manner as terrestrial rootless cones, then equatorial ground ice or ground water must have been present near the surface in geologically recent times.

Fields of small cratered cones on Mars are interpreted to have formed by rootless eruptions due to explosive interaction of lava with ground ice contained within the regolith beneath the flow. Melting and vaporization of the ice, and subsequent explosive expansion of the vapour, act to excavate the lava and construct a rootless cone around the explosion site. Similar features are found in Iceland, where flowing lavas encountered water-saturated substrates. The martian cones have basal diameters of c. 30-1000m and are located predominantly in the northern volcanic plains. High-resolution Mars Orbiter Camera images offer significant improvements over Viking data for interpretation of cone origins. A new model of the dynamics of cone formation indicates that very modest amounts of water ice are required to initiate and sustain the explosive interactions that produced the observed features. This is consistent with the likely low availability of water ice in the martian regolith. The scarcity of impact craters on many of the host lava flows indicates very young ages, suggesting that ground ice was present as recently as <10-100Ma, and may persist today. Rootless cones therefore act as a spatial and temporal probe of the distribution of ground ice on Mars, which is of key significance in understanding the evolution of the martian climate. The location of water in liquid or solid form is of great importance to future robotic and human exploration strategies, and to the search for extraterrestrial life.

Dust devils, and other columnar vortices, are associated with local surface pressure drops that can be observed in time-series data on both Earth and Mars. High cadence measurements are needed to resolve these small structures, and we report a month-long survey (June/July 2012) on a Nevada desert playa using microbarographs sampled multiple times per second. Candidate dust-devil signatures are classified, with detections being robust at about one per day for pressure drops exceeding 0.3 hPa (roughly a 5:1 signal-to-noise threshold, where the observed noise level corresponds reasonably well with the dynamic pressure associated with the estimate convective velocity scale). The vortex population is evaluated and compared with those observed on Mars : a broken power law or a more convex distribution describes the terrestrial data. A single station observes about three events per week (for normalized pressure drops of 0.06%), about three times fewer than Mars observations for the same normalized drop. We find evidence for clustering of vortex events in a pseudoperiodic manner with a 20-min period, consistent with the size of boundary-layer convection cells.