We have produced a multiannual climatology of airborne dust from Martian year 24 to 31 using multiple datasets of retrieved or estimated column optical depths. The datasets are based on observations of the Martian atmosphere from April 1999 to July 2013 made by different orbiting instruments: the Thermal Emission Spectrometer (TES) aboard Mars Global Surveyor, the Thermal Emission Imaging System (THEMIS) aboard Mars Odyssey, and the Mars Climate Sounder (MCS) aboard Mars Reconnaissance Orbiter (MRO). The procedure we have adopted consists of gridding the available retrievals of column dust optical depth (CDOD) from TES and THEMIS nadir observations, as well as the estimates of this quantity from MCS limb observations. Our gridding method calculates averages on a regularly spaced, but possibly incomplete, spatio-temporal grid, using an iterative procedure weighted in space, time, and retrieval uncertainty. In order to evaluate strengths and weaknesses of the resulting gridded maps, we associate values of weighted standard deviation with every grid point average, and compare with independent observations of CDOD by PanCam cameras and Mini-TES spectrometers aboard the Mars Exploration Rovers ("Spirit" and "Opportunity"), as well as the Compact Reconnaissance Imaging Spectrometer for Mars aboard MRO. We have statistically analyzed the irregularly gridded maps to provide an overview of the dust climatology on Mars over eight years, specifically in relation to its interseasonal and interannual variability. Finally, we have produced multiannual, regular daily maps of CDOD by spatially interpolating the irregularly gridded maps using a kriging method. These synoptic maps are used as dust scenarios in the Mars Climate Database version 5, and are useful in many modelling applications in addition to forming a basis for instrument intercomparisons. The derived dust maps for the eight available Martian years (currently version 1.5) are publicly available and distributed with open access.

International audienceWe discuss 3D global simulations of the early martian climate that we have performed assuming a faint young Sun and denser CO2 atmosphere. We include a self-consistent representation of the water cycle, with atmosphere-surface interactions, atmospheric transport, and the radiative effects of CO2 and H2O gas and clouds taken into account. We find that for atmospheric pressures greater than a fraction of a bar, the adiabatic cooling effect causes temperatures in the southern highland valley network regions to fall significantly below the global average. Long-term climate evolution simulations indicate that in these circumstances, water ice is transported to the highlands from low-lying regions for a wide range of orbital obliquities, regardless of the extent of the Tharsis bulge. In addition, an extended water ice cap forms on the southern pole, approximately corresponding to the location of the Noachian/Hesperian era Dorsa Argentea Formation. Even for a multiple-bar CO2 atmosphere, conditions are too cold to allow long-term surface liquid water. Limited melting occurs on warm summer days in some locations, but only for surface albedo and thermal inertia conditions that may be unrealistic for water ice. Nonetheless, meteorite impacts and volcanism could potentially cause intense episodic melting under such conditions. Because ice migration to higher altitudes is a robust mechanism for recharging highland water sources after such events, we suggest that this globally sub-zero, 'icy highlands' scenario for the late Noachian climate may be sufficient to explain most of the fluvial geology without the need to invoke additional long-term warming mechanisms or an early warm, wet Mars. (C) 2012 Elsevier Inc. All rights reserved

No abstract