An outline of the paper: 3D modelling of the early Martian Climate under a denser CO2 atmosphere: Temperatures and CO2 ice clouds. by Forget et al.
The authors use climate models (general circulation models) to simulate the climate of Mars with a thicker atmosphere than at present. They consider a range of cases from 0.1 to 7 bars. Mars has 0.006 bars of atmosphere at present, but the early atmosphere would have been thicker. It was previously thought that an early pressure of several bars of CO2 was possible, but experts now think a value above 1 bar is unlikely.
Apart from the total pressure, the paper also examines the effect of different inclinations and orbital eccentricities, as these can vary widely in the history of Mars.
One key question about early Mars is weather the early atmosphere would allow the persistence of liquid water in lakes and open seas. Earlier results from 1-D calculations gave conflicting results.
What this paper finds is that no atmospheric pressure allows for extended periods of temperature above freezing. If the atmosphere is thin (<1 bar), then the CO2 greenhouse effect is insufficient. But when the pressure gets higher, CO2 starts snowing out of the atmosphere (the freezing temperature of CO2 increases with pressure), which cools the surface by the extraction of latent heat and by making the surface whiter and more reflective.
So it would seem that the ancient traces of flow seen on Mars are not due to stable lakes and seas, which leaves occasional impacts and volcanic heating as the main contenders.
Another interesting feature of early Mars according to these models is the dependence of temperature with altitude. In present Mars, the temperature is largely independent of height, despite the huge relief of >30km, because the atmosphere is too thin to make a difference. Ground temperatures are mainly dependent on the incoming sunlight, and undergo huge diurnal and seasonal variatoins. But when the pressure reaches a fraction of bar, the regime becomes nearer to that we know on Earth, with the lower part of the atmosphere, opaque to infrared heat, adjusting along an adiabatic profile. Lowlands like the northern plains and the Hellas basin would then stay warmer than the highlands.
Of course, the complexity of climate models means that they always have to be closely validated with observations, and in the case of early Mars, this is difficult. Could traces of ancient CO2 glaciers, for instance, be used to test these predictions? The authors do not offer a specific testable prediction at this stage.