Despite Mars's bleak, cold climate today, there is abundant geologic evidence that liquid water once flowed on the Martian surface (and even recent evidence that some water still periodically lows today). The evidence of past flows includes river channels, now dry, that were once carved by powerful floods, valley networks of branching streams and tributaries indicating precipitation falling over wide areas, river deltas and layered sediments suggesting flow into standing bodies of liquid water, and salt and other minerals that dissolved in liquid water and were left behind when water evaporated. When that ancient wet period existed and when it ended is difficult to say.

The best method for dating a surface uses radioactivity, the process by which certain elements decay over time. One collects rocks from the surface and measures their radioactive elements to see how much of the decay products have accumulated in the rock since it formed. Transferring this age to the age of the surface can be tricky unless one knows where the rock came from. And one needs to collect the samples and bring them to a well-equipped laboratory.

The other way to date a planetary surface, which is used when rock samples are unavailable, is by counting craters. The longer a surface has been exposed to meteorite bombardment, the more craters it will have. This gives the relative age, and allows one to determine which surfaces are older and which surfaces are younger.To get the actual age one needs a standard. This was established for the Moon when lunar samples were returned to Earth as part of the United States Apollo program. The crater counts of the areas of the Moon from which the samples came were carefully noted. The samples were dated in terrestrial laboratories using radioactive relation between absolute age in years and crater density-numbers per unit area-was determined. This relation is valid for the Moon, and it is possible that it is valid for Mars as well. Then counting craters on a Martian surface should yield its actual age. It is possible, however, that the number of impacting objects is different for Mars. The problem has been carefully studied, and adjustments have been made, but the general conclusion is that a wet period occurred early in Mars's 4.5-billion-year history.

This causes a problem for Mars climate science, because the Sun was supposed to be fainter, perhaps by 20 or 30 percent, during the Mars wet period. Yet a watery Mars arguably would have required relatively warm temperatures. The power output of the Sun-its luminosity over time-is based on models of its interior and evolution. The models are well tested and agree with the luminosities of stars like the Sun whose ages are known. The same problem exists for Earth, which appears to have been warmer billions of years ago than it is today. Geologists have called this the "faint young Sun paradox." The solution to the paradox probably lies in climate science and not in the theory of stellar evolution, but the issue has not been entirely settled.

If Mars was wet billions of years ago, was it also warm? There could have been catastrophic floods released from underground aquifers (layers of rocks that can contain water) where the top surface of the water froze like the top of a Hawaiian lava flow. The solid crust forms an insulating layer, trapping the heat of the liquid and allowing it to flow for great distances without freezing. Lava tubes on Hawaii are empty channels where lava once flowed in this way. Thus the climate of Mars could have been wet but not warm, with the water flowing entirely under the ice. Ice sheets and glaciers on Earth today have extensive plumbing systems that allow water to flow and then drain away. Liquid water at the base of glaciers acts as a temporary lubricant and allows them to surge and slide downhill. Flood channels, river deltas, and standing bodies of liquid water all could have existed on Mars under thick layers of ice. The wet-but-not-warm hypothesis has the most trouble with areas that imply sources like precipitation over a wide area rather than the bursting of an underground aquifer. Further study of Mar’s geology will clarify these issues.