And a miniaturized version of the same process -- breaking up solid surfaces into crumbly, powdery material -- may be an alternative to glaciers to explain much of the "viscous creep" of sliding soil and rock that has eroded and smoothed out slopes and hills in this zone.

All the same processes are probably also going on right now on Mars' surface above 60 deg latitude; but they're currently concealed from view by the smooth near-surface ice layer that is currently formed over them there -- whereas they're currently exposed to view in the 30-60 degree latitude zone, where the thin ice layer that covered them tens of thousands of years ago has now disappeared as that zone warms up in the current stage of the obliquity cycle (regardless of whether that ice layer was created from above as a snowfall layer or from below as an ice lens).

All this explains Mars' surface features down to within 30 degrees of the equator quite well. But that's just that part of its surface that would be affected by such ice-transfer processes during the current part of its obliquity cycle -- in which, for the last 3 million years according to computer simulations, it's been slowly rocking back and forth between a tilt of 15 degrees and 35 degrees.

At other times in Mars' past, it's had a more dramatic tilt -- for instance, the same simulations indicate that during the period between 5 to 10 million years ago it was rocking back and forth between tilts of 25 and 45 degrees, with the climate effects from the most extreme part of THAT obliquity cycle being capable of making Mars' ice belts form all the way down at its equator!

Do we see any evidence that this also happened?

Well, during the millions of years since near-surface ice layers may have last been deposited at Mars' equatorial latitudes, the wind erosion there has apparently swept away most of the leftover kinds of evidence that can still be found at Mars' more recently iced-up middle latitudes -- including remaining patches of weakly salt-cemented soil from a near-surface ice layer, and any layer of ice-crumbled dry surface material that may have flowed down and blurred hills and slopes at the equator.

But we can still see some traces of evidence of past ice even in the tropics. For instance, climate simulations suggest that during such periods of maximum planetary tilt, the ice deposits forming at the equator would be deepest on the flanks of the four great shield volcanoes in the Tharsis region -- and photos show what looks very much like the marks left by now-evaporated glaciers in just those regions.

And Odyssey's gamma-ray and neutron measurements revealed something even stranger and harder to explain than Mars' highly concentrated near-surface high-latitude ice sheets -- two areas on the planet, over 2500 kilometers wide, which seem very rich in near-surface soil hydrogen right now despite the fact that they're located in its equatorial region. One is located in the Arabia Highlands; the other on roughly the planet's opposite side, south of the Amazonis and Elysium Plains. Their soils were originally thought to contain at least 4% ice by mass, assuming it's ice that's located there -- but the recent improved recalibration of Odyssey's GRS has now raised this figure to at least 8% by mass (13% if the hydrogen-rich material is shallowly buried beneath other dirt, as with Mars' near-polar ice layer).

At first it was thought that these areas might have a lot of minerals that had been hydrated by exposure to past liquid water, converting them into hydrogen-rich minerals such as clays or hydrated magnesium sulfate. But the new, higher measurement of water there probably rules that out as a full explanation -- such minerals would have to be extremely concentrated in these regions, but infrared spectra show no unusual concentration of them at all there.

The same puzzling conflict exists in these regions as in Mars' near-polar regions between Odyssey's thermal-neutron data and its other measurements as to just how deeply the hydrogen-rich layer is buried here -- but once again it looks as though we have a layer of dirt with a considerable amount of water ice mixed into it, buried beneath a layer of dry dirt only 30 cm or less thick. (Although the underlying ice-rich layer is here IS mostly dirt, rather than mostly ice like the near-polar ice layer).

And this is very hard to explain, since the noontime temperatures at Mars' equator right now are so high that any ground ice that was deposited that close to the surface during Mars' last period of very high obliquity should long ago have vaporized away during the several million years since.

These two water-rich regions show absolutely no other signs distinguishing them from their surroundings -- no concentration of valley networks, no other visible signs that they differ from the drier neighboring regions, and they aren't especially high-altitude -- which suggests that the ice may indeed have been deposited there during the highest-tilt phases of Mars' more seriously tilted obliquity cycle millions of years ago, by the complex wind patterns that happened to exist then. But how can that permafrost still be there, so close to the surface, at this point?