I'd prefer to describe the H2O on Mars as "usually frozen". The really important stuff is how often, when, and where it was NOT frozen.

A mental focus on "water" does not help us to make objective scientific assessments of the how, when , and where.

I have always tried to suggest that Mars is best seen as a slightly warm comet / asteroid / moon, rather than a cold Earth. We make automatic assumptions about processes, chemical activities, and rates on Earth, that are clearly not valid on Mars. There are stronger parallels with the outer solar system, imho, than with the inner.

Having made the declaration above, let's look at what we *know* or can infer about the Phoenix location.

It's an area like Siberia on earth, where there is permafrost. It's an area where every winter CO2 and H2O ices condense from the atmosphere onto the surface, and in the spring boil-off. This process is almost certainly solid-state, with no liquid water phase (otherwise, the surface chemistry would be very different) In the summer, the depth of sublimation extends several cm below ground surface, resulting in a dry crumbly surface layer, then an ice-cemented layer with possible lenses of ice segregation, as we see in Siberia. There are things we *don't* understand, because we have no equivalent environments on Earth - such as the demonstrable cohesion of the soils, that some people say "must" be due to wetness (but recall that the chemistry precludes wetness). Clearly, something is happenig that we don't understand. This is how we learn...

There are some interesting roles for extreme salts, which would over time form rare liquids, then leach downwards into the subsurface.

The probable answer is that what we see today is 99.99% the result of solid state deposition from sublimation, and that *almost* no liquids are ever involved.

But that residual 0.01% is fascinating.

We'll only understand it if we stop pretending that Mars was often, or frequently, or recently, "wet".