My ignorance is at least equal to yours here nprev, but I am under the impression that forming frost requires a mobile, or 'liquid' monolayer at the forming surface of the crystals, kept liquid by the pressure of the van der waals forces between the atoms. I think SickNick or Dburt would have some knowledge on whether this is always true. I don't know if any studies have been done to see if frost forms without the monolayer under mars-like conditions?
If it is always true this implies some difference between the surface and near surface material.
Surfaces without thin films of (usually but not always) water experience much higher friction coefficients, as high relief areas, known as asperities, on the two surfaces bond together. This might explain some of the clumping, but not why the clumping seems to lessen over time.
EDIT: Actually it might. If the regolith beneath the surface is ultra dry, and the atmosphere in the soil pores to, then soil freshly removed from the subsurface will have no water monolayers and particles will be held together by cold welds between asperities. After some days exposed to the atmosphere above the surface, which does have some water content, monolayers will start to form, lubricate the particles and hey presto the stickiness goes away.
If there's enough H2O in the Martian air to make that a possibility (no idea myself) then this could be tested by scooping up some subsurface, testing it for monolayers, leaving it in a heap on the deck for a few days then testing it again. If the soil becomes 'damp' after a few days on the deck then we have an explanation in the making and I shall buy myself a new bottle of absinthe!
![rolleyes.gif](http://www.unmannedspaceflight.com/style_emoticons/default/rolleyes.gif)
END EDIT.
It's been pointed out to me on the the yellow and black rover forum that relative humidity is temperature dependent and would be expected to vary significantly as the temperature changes, as relative humidity is the partial pressure of the water vapour divided by the saturation vapour pressure of water at a given temperature. Is it possible there has been a misinterpretation somewhere, and the change in relative humidity is entirely due to temperature, and little or no vapour is coming off the ice? Again my ignorance of humidity and atmospheric physics s pretty profound