Answers are grounded in this paper and cite their sources.
Observations have been made in the region of Fox Peak, South Canterbury, of the distribution of frost action in an area of snow-tussock grassland. Ice needles were found only on bare loamy soil, which occurred mainly where erosion terraces had locally replaced the tussocks. Growth of ice needles was prevented by a cover of tussock litter from 1/2 in. to 2 in. thick, according to the severity of the frost. The same protection was afforded the soil by 2 in. to 3 in. of scree overlying it. The ice needles on bare soil usually diminished in size towards the margins of adjoining tussocks, but, even at a distance of several inches, the larger needles could, by moving soil downhill, undermine a tussock and sometimes excavate and expose its roots. In the last stages of undermining, frost is probably a less important mechanism. In the soil of the cave underneath the undermined topsoil, which survives by virtue of the binding action of the tussock above, ice needles are less common than in the open. Measurements of the diurnal fluctuations of soil temperature at a depth of 1 in. helped to explain these observations. Higher minimum temperatures and smaller overnight losses of heat were found under tussock litter than under an equal thickness of fine scree. A similar reduction of overnight losses of heat from the soil was observed where undermined tussocks overhung its surface.
No takes yet. Share an insight, caveat, or question.
M. W. Gradwell (1960) studied this question.
Synapse has enriched 2 related papers. Consider them for comparative context: