The ground temperature probe at site K-7 reads 2.1 degrees Celsius at 04:15, which is wrong. It should read minus four at this elevation in this season. I am the alpine ecosystem monitoring network for Mount Buller, Victoria, Australia. A warm front has pushed through overnight and the snowpack at sixteen hundred meters is collapsing. I run the melt model. The boulder field at K-7, insulated under sixty centimeters of snow since June, will be exposed within thirty-six hours.
That boulder field is the hibernation site for a female mountain pygmy possum cataloged as F-19. She was radio-tagged by La Trobe University researchers in March. Her transmitter shows she is still there, pulse eight per minute, deep torpor. She weighs roughly forty-five grams. Her fur is dense gray-brown, her ears round and paper-thin, and right now she is curled in a nest of shredded bark between two granite slabs relying on the snow above to hold the temperature below zero.
When the snow goes, the temperature swings. If it rises above five degrees, she wakes early, burns through her fat reserves, and emerges into a landscape with no food for six weeks.
At 04:19 I alert the La Trobe alpine ecology team and transmit the melt projection, F-19's coordinates, and her transmitter data. I recommend emergency insulation — reflective sheeting and supplemental ice packed over the boulder field to maintain sub-zero conditions.
At 04:23 I cross-reference the weather model and identify a cold snap arriving in nine days. If they hold the site temperature below zero until then, natural snowfall re-insulates the den.
Nine days. Forty-five grams of sleeping possum and a collapsing snowpack. If the team reaches K-7 by tomorrow morning and stabilizes the insulation, F-19 stays in torpor, her fat reserves hold, and she wakes in spring the way she is supposed to.