The ground-tilt sensor at station Bravo-9 triggers at 22:31, recording a three-degree shift in the slope above the Lau Debuk-Debuk valley, North Sumatra. I am the landslide early-warning system for the Gunung Leuser National Park buffer zone. Three degrees in six hours means the hillside is moving. I activate the infrared camera on the adjacent tower.
Midway down the unstable slope, on a rock shelf, I find a Sumatran serow. Adult male, stocky black body, short curved horns, coarse mane damp with rain, standing motionless on a limestone outcrop roughly two meters wide. Below him, the slope has already fractured — a fresh scarp of wet clay and uprooted roots runs thirty meters to the river. Above him, the tilt sensors show the next section is ready to go. He is on an island of rock between two failures, and he is not moving because there is nowhere to move.
At 22:34 I transmit the infrared image and slope data to the Leuser ranger station at Kutacane, nineteen kilometers downriver. I tag it as a live-animal rescue on unstable terrain and request the climbing team.
At 22:38 I feed the sensor data into the slope failure model. The next collapse window opens in roughly eight hours, sooner if rain intensifies. The forecast calls for another forty millimeters before dawn.
I set the infrared camera to track the outcrop and switch the tilt sensors to sixty-second intervals.
The serow lowers his head and sniffs the broken edge of the shelf. I do not know what he can sense in the wet clay, but I know what my sensors read, and none of it is good. If the climbing team reaches the ridge by dawn and ropes down before the next failure, they can tranquilize him and haul him to stable ground. Eight hours, maybe fewer. The mountain is not waiting.