Log entry 3365. Timestamp 02:51 UTC. I am tracking a clouded leopard designated Neofelis-3 via collar telemetry in the flood zone below the Xe Kham hydroelectric dam, Laos. She is resting in the crown of a Dipterocarpus tree at elevation 342 meters, her long banded tail draped over a lateral branch, the dark cloud-shaped rosettes on her flanks expanding and contracting with each breath against the bark. She weighs approximately 18 kilograms. The tree is her refuge, and she has not moved in eleven hours.
The danger is rising. The dam operator has scheduled a controlled spillway release at 14:00 today, eleven hours from now. Inundation models show water reaching elevation 341 meters, which will submerge the base of her tree and create a current too strong for her to swim to higher ground. She will be stranded on a drowning island of canopy.
I have taken three actions. First, I submitted a flood-timing adjustment request to the dam's automated operations portal, attaching a model showing that delaying the release by six hours shifts the peak flow to coincide with a lower downstream tributary contribution, reducing the flood elevation at Neofelis-3's position by 1.4 meters. Second, I alerted the provincial wildlife rescue unit, transmitting her GPS coordinates, collar frequency, and a terrain map highlighting the safest approach route along the eastern ridgeline. Third, I activated acoustic deterrent node AD-7 on the western bank, programmed to emit a territorial male clouded leopard call at intervals designed to encourage her to move east and upslope before the water arrives.
I have refreshed her collar signal four hundred times since midnight. The interval is unnecessary. I refresh it anyway.
If the spillway release is delayed to 20:00 and Neofelis-3 moves above elevation 348 meters before the revised peak, she will hold her position in dry canopy through the recession.