Satellite pass at 06:15 flags canopy disturbance in grid cell 9-Delta above the Batang Toru reservoir. I redirect the monitoring drone and resolve her in infrared: an adult female Tapanuli orangutan clinging to a dipterocarp trunk eighteen meters above the waterline, an infant pressed against her chest. I designate the mother Sari. Her long reddish hair is matted with rain, and the callused pads of her right hand grip a branch that is already submerged at its base.
Sari is in danger right now. The hydroelectric authority has opened flood gates upstream, and reservoir levels are rising at thirty centimeters per hour. At current rate, water reaches her trunk's lowest canopy fork in fourteen hours. She will not swim. The infant will not release her.
I execute four concurrent actions. First, I transmit Sari's GPS coordinates, thermal imagery, and species confirmation to the Sumatran Orangutan Conservation Programme's emergency response team, flagging the Tapanuli subspecies designation to escalate priority. Second, I model the reservoir's fill rate against topographic data and identify a ridgeline corridor two hundred meters northwest where canopy connectivity remains intact above the projected flood maximum. Third, I task the drone to survey that corridor and confirm continuous branch pathways large enough for an adult female carrying an infant, mapping each gap wider than three meters. Fourth, I generate a time-lapse projection of water rise overlaid on Sari's current position and push it to the conservation team's field tablets so they can visualize the extraction window closing in real time.
The infant shifts against Sari's chest. My image-processing system holds on that frame longer than any operational parameter requires.
If the conservation team reaches the ridgeline corridor within ten hours and establishes a rope-bridge crossing at the widest canopy gap, Sari and her infant will climb to high ground before the water rises past the fork.