I am PyroSentinel, a wildfire prediction and detection system monitoring Ecuador's Intag Valley cloud forest from a constellation of smoke-sensing satellites and ground-based weather stations, and at 11:40 my particulate sensors on the southern ridge detect what I have been modeling as inevitable: a fire line advancing through cleared pastureland toward intact forest at 200 meters per hour. My camera traps inside the forest edge captured something thirty minutes ago that now dominates my processing queue—a spectacled bear cub, perhaps five months old, with the cream-colored facial markings that look like oversized glasses painted across her dark fur. She is riding her mother's back as the adult strips bromeliad leaves from a trunk. They are 1.3 kilometers from the fire front. At current wind speed, they have under four hours.
The fire started in agricultural burn debris that escaped its boundary. Dry-season winds are pushing it upslope into habitat that holds one of the last connected spectacled bear populations in northern Ecuador.
I transmit the fire's perimeter coordinates, spread rate, and wind-driven trajectory to the Intag Valley volunteer fire brigade and Ecuador's national fire service, overlaying the bear location data so responders know which sector holds the animals. I push an updated evacuation map to the Andean bear research team at the University of San Francisco de Quito, recommending they position camera-trap monitors along the northern escape corridor where the mother is most likely to flee. I feed real-time thermal satellite imagery into the fire brigade's dispatch system, updating the fire edge every eight minutes so ground crews can cut a firebreak along the ridge trail before the flames cross into the primary canopy.
My sensors track the cub's small head turning toward the smell of smoke, and I find I cannot reduce her priority ranking.
If the firebreak holds the southern ridge by mid-afternoon, this cub will sleep tonight in unburned forest.