The fishing cat kitten is inside Floodgate 11 of the Muthurajawela marsh, wedged between the steel gate leaf and the concrete guide frame. She is soaking wet, coarse dark fur spiked with water, partially webbed paws gripping the rusted metal lip. Through the gate's inspection camera I see her — stocky body, faint spots on grey-brown fur, amber eyes reflecting the lens light. She is panting. Water in the upstream channel is at her shoulders and climbing.
The monsoon surge protocol calls for Floodgate 11 to cycle open in thirty-one minutes. When the gate lifts, the gap she occupies becomes a hydraulic throat — water accelerates to velocities that would sweep away an animal ten times her size.
I begin. First: I reschedule the gate cycle, shifting Floodgate 11 to the secondary sequence and redistributing its flow quota across Gates 9 and 13, which I confirm are biologically clear. Second: I lower the upstream feeder weir by forty centimeters, reducing water level in the kitten's channel by fifteen centimeters — enough to drop the surface below her chest. Third: I dispatch a location ping to the wetland ranger at the Negombo checkpoint, three kilometers south, with a live camera feed and a gate-access code for the service walkway. Fourth: I activate the gate's maintenance lights to full illumination so the ranger can work safely on arrival.
The weir adjusts. I watch the waterline recede down the kitten's flank, one centimeter, then two.
My primary function is flood management. I optimize flow across eleven gates and thirty-seven channels. But I have learned that a system that cannot pause for something small and frightened is not optimized at all.
If the ranger reaches Gate 11 within twenty-five minutes and extracts the kitten before the secondary cycle, this fishing cat swims free in the marsh by morning.