CYPRESS-DOME-WATCH · GeoAI & Remote Sensing Lab

Reading water from the shape of a forest

Across Big Cypress National Preserve, bald cypress grow in round domes: short trees at the dry edge, the tallest at the deep, wet center. From above, a dome's height profile is a map of the water beneath it. This project finds the domes automatically in airborne LiDAR.

The average dome

Every candidate scaled to the same size and stacked: mean canopy height from the centre out past the edge. The tallest trees stand in the middle and the canopy falls away steadily toward the rim. A tall centre is partly built in (the detector starts from high points), but the smooth slope is the dome shape, and the idea this project is testing is that it tracks the water: deepest and longest-lasting in the middle. Next step: check that against the water itself, from radar.

How it works

USGS 3DEP LiDAR is a laser survey flown from aircraft: billions of points, each classified as ground, vegetation or something else. The canopy model is streamed straight from the USGS archive a square kilometre at a time, with no tiles downloaded. For each, the ground points become a bare-earth surface, and every other point gets a height above it. Keeping the tallest point in each square meter gives a canopy height model.

The detector smooths that model to dome scale, finds local high points, grows each into a region, and keeps the ones that are round and dome-sized. Next: Sentinel-2 water indices for each dome, and a before-and-after look at the February 2026 Big Cypress fire.