Visual anemometry: functional dependence of leaf fluctuating speed on impinging flow

High-resolution, near-ground wind data is critical for improving the accuracy of weather and climate models, supporting wildfire control efforts, and ensuring safe aircraft takeoffs and landings. However, current wind sensing (anemometry) techniques rely on physical installation at specific locations and only provide single-point or line measurements, which limits their ability to offer the global coverage needed. While traditional methods like the Beaufort scale offer qualitative wind estimates based on visual cues like smoke movement and tree sway, their reliance on subjective interpretation hinders precise data collection, limiting their usefulness for robust predictions. I will present a physics-based single-camera visual anemometry method for quantifying wind speeds observed across an entire camera's field of view. This approach leverages the motion of vegetation, like trees and grasses, as natural wind sensors. Previous methods require a reference anemometer on site, but this new physics-based approach eliminates the need for the reference anemometer at wind speeds under 20 meters per second.  We discover the dominant physics driving the observed vegetation fluctuations, which is used to collapse data from vastly different plants onto a single master curve. The resulting wind speed estimates approach the theoretical accuracy limit imposed by atmospheric turbulence.