Supplementary information files for "Seeing gusty winds: Optical tracking of tree motion captures peak 3-second gust velocities"

Supplementary files for article "Seeing gusty winds: Optical tracking of tree motion captures peak 3-second gust velocities"<br><br>Peak 3-second gust speed (𝑠3) is a key meteorological metric for potentially impactful winds. It is widely used in meteorology and engineering, but traditional anemometry is limited by sparse measurements. This proof-of-concept study is the first to relate Optical Flow Tracking Velocimetry (OFTV) derived wind estimates explicitly to 𝑠3, by capturing fine-scale wind variability from video footage of trees. While calibration is required, this approach is simple, transparent and avoids the descriptor-based Beaufort scale. So, it is a potential improvement or a component of frameworks that use machine learning or physics-based structural approaches. Results from a pear tree in a domestic setting show correlations between OFTV-𝑠3 and cup-anemometer data (R2 = 0.65, p &lt; 0.05), confirming that 𝑠3 can be extracted using only OFTV. Interestingly, OFTV-𝑠3 are consistent and strongly correlated (R2 ≥ 0.92) across a range of reference objects (i.e., a flag, parts of the tree crown, grass). This insight suggests that identifying specific objects is unnecessary in simple scenes, potentially simplifying future crowd-sourced visual anemometry, although it may still be required in cluttered urban environments. Furthermore, 𝑠3 occur across objects within ±1–5 s of each other, indicating a robustness of measurement. Collectively, these insights are a valuable step towards improved wind distribution mapping (e.g. with a nation’s road network CCTV cameras) and detailed site hazard assessments for the (re)insurance sector.<br><br>© The Author(s), CC-BY 4.0