Spectral Analysis of Dynamic Wind Loads on Trees

The issue of tree instability and failure, particularly during windstorms or wind gusts, is of great concern. Its extensive impact on public safety is becoming more widely acknowledged. When blowing at critically high speed, the kinetic energy carried by winds, loading onto trees, may forcefully exceed tree strength. This may cause the stem to break, the trunk to split, or the entire tree to be uprooted. Even if surviving bursts of strong wind, the structural elements of trees may have undergone irreversible plastic deformation. Permanently damaged trees have a greater risk of failure and they need to be properly trimmed, pruned, or removed. Realizing these dangerous wind effects, it follows that the assessment of dynamic wind loading on tree stress and stains is a key to preventing tree hazards. We have developed a small-scale inertial measurement sensor unit on a single-board microcontroller. Attached to a tree branch, it constantly measures a time series of oscillatory motion of the branch in response to natural wind forces. Decomposing this signal into its constituent spectral components, the tuned mass damping effects of tree sways can effectively be estimated. Our experiments verified and validated that the wind directions can be determined, relative dynamic wind loads can be approximated, types of trees can be categorized, and the signal energy of tree swaying accelerations can be analyzed. Affordability and ease of use of our device make it highly practical to monitor not only multiple trees at once but also many branches of a single tree. Resulting analysis benefits tree management teams and landscape architects, assisting them in assessing which trees to prune and when.