Sway aerodynamics of urban tree in response to strong winds

Abstract

This study examines the sway aerodynamics of a Delonix regia (Flame of Forest), a common urban tree species, at the University of Hong Kong campus. The study area is proximal to the central business district. The sway aerodynamics is quantified in time-domain and frequency-domain. The trunk and four branches were selected for monitoring during a four-day period of tropical-cyclone strike. Three branches bear structural defects which includes crossing, cavity and open wound.

Wind exerts a wide range of impacts on trees. Sway aerodynamics measure the response of various parts of a tree to wind. Most studies covered trees growing in natural landscape and coniferous species in the mid-latitude. Trees growing in urban landscape have received little attention. The sway aerodynamics of deciduous species, particularly in the tropics, is inadequately understood. The effects of physical compromises on sway aerodynamics of the branches demand research.

Wind speed was measured by an ultrasonic anemometer. Tree Motion Sensor(TMS), a precision electronic transducer, measured the three-dimensional sway motion. TMS were mounted at two positions on each sampled branch. Data were analyzed with interpolation which enhanced the time-series resolution. The original time-series data was converted to amplitude which offered an effective approach for Fast Fourier Transform to prepare the data for frequency-domain analysis.

The computation procedure of amplitude served to suppress the effect of remnant sway motion and the delay between two signals. The amplitude approach provided a sensitive monitoring of sway response to wind, and differentiate the sway characteristics of different branches. The difference is attributed to variations in branch dimension and configuration. Sway amplitude is greater at further distance from branch base, with more slender growth form and during greater wind amplitude. Physical compromise did not bring discernible response in sway amplitude.

The natural frequency of the monitored branches was estimated to be 0.5 ̶ 0.6 Hz. With frequency-domain analysis, the role of mass and foliage distribution in sway aerodynamics was identified. Sway frequency at different positions on a branch might differ, and they were mainly positively related. Branch dimension is generally related to sway frequency, but the relationship could be positive or negative along different sway-motion axes. Physical compromise was not related to sway frequency. Frequency-domain analysis can be applied to wind speed. Wind frequency could not explain variations in sway frequency. The urban-fabric effect on wind frequency demands more exploration. Sway aerodynamics can add a new dimension to urban tree-risk assessment. Frequency-domain and the amplitude-approach time-domain analyses can supplement the widely used visual tree assessment. The tolerance level of physical compromises can be established with a firm and objective scientific basis. Because wind frequency may influence sway aerodynamics, tree-stability assessment may incorporate the evaluation of wind environment around the subject trees.