Modelling Wind Damage to Southeastern U.S. Trees: Effects of Wind Profile, Gaps, Neighborhood Interactions, and Wind Direction

Abstract

Tree damage from a variety of types of wind events is widespread and of great ecological and economic importance. In terms of areas impacted, tropical storms have the most widespread effects on tropical and temperate forests, with southeastern U.S. forests particularly prone to tropical storm damage. This impact motivates attempts to understand the tree and forest characteristics that influence levels of damage. This study presents initial findings from a spatially explicit, individual-based mechanistic wind severity model, ForSTORM, parameterized from winching research on trees in southeastern U.S. This model allows independent control of six wind and neighborhood parameters likely to influence the patterns of wind damage, such as gap formation, the shape of the vertical wind profile, indirect damage, and support from neighbors. We arranged the subject trees in two virtual stands orientations with identical positions relative to each other, but with one virtual stand rotated 90 degrees from the other virtual stand – to explore the effect of wind coming from two alternative directions. The model reproduces several trends observed in field damage surveys, as well as analogous CWS models developed for other forests, and reveals unexpected insights. Wind profiles with higher extinction coefficients, or steeper decrease in wind speed from canopy top to lower levels, resulted in significantly higher critical wind speeds, thus reducing level of damage for a given wind speed. Three alternative formulations of wind profiles also led to significant differences in critical wind speed (CWS), although the effect of profile was less than effect of different extinction coefficients. The CWS differed little between the two alternative stand orientations. Support from neighboring trees resulted in significantly higher critical wind speeds, regardless of type of wind profile or spatial arrangement of trees. The presence or absence of gaps caused marginally significant different in CWS, while inclusion of indirect damage along with direct damage did not significantly change CWS from those caused by direct damage alone. Empirical research that could most benefit this modelling approach includes improving crown area measurement, refining drag coefficients, and development of a biomechanical framework for neighbor support.