Hurricanes represent an important natural disturbance process to tropi
cal and temperate forests in many coastal areas of the world. The comp
lex patterns of damage created in forests by hurricane winds result fr
om the interaction of meteorological, physiographic, and biotic factor
s on a range of spatial scales. To improve our understanding of these
factors and of the role of catastrophic hurricane wind as a disturbanc
e process, we take an integrative approach. A simple meteorological mo
del (HURRECON) utilizes meteorological data to reconstruct wind condit
ions at specific sites and regional gradients in wind speed and direct
ion during a hurricane. A simple topographic exposure model (EXPOS) ut
ilizes wind direction predicted by HURRECON and a digital elevation ma
p to estimate landscape-level exposure to the strongest winds. Actual
damage to forest stands is assessed through analysis of remotely sense
d, historical, and field data. These techniques were used to evaluate
the characteristics and impacts of two important hurricanes: Hurricane
Hugo (1989) in Puerto Rico and the 1938 New England Hurricane, storms
of comparable magnitude in regions that differ greatly in climate, ve
getation, physiography, and disturbance regimes. In both cases pattern
s of damage on a regional scale were found to agree with the predicted
distribution of peak wind gust velocities. On a landscape scale there
was also good agreement between patterns of forest damage and predict
ed exposure in the Luquillo Experimental Forest in Puerto Rico and the
town of Petersham, Massachusetts. At the Harvard and Pisgah Forests i
n central New England the average orientation of wind-thrown trees was
very close to the predicted peak wind direction, while at Luquillo th
ere was also good agreement, with some apparent modification of wind d
irection by the mountainous terrain. At Harvard Forest there was evide
nce that trees more susceptible to windthrow were felled earlier in th
e storm. This approach may be used to study the effects of topography
on wind direction and the relation of forest damage to wind speed and
duration; to establish broad-scale gradients of hurricane frequency, i
ntensity, and wind direction for particular regions; and to determine
landscape-level exposure to long-term hurricane disturbance at particu
lar sites.