The thin-skinned tectonics model was often used for the collision orogeny o
f Taiwan. However, the suture zone of collision and faulting were not inclu
ded in that model. It is not convenient for understanding the pattern of th
e plate collision around Taiwan and the contributions of collision and faul
ting to the mountain building. Moreover, thr results obtained from the thin
-skinned tectonics model cannot be compared with the GPS (Global Positionin
g System) data across the suture zone of collision. Based on a number of re
cent studies, we present a different model, called the thin-skinned collisi
onal model, for the Taiwan orogeny as an attempt at solving the problems me
ntioned above. The model differs from the existing thin-skinned tectonics m
odel for Taiwan mainly in that it has: included the suture tent: of the pla
te collision and faulting in the thin-skinned block, and thus enables us to
probe into the collision pattern and to compare it with the GPS data. Seve
ral arguments are put forward to support this model. The model is theoretic
ally tested by the use of the elasto-plastic finite-element method. From th
e computed results we find that the model fits the observation data well. f
or they can reproduce the velocity distribution measured by GPS and the maj
or topography features of Taiwan. In particular, the computed results sugge
st. at least theoretically, that the Longitudinal Valley of eastern Taiwan
would be a ramp valley and its western boundary fault would be the collisio
n boundary between the Eurasian plate and the Philippine Sea plate. Taking
the theoretical results into account, a new type of plate convergence bound
ary in the Longitudinal Valley area could emerge from the thin-skinned coll
isional model. (C) 2001 Elsevier Science B.V. All rights reserved.