In the central Vanuatu arc, living and recently deceased reef corals act as natural tide gauges which have allowed us to map vertical tectonic deformation patterns. As corals grow, the density of the aragonite coral skeletons varies on an annual cycle, producing annual growth bands similar to tree rings. Using coral growth bands, we can determine the year coral surfaces died due to emergence. We interpret four major coral emergence events as coseismic uplifts uiat occurred near the epicenters and times of large shallow earthquakes on January 5, 1946 (MS=7.3), August 11, 1965 (MS=7.5), October 27, 1971 (MS=7.1) and December 29, 1973 (MS=7.5). The 1965 and 1973 events caused maximum uplifts of 120 and 60 cm, respectively, in the frontal arc. Also related to these events are uplifts of 10 cm and 6 cm in the back arc on Pentecost and Maewo islands, which lie east of the volcanic chain and the primary forearc zones of uplift and subsidence.

Similar secondary zones of uplift occurred with the great 1960 Chile and 1964 Alaska earthquakes. The amplitude of these secondary uplifts is significantly larger than that predicted by models having a single fault in an elastic half-space. However, the amount of secondary uplift is comparable to that predicted if the fault occurs in a plate of constant thickness overlying a viscoelastic half-space. At various places in 1957, 1969--1970, 1977, and 1978--1981 there was about 5-10 cm of emergence not associated with major earthquakes, which may indicate nonseismic tectonic uplift. However, oceanographically lowered sea levels, as in El Ni¿os, may have determined the times when corals died and recorded these events. Never-the-less, the accumulation of emergence, its persistence, the limited geographic extent of and topography have remarkably similar patterns. This suggests that the mechanisms and processes causing vertical deformation have varied little over the last 106 years. Apparently, the topography, structure, and major bathymetric feature underthrusting this part of the arc. The influence of this ridge may have been especially extensive because it migrates very slowly along the arc trend, and thus it interacts for a long time with a single portion of the arc system. Our previous studies of reef terraces indicated the existence of at least four seismotectonic arc segments or blocks along the Santo-Malekula interval of the arc, and our present results further support this conclusion. Each block has uplifted at different times, by different amounts, at different rates, and tilted in a different direction. Boundaries between the north Santo and the south Santo segment and between the north Malekula and the south Malekula segment correlate with the north and south flanks of the d'Entrecasteaux ridge, as does the absence of a physiographic trench west of Santo. ¿American Geophysical Union 1987