Oceanic plateaus and swells are a major component of the seafloor topography, yet they remain among the most poorly understood features. This is especially true of the oceanic plateaus which show large variations in crustal thickness. To determine the depth and mode of compensation for 53 of the largest plateaus and swells, we analyzed the relationship between geoid height and topography in polygonal areas containing each feature. Both geoid height and topography were first band‐pass filtered (400 km < l < 4000 km) to isolate the signal associated with local compensation from flexural and deep mantle signals. The ratio of geoid height to topography was then determined by fitting a straight line to the data. Except for nine of the smaller features there is a high correlation between geoid height and topography that is positive in accordance with Airy and thermal compensation models. Eighteen features have high geoid/topography ratios that cannot be explained by the Airy compensation model of crustal thickening. These features (thermal swells) are partially supported by thermal buoyancy forces in the lower half of the lithosphere. The ratios are highest for active hot spot swells and decay, with the thermal age of the swell, to values consistent with Airy compensation of the enduring volcanic edifice. The remaining features (plateaus) have lower geoid/topography ratios in agreement with the Airy compensation model. Those plateaus with average height greater than 4 km are thought to be continental fragments; the shorter plateaus tend to be volcanic features. Modified continental plateaus, presumably small fragments of extended and intruded continental margin crust, cluster around heights of ∼3 km, overlapping the range associated with oceanic plateaus. Since the origin of many plateaus is poorly understood, this global geoid/topography analysis provides a new technique for comparing the deep structure of oceanic plateaus and swells.

SUMMARYIncompatible elements and isotopic ratios identify three endmember mantle components in oceanic island basalt (OIB); EM1, EM2, and HIMU. We estimate compatible to mildly incompatible transition metal abundance trends (Ni, Co, Fe, Cu, Cr, V, Mn, Sc, and Zn) in 'primitive' basalt suites (Mg# = Mg/(Mg + 0.9*Fe) atomic = 0.72) from 12 end-member oceanic islands by regressing metals against Fe/Mg ratios in sample suites, and solving for concentrations at Mg/Fe = 1 (Mg# = 0.72). Using the transition metal estimates, exploratory statistics reveal that islands 'group' based on mantle component type even when La/Yb ratios are used to compensate metal concentrations for percentage melting. Higher chalcophile Zn (and Pb, earlier work) in EM1 and EM2 compared to HIMU, and higher Cr (3+) and Sc in HIMU relative to EM1, support views that HIMU represents subductionprocessed ocean floor basalt. Incompatible elements, ratios and isotopes indicate that EM1 is Archean, EM2 is Proterozoic or younger, and both are related to sediment subduction. As found with incompatible elements, EM1 and EM2 show similar 'compatible' element concentrations, but lower (multivalence) Cr, Fe and Mn in EM1 could indirectly reflect increasing oxidation of subducted sediment between the Archean and Proterozoic. Alternatively, changes in subduction processes that yielded peak continental formation in the Neoarchean, and craton-suturing in the Paleoproterozoic may account for EM1-EM2 differences. EM1 shows similar or lower Cr, Ni and Co compared to HIMU and EM2 suggesting that economic viability of layered intrusions, which have extreme EM1-like signatures, is unrelated to high metals in EM1 mantle sources, but that high % melting appears important. Because core-concentrated transition metals correlate with mantle component type, lithospheric recycling apparently controls their concentrations in OIB and core-mantle interaction may be unimportant. RÉSUMÉLes éléments incompatibles et les rapports isotopiques permettent de délimiter trois termes extrêmes de composants mantéliques dans des basaltes insulaires océaniques (OIB), soit EM1, EM2, et HIMU. Nous estimons les tendances d'abondance de métaux de transition (Ni, Co, Fe, Cu, Cr, V, Mn, Sc, and Zn) compatibles à modérément incompatibles dans des suites de basaltes « primitifs » (Mg# = Mg/(Mg + 0,9*Fe) rayon atomique = 0,72) sur 12 termes extrêmes de matériaux insulaires océaniques, par régression des concentrations des métaux sur les rapports Fe/Mg dans des échantillons des suites, la détermination étant définie au rapport Mg/Fe = 1 (Mg# = 0,72). L'utilisation d'une approche statistique exploratoire sur les estimations de métaux de transition montre que la composition des îles se « regroupent » en fonction du type de composition du manteau, cela même lorsque les ratios La/Yb sont utilisés pour compenser les concentrations de métaux pour déterminer le pourcentuel de fusion. Le caractère plus chalcophile du Zn (et Pb, travail antérieur) dans EM1 et EM2 comparé à HIMU, et la plus grande teneur en Cr ...

In the summer of 1978 the Iceland Research Drilling Project undertook the drilling of a deep crustal hole near Reydarfjörder in eastern Iceland. As a part of this project, the Scripps Institution of Oceanography and the University of Washington undertook a small‐scale seismic refraction experiment near the drill site in an attempt to compare surface geophysical measurements with observations of samples from and logging in the hole. Using recent advances in the methods of extremal inversion of seismic data, we have determined an approximate one‐dimensional velocity structure for the drill site. This structure indicates that the 1.9‐km hole failed to penetrate the layer 2 layer ‐3 transition which was at some 3.0–4.5 km beneath the drill site. The transition appears to be rather abrupt, unlike that beneath the ocean, with velocity increasing from 5.2–5.5 km/s in the upper crust to about 6.7 km/s in layer 3. We observe a steep eastward dip and a shallow westward dip in the lower crust away from the nearby Thingmúli and Reydarfjördur volcanic centers, respectively, in agreement with previous work associating shallow depths to layer 3 with Tertiary volcanic centers as a result of increasing metamorphic grade and increased dyke swarm intensity.

No abstract