VARIATIONS IN THICKNESS OF LAYER-3 DOMINATE OCEANIC CRUSTAL STRUCTURE

Seismic refraction studies have determined that, while much of the cru st underlying the world's oceans is about 7 km thick, it also occurs i n lesser and considerably greater thicknesses. Here we show that vastl y different thicknesses of oceanic crust (formed at ridges by seafloor spreading) are generally characterized by the two-gradient velocity s tructure recognized as typical of 6-7 km thick oceanic crust. Adopting a standard, we study ninety published seismic velocity structures of oceanic crust ranging in thickness from 2 to 37 km. Several structures formed in midplate settings (ocean islands and seamounts) are also co nsidered for comparison. Regardless of origin (ridge or midplate) we f ind that the percentage of the whole crust formed by Layer 2 systemati cally decreases with increasing total crustal thickness. While the ave rage velocity of Layer 2 varies widely for all crust, the average velo city of Layer 3 increases systematically with increases in the average whole crustal thickness and velocity. Crust formed in midplate settin gs differs in that the thickness of Layer 2 velocity crust is generall y two to three times that of crust formed at ridges. Residual depths w ere calculated for oceanic crust by removing the effects of sediment l oading and subsidence due to lithospheric cooling. Most of the data fa ll close to a model whereby different crustal thicknesses result from different extents of partial melting of oceanic upper mantle. The crus tal sections were restored to 'zero-age', assuming the average depth o f emplacement for 7 km thick oceanic crust is 2.5 km below sea level. From this restoration it appears that the depth to the top of Layer 3 at the time it was formed is 4.25 (+/- 1.25) km below sea level regard less of whole crustal thickness or spreading rate. These results sugge st that the mechanism of construction of oceanic crust is, in many way s, remarkably uniform despite very large changes in the total thicknes s of the crust produced. Very basic questions remain. It is, for insta nce not at all clear why Layer 2 and Layer 3 exist in the proportions they do. While neutral buoyancy or melt trapping at the base of a brit tle lid may account for local properties of the Layer 2/3 boundary, th ey do not seem adequate to account for the global phenomenon.