Lower Crustal Cooling Rates at Slow-Spreading Ridges

The cooling rate, and thus thermal structure, of newly formed lower oceanic crust is poorly constrained despite the constraints on the vigour of hydrothermal circulation, magmatic accretion, and rheology that a better understanding would provide. We have recently developed a method to constrain the cooling rate of samples of oceanic plutonics [Coogan et al. 2002]. This is based on measuring the closure temperature of down-temperature diffusive exchange of Ca from olivine to clinopyroxene and modelling this to determine the cooling rate at high, sub-subsolidus, temperatures (600-900^oC). Applying this methodology to the Oman ophiolite, thought to have formed at a fast-spreading ridge reveals a systematic decrease in cooling rate (of >4 orders of magnitude) with depth from the uppermost gabbros to the petrological Moho. In contrast, ODP Hole 735B from the ultra-slow-spreading SWIR shows generally more rapid cooling rates and no variation in cooling rate with depth over its 1500 m length. ODP Hole 923A from the MAR cooled at approximately the same rate as Hole 735B. The variation in cooling rate with depth in the plutonic section in the Oman ophiolite can be explained by a half-space cooling model although uncertainty in the diffusion coefficient for calcium in olivine prevents quantification of the importance of hydrothermal heat extraction. The constant cooling rate throughout the entire 1500 m of ODP Hole 735B is not consistent with a plate cooling model. Instead, this can be explained if the SWIR lithosphere is thick and cold and that plutons are emplaced into it intermittently. In this scenario, isotherm locations migrate in space and time and crystal subsidence is not required within the lower crust. This difference between the plutonics in Oman and at the SWIR suggest that lower crustal accretion and/or hydrothermal circulation must operate very differently at fast- and slow-spreading ridges. It also supports models in which the lower crust is accreted during discrete magmatic episodes at slow-spreading ridge. These magma chambers must crystallise and cool in place, cooling to well below the closure temperature of calcium exchange between olivine and clinopyroxene, prior to the emplacement of the next intrusion. Small plutons will rapidly freeze meaning that pluton formation must be coupled to rapid dyke and lava emplacement within the short time that eruptible partial melt exists in the lower crust. The rapid cooling of the lower crust at slow-spreading ridges also means that ductile deformation can only occur transiently within this region.