SUMMARY The lithospheric response to seamounts and ocean islands has been successfully described by deformation of an elastic plate induced by a given volcanic load. If the shape and mass of a seamount are known, the lithospheric flexure due to the seamount is determined by the thickness of an elastic plate, Te, which depends on the load density and the age of the plate at the time of seamount construction. We can thus infer important thermomechanical properties of the lithosphere from Te estimates at seamounts and their correlation with other geophysical inferences, such as cooling of the plate. Whereas the bathymetry (i.e. shape) of a seamount is directly observable, the total mass often requires an assumption of the internal seamount structure. The conventional approach considers the seamount to have a uniform density (e.g. density of the crust). This choice, however, tends to bias the total mass acting on an elastic plate. In this study, we will explore a simple approximation to the seamount’s internal structure that considers a dense core and a less dense outer edifice. Although the existence of a core is supported by various gravity and seismic studies, the role of such volcanic cores in flexure modelling has not been fully addressed. Here, we present new analytic solutions for plate flexure due to axisymmetric dense core loads, and use them to examine the effects of dense cores in flexure calculations for a variety of synthetic cases. Comparing analytic solutions with and without a core indicates that the flexure model with uniform density underestimates Te by at least 25 per cent. This bias increases when the uniform density is taken to be equal to the crustal density. We also propose a practical application of the dense core model by constructing a uniform density load of same mass as the dense core load. This approximation allows us to compute the flexural deflection and gravity anomaly of a seamount in the wavenumber domain and minimize the limitations recognized from the analytic tests. Then, the dense core model is applied to predict the lithospheric flexure beneath Howland Island in the Tokelau seamount chain; these results are compared with the predictions of the uniform density model. Based on age dating of Howland and the age of the seafloor, traditional Te versus age curves predict the elastic plate thickness beneath the seamount to be around 20 km, which is comparable to the best dense core model of Te = 26 km. However, the best uniform density model is found at Te = 12 km, which is significantly less than the predicted. From our investigations of synthetic and real seamount cases, we conclude that the dense core model approximates the true mass distribution of a seamount better than the uniform density model. Finally, we suggest that the role of underplating in flexure modelling may need to be reexamined because the dense core model predicts substantially less deflections than the uniform density model without requiring additional buoyancy caused by underplated material.
Read full abstract