Analysis of Pacific Hotspot Chains and a Model for Recent Plume Drift

Abstract

Age-progressive seamount trails created by long-lived deep-mantle plumes have been used to establish absolute reference frames for plate motion. However, when plume drift is considered, changes in seamount trail direction and age progression rate cannot be attributed to plate motion change alone. A better understanding of these absolute motions are needed for studying the mantle dynamic processes that drive both plate tectonics and plume drift. For this study, improvements to age-progressive curves of eleven Pacific hotspot chains are made independently of past plate motion models. Our approach involves bathymetry processing to robustly predict a smooth and continuous hotspot path by connecting high points in seamount bathymetry, with uncertainties in the path based on seamount trail width and amplitude. Ages found using radiometric dating techniques from seamount samples are projected onto the inferred hotspot trail. A best-fit model of age as a function of along-track distance is determined, giving continuous age progressions for each seamount chain with uncertainties in both age and path. Three different types of paleolatitudes are also examined by incorporating data from the magnetization of seamount drill core samples, paleo-poles from marine magnetic anomaly skewness, and paleo-spin-axes from shifts in equatorial sediments. Improved paleolatitude curves for the Hawaiian-Emperor and Louisville chains are determined by combining these three different types of data. Paleolatitude curves are also determined for other chains where a sufficient amount of paleo-pole or paleo-spin-axis data are available. The data analysis of these eleven Pacific seamount chains provide prime constraints for future plate and plume motion models. For this study, we examine the data by accessing the change over time in distance between coeval seamounts, which infers relative drifts between the hotspots at different times in the past. We find that the inter-hotspot distance change since 6 Ma have been linear, given the (relatively large) errors, which prompted the development of a novel modeling approach. We end with inverting these linear relative rates via our new modeling scheme and solve for recent plume drifts for the time frame of 6 Ma to the present.