A Two-Phase, Multi-Component Geochemical Model of Mid-Ocean Ridge Magmatism

Understanding melt generation, transport, and crust formation within a mid-ocean ridge context is a compelling challenge in geoscience. These systems are indirectly observable, both spatially and temporally, and our current understanding therefore relies on poorly resolved geophysical imaging and geochemical signatures preserved in erupted products.  Previous numerical studies incorporating two-phase melt transport have greatly improved our understanding of melt migration and focusing beneath mid-ocean ridges [1,2,3]. However, these models typically simplify the treatment of crustal formation and have a limited ability to make a direct comparison between model predictions and observed mid-ocean ridge basalt (MORB) compositions.  

We present a new two-dimensional staggered-grid finite-difference model based on the framework of [3,4]. Implemented in MATLAB, the model is designed to simulate magmatic systems at mid-ocean ridges. The model solves fully compressible solid-state mantle flow coupled to two-phase melt transport and includes a novel multi-component model of mantle melting and crust formation. 

A key advance of this framework is an in-situ melt extraction and crust formation algorithm that conserves mass and enables the development of a crustal layer along the seafloor rather than artificially removing melt from the ridge axis as most previous models do. The model further includes a multi-component model of major, trace, and isotopic composition to understand petrogenesis and geochemical evolution through melt production, focusing, and extraction. This allows for a more detailed comparison with real-world geochemical datasets.

The petrogenesis component of the model is calibrated to allow for the prediction of MORB compositions based on the underlying physical dynamics. This enables us to test the sensitivity of crustal production and composition to variations in physical parameters such as spreading rate, mantle potential temperature, mantle composition, and mantle rheology. Additionally, it allows us to assess whether different melt focusing end members from active to passive flow regimes result in a detectable geochemical signature.

The primary aim of this work is to develop a flexible modelling framework that can be used to explore the parameter space governing passive and active melt focusing and understand how mantle and melt dynamic regimes are expressed in petrological and geochemical observables. 

[1] Katz, 2008: https://doi.org/10.1093/petrology/egn058 

[2] Katz, 2010: https://doi.org/10.1029/2010GC003282

[3] Keller et al., 2017: https://doi.org/10.1016/j.epsl.2017.02.006 

[4] Keller and Suckale, 2019: https://doi.org/10.1093/gji/ggz287