Investigating the response of hydrothermal convection to decadal cycles of magmatic inflation at the East Pacific Rise, 9º50'N

Hydrothermal circulation at the axis of fast-spreading mid-ocean ridges is intrinsically linked to magmatic activity, which typically fluctuates on decadal time scales, i.e., the characteristic recurrence time of eruptions. While hydrothermal vent temperatures are known to fluctuate in response to sudden events such as dike intrusions or seismic swarms, their response to longer-term processes such as the replenishment of an axial melt lens (AML) remain poorly documented. Here we focus on high-temperature vents from the 9°50'N segment of the East Pacific Rise, which experienced eruptions in 1991/1992 and 2005/2006, and has been extensively monitored over the last 3 decades. There, a compilation of legacy data complemented by recently acquired temperature measurements from the Bio9 vent site (cruise AT50-21, February-March 2024) reveal decadal trends where maximum vent temperatures increase by ~30ºC in ~15 yr between eruptions, and drop by a commensurate amount within a few years of each eruption. In this study we use numerical models of hydrothermal convection to test the hypothesis that decadal increases in vent temperatures are caused by AML inflation pressurizing the upper crust and decreasing its permeability.

We simulate 2-D porous convection driven by a constant basal heat flux, where permeability decreases exponentially with pressure, as suggested by rock deformation experiments. We first benchmark the relationship between average maximal vent temperature and mean permeability against the analytical model of Driesner (2010). Then, we perturb the permeability field using a mechanical model of sill inflation that imparts isotropic compression across the upper oceanic crust, resulting in exponentially-decaying permeability above the 1.5 km deep AML. When using a narrow basal heat source, we obtain a single plume of rising hot fluid, whose flow progressively slows down in the basal conductive boundary layer. This creates a positive thermal anomaly which is then advected to the seafloor by the plume. However, when the heat source is broader and the convection geometry more intricate, variations in permeability modify fluid pathways, leading to a more complex response. Lastly, simulating cycles of AML inflation and deflation yields oscillations in vent temperatures with periods representative of the duration of a replenishment cycle, but with a lag strongly modulated by the vigor of the convective system.