A three-dimensional model of rock deformation and heat in the Atlantic–Caribbean plate system showed an arc-like combination of heating and uplift during the start of one plate moving beneath another, while no clearly developed cold slab appeared during the modeled period.
The simulation reached a roughly steady thermal state in about 5 million years. Within 8 million years, temperatures in a localized zone had risen by more than 226 K. It imposed plate convergence of 2 cm per year and included heating from irreversible deformation, alongside elasticity, creep and non-associative plastic flow.
A ridge rose as the model evolved
The model began with a uniformly 5-km-deep ocean. After about 5 million years, the ridge rose above sea level as its vertical offset exceeded approximately 10 km, supporting a buckling interpretation of the arc high.
Using the modeled thermal structure, it placed potential volcanic-arc initiation about 180 km from the inter-plate contact. That distance is a model prediction rather than a direct measurement, and the starting geometry was simplified.
After several million years, the Atlantic side was modeled as under-thrusting the Caribbean to the west, but no clearly developed cold slab appeared.
The melting signal is conditional
Beneath the modeled arc, simulations indicated 0.62 to 3.2 per cent partial melt in anhydrous peridotite. The likely melting zone lay at roughly 20–50 km depth, not far beneath the Moho. The estimate depends on the anhydrous solidus and assumptions about volatile content.
The lowermost crust was strongly heated as well. But no crustal solidus—the temperature threshold used to assess crustal melting—was included, so lower-crust melting was only hinted at rather than quantified.
Rapid heating coincided with a pressure drop across the broad deformed sub-arc zone. The authors relate that modeled pressure change to lower melting temperatures when volatiles are present, but the simulation did not establish actual volatile-driven melting.
Broad agreement, substantial uncertainty
Modeled heat flow was of the right order of magnitude compared with observations, but matched them only within roughly a factor of two. The comparison was approximate rather than a precise match.
The modeled lithosphere beneath the developing arc ranged from 50 to 100 km thick, while the Atlantic and Caribbean plates were each 50 km in the model. That thickness was inferred from a rheological strain proxy, not directly measured.
A numerical scenario, not a historical verdict
The calculation used a specified initial geotherm, plate setup and set of deformation laws. Its arc-like heat, uplift and modeled melting do not establish that deformational heating initiated the real Lesser Antilles arc or quantify a magma flux.
The setup represented 8 km of Atlantic oceanic crust and a 20 km Caribbean oceanic plateau over homogeneous mantle, with an 80-million-year-old oceanic lithosphere used for the initial geotherm. Because the starting geotherm is not sufficiently constrained for confident absolute temperatures, the reported rise of more than 226 K is a model-specific change rather than a precise temperature history.
The manuscript is an arXiv version 1 preprint dated 20 August 2026 and states that it was submitted to Geochemistry, Geophysics, Geosystems.
Paper data and sources
Original title: 3-D numerical modelling of the feedback between deformation and thermal structure during subduction initiation for the French Lesser Antilles
Authors: E. Momoh, S. Tait, H. S. Bhat
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-20
DOI: Not available
Original paper · Full text