Long‐Lasting Evolution of Layer 2A in the Western South Atlantic: Evidence for Low‐Temperature Hydrothermal Circulation in Old Oceanic Crust

Abstract

Previous seismic studies suggest that hydrothermal processes are active only within young oceanic crust (<10–16 Ma). However, differences between measured and predicted heat flow at the ocean floor indicate that hydrothermal fluids may be transporting heat advectively in crust up to ages of 65 Ma. We report on seismic velocities of 0–71 Ma slow to intermediate spreading rate upper crust in the western South Atlantic. Thirteen high‐resolution 2‐D velocity models were built using traveltime tomography on downward continued streamer data acquired during the Crustal Reflectivity Experiment Southern Transect. In the Crustal Reflectivity Experiment Southern Transect area, velocities at the top of seismic layer 2A increase rapidly from ~2.4 km/s at 0 Ma to ~4.2 km/s at 6 Ma and then undergo a more gradual increase to ~4.9 km/s at 58 Ma. These new results resolve the long‐standing debate about the duration of interaction between ocean crust and seawater, providing seismic evidence for hydrothermal circulation continuing to crustal ages predicted by heat flow studies. Seismic layer 2B does not show a systematic off‐axis velocity trend but has an average value of 5.15 km/s. We interpret this result to indicate that the hydrothermal system becomes too shallow to affect the physical properties of layer 2B shortly after crustal accretion. Upper crustal heterogeneity in ridge‐parallel profile orientation is more pronounced for crust accreted at slow spreading rates, compared to intermediate rates. This result is consistent with shorter magmatic segments at slower spreading rates, increasing the frequency of tectonic and magmatic accretion alternately occurring along the ridge.

Seismic Layer 2A: Evolution and Thickness From 0‐ to 70‐Ma Crust in the Slow‐Intermediate Spreading South Atlantic

“…Our layer 2A thickness estimates contain significant variability, which introduces a broad distribution and substantial scatter when examining trends. The values of layer 2A thickness reported by Kardell et al () contain a similar standard deviation as ours, but their results do not include the rapid variation we observe (Figures and ). We believe two primary contributing factors to be predominantly responsible for the scatter we observe in our thickness values.…”

“…Furthermore, our thickness range roughly brackets those of previous studies, indicating our values are reasonable. Kardell et al () found slightly different layer 2A thickness values along the CREST transect where they obtained velocity models. We believe the difference is a product of both data and methods.…”

“…The lack of continuity of the layer 2A event we observe, which results in the gaps in our thickness data (Figure ), is possibly a result of the roughness of the basement interface at the slower spreading rates. Also possible is that the sharp gradient change necessary to produce the triplication may not be present everywhere, even when the absolute magnitude of velocities seen in the streamer tomography models of Kardell et al () is indicative of layer 2A. Kardell et al () report that their models cannot resolve the rapid gradient change needed to produce the triplication we use to estimate the base of layer 2A, so further determination of whether or not the gradient change exists where we do not image the layer 2A event will require higher‐resolution velocity modeling.…”

“…To convert seismic reflection two‐way travel time values to thickness and depth in meters, we used velocities extracted from the streamer tomography models of Kardell et al (), which are derived from the same CREST MCS profiles used in this study. We converted the arrival time of our layer 2A event to depth then extracted the mean vertical velocity between the top of the igneous basement and the layer 2A event at every CMP where the event was present in the velocity model.…”

“…We converted the arrival time of our layer 2A event to depth then extracted the mean vertical velocity between the top of the igneous basement and the layer 2A event at every CMP where the event was present in the velocity model. To reduce scatter and to extend velocities across the length of the transect, we fit a sixth order polynomial curve to the equally weighted velocity values of Kardell et al (), which provided an estimated mean layer 2A velocity at every shotpoint along the transect (Figure ).…”