Seismic ocean thermometry

Despite much progress in recent decades, sampling the ocean with sufficient spatial and temporal resolution to map the uptake of anthropogenic heat remains a challenge. Large-scale and long-term trends are often much weaker than small-scale and short-term fluctuations. Data was particularly sparse before the implementation of the Argo program and is still sparse below 2000 m depth. Here we propose to augment the existing observing system with a novel and low-cost method that uses acoustic waves generated by natural earthquakes. As in previously employed acoustic thermometry, acoustic travel times are integrating — and thus intrinsically averaging — measurements of temperature anomalies along the waves’ path. We apply this method to a 3000 km long section in the East Indian Ocean that was insonified by the 2005 Nias–Simeulue earthquake and its aftershock sequence. Over the period 2005–2016, we find a warming trend and seasonal fluctuations that match Argo data, demonstrating the feasibility and accuracy of the approach. This seismic thermometry has the potential to accurately measure large-scale ocean temperature change back to the 1990's, when modern global seismic data became available. The very low frequency seismic waves (around 2 Hz) have considerable sensitivity to the ocean below 2000 m depth. Our reliance on natural sources avoids any potential interference with marine life.