CoAuthors

Lana Opel (University of Bonn, Germany); Michael Schindelegger (University of Bonn, Germany)

A number of investigations have reported ocean stratification increasing over the past half-century (e.g., Li et al., Nat Clim Change, 2020). A possible byproduct of stratification changes is a corresponding change in barotropic-to-baroclinic tidal conversion, a process known to account for about a third of the M2 barotropic energy loss. Increasing stratification could potentially induce an increase in this tidal conversion, with a tendency for amplified internal tides at the expense of the surface tide. We here report that such changes in both barotropic and baroclinic tides are indeed occurring, as revealed by three decades of satellite altimetry and, covering the same time period, 3-D numerical ocean simulations of the tides. The latter can delineate how the tides change in response to changes in stratification versus changes from sea level rise; in fact, stratification is found to dominate. Our altimeter analyses are based strictly on Topex-Poseidon, Jason, and Sentinel-6 data on the primary ground-track. The altimeter results confirm, for the most part, some initial reports by Bij de Vaate and colleagues on the barotropic tide and by Zhao (see OSTST poster) on the baroclinic tide. Altimetric trends in the barotropic tide are prone to potentially large systematic errors, an especially significant one arising from tidal leakage in the DAC correction, which is partly caused by errors in ECMWF atmospheric tides. Our numerical results are based on annual MITgcm simulations with stratification imposed from the GLORYS12 reanalysis. For the 30-y trend in the M2 barotropic tide, altimetry and model simulations are surprisingly consistent, both showing open-ocean amplitude trends predominantly negative, roughly 0.1-0.2 mm/y (more in some locations). Agreement is particularly evident in the Indian Ocean (including the Mozambique Channel), the Southeast Pacific, the Gulf of Alaska, Baffin Bay, and the equatorial Atlantic, but is poor in the Northwest Pacific. Mapped baroclinic tides usually display positive trends, at least near large generation sites, with some interesting exceptions.