Low-frequency non-astronomical changes of ocean tides of 
(1 cm) have been documented in water level measurements around the globe, but their causative mechanisms remain poorly understood in many cases. While anthropogenic developments (e.g., harbor dredging) are certainly a leading factor at individual sites, the spatially coherent tidal variability seen in areas with distributed tide gauge information is revealing of natural processes. Here we use a general circulation model, configured on a 1/12° horizontal grid, to spatially map the influence of ocean stratification changes on the global M2 tide from 1993 to 2019. We partition the problem into separate yearly simulations of short duration (40 days) and relax each forward integration to the year’s “true” stratification, as provided by an eddying ocean reanalysis. The simulations reveal typical stratification-driven M2 amplitude changes of 0.5 cm on interannual time scales, as calculated at positions of 40 coastal tide gauges in three particular regions (New Zealand & Australia, Florida & Gulf of Mexico, Northeast Pacific). Most of the identified fluctuations at the coast are present in the barotropic tidal component, suggesting an origin in changing tidal conversion at remote topography or turbulent energy dissipation in shallow water. In addition, we fit linear rates to the yearly M2 solutions over the 1993–2019 time span and compare the resulting in-phase and quadrature trends to a novel (but still tentative) estimate of M2 trends in the open ocean from TOPEX-Jason satellite altimetry. The two solutions bear gross resemblance to each other and indicate large spatial-scale trends of ~1 cm cy-1 in the barotropic M2 tide in the Indian Ocean, the western and northern Pacific (e.g., in the Gulf of Alaska), and Baffin Bay. Our results highlight that efforts seeking to explain interannual to secular changes of tides at the coast and in the open ocean must consider both sea level rise and contemporary changes in ocean stratification.
