Meeting Documents

Sea Level Variability Prediction and Attribution Using the ECCO Ocean State Estimation Framework

Abstract

Sea level (SL) variability on synoptic-to-decadal timescales is driven mostly by changing inputs at the ocean surface, but ocean dynamics may convey the impact of these forcings across large distances in space and time. Coupled seasonal prediction and climate models frequently struggle to represent these remotely-forced and lagging impacts, due to model drifts that may result from the initial conditions or model physics. The Estimating the Circulation and Climate of the Ocean (ECCO) state estimate addresses the challenges of model drift in the ocean by using observations to adjust model parameterizations as well as initial and boundary conditions for a more realistic ocean state, while maintaining physical conservation laws in the ocean interior. While the ECCO state estimate itself is not a prediction model for the future ocean state, this effort exploits the improved ocean model, adjoint, and attribution capabilities of ECCO to better understand and predict synoptic-to-decadal SL variability.

Two studies are discussed: the first study implements ECCO adjoint sensitivities in combination with ERA5 reanalysis and coupled seasonal prediction models to evaluate a framework for prediction of sea level changes along the U.S. Gulf Coast, on timescales of weeks to months. The sea level variations in ECCO adjoint-based hindcasts are attributed to fluxes in the Gulf as well as the Atlantic coast of North America and the tropical Atlantic. Areas of high-impact forcing are identified where improved seasonal forecasts will have the most benefit for sea level predictions. The second study examines forward sensitivity runs of the ECCO modeling framework over 1992-2019, assessing the impact of surface fluxes (freshwater, heat, wind stress) on decadal SL reversals. Decadal reversals in regional SL trends are largely attributed to wind stress and heat fluxes, with the heat fluxes sometimes compensating the regional wind stress effect. These results have implications for SL predictions on a wide range of timescales, and may help prioritize infrastructure and ecosystem adaptations in a world of rising global mean sea levels.