Sea-level rise (SLR), a significant consequence of climate change, poses a severe threat to coastal communities globally. Global mean sea level (MSL) has risen approximately 1.5 mm/yr since 1900, a rate unprecedented in at least the last 3000 years, primarily due to thermal expansion and melting ice. Since the 1960s, this rise has accelerated, exceeding 3 mm/yr in the satellite altimeter era. The U.S. Southeast and Gulf coasts experience SLR rates higher than the global average, influenced by factors such as vertical land motion (VLM), sterodynamic sea level (SDSL) changes, and gravitational, rotational, and deformation (GRD) effects. These rapid increases contribute to nuisance flooding, amplified storm damage, and land loss, impacting national security and hindering adaptation efforts. The National Oceanic and Atmospheric Administration (NOAA) and National Aeronautics and Space Administration (NASA) sea-level rise projections show inconsistencies between process-based models and observed trajectories along the eastern Gulf coast, particularly concerning the acceleration rate. This study aims to determine whether the observed acceleration is a robust indicator of high-end SLR trajectories, due to unresolved processes (e.g., nonlinear VLM), or an amplification of climate-driven acceleration by natural ocean dynamic variability.

Existing research documents a global acceleration in MSL since the 1960s, but local detection is challenging due to natural variability. Studies on the North American East and Gulf coasts report MSL trends ranging from 1.7 to 8.4 mm/yr between 1900 and 2021, generally exceeding the global average. These trends are influenced by a combination of natural and anthropogenic VLM, SDSL changes, and GRD effects. Previous work identified an acceleration hotspot in the Mid-Atlantic Bight and Chesapeake Bay, but the recent acceleration along the Southeastern U.S. coast and its causes remain uncertain. NOAA/NASA projections show a discrepancy between process-based models and observed trajectories, especially along the eastern Gulf coast, where observed rates exceed even the highest projected scenarios, raising questions about the role of high-end emission pathways, ice-sheet sensitivities, and natural variability.

This study analyzes 66 tide gauge records from the Permanent Service for Mean Sea Level (PSMSL) covering 1900–2021 along the North American East and Gulf coasts. Records were gap-filled and corrected for linear VLM using data from the literature. Nonlinear VLM was considered for Louisiana and Texas coastlines using an approach based on differences from the relatively stable Florida Panhandle. Singular Spectrum Analysis (SSA) was used to calculate nonlinear MSL trends with frequencies longer than 30 years. A Monte Carlo experiment, generating 1000 artificial time series with similar noise characteristics, assessed the statistical significance of observed accelerations. Satellite altimetry data, unaffected by VLM, was used to compare acceleration coefficients with tide gauge data and to investigate the spatial structure of MSL variability. To isolate coastal SDSL signals, GRD effects related to barystatic changes and the inverted barometer effect were removed from the tide gauge records. A 1.5-layer reduced gravity model simulated the propagation of Rossby waves in the North Atlantic, forced by wind stress. Principal Component Analysis (PCA) was used to identify the leading modes of variability from this model. Finally, correlations between the Rossby wave signals and other internal forcing factors (river discharge, coastal winds) were evaluated to estimate their combined effect on coastal sea level.

The study found a significant acceleration (P ≥ 0.95) in MSL along the U.S. Southeast and Gulf coasts south of Cape Hatteras since the mid-2000s, with rates exceeding 10 mm/yr by 2021. This acceleration is mainly sterodynamic in origin and extends offshore into the Subtropical Gyre and Caribbean Sea. The acceleration surpasses historical climate model simulations and projections. However, the exceedance is attributed to a combination of externally forced acceleration (approximately 40%, as predicted by climate models) and large internal North Atlantic decadal variability (approximately 60%), which is out of phase with model simulations. A significant portion of the residual SDSL variability is coherent with open-ocean wind stress forcing via westward-propagating Rossby waves in the tropical North Atlantic. These waves influence water mass inflow into the Caribbean Sea, Gulf of Mexico, and the Subtropical Gyre. The combined effect of Rossby waves, river discharge, and coastal winds shows strong correlations with unforced SDSL variability, indicating that internal variability significantly contributes to the observed acceleration. The study highlights the importance of both external and internal factors in shaping regional sea-level trends.

The findings demonstrate that the observed MSL acceleration along the U.S. Southeast and Gulf coasts is not solely due to externally forced climate change but is significantly influenced by internal climate variability. The dominance of internal variability, specifically Rossby waves, explains the discrepancies between observed rates and climate model projections. This implies that MSL rates in the region may revert to rates projected by climate models within the next decade. The study underscores the challenges in early detection of acceleration signals and the need for cautious interpretation of climate model projections at the regional scale. A mechanistic understanding of regional SLR accelerations is critical for accurate sea-level projections.

This study reveals a significant MSL acceleration along the U.S. Southeast and Gulf coasts, primarily driven by a combination of externally forced SLR and amplified internal North Atlantic decadal variability, mainly through wind-forced Rossby waves. The observed acceleration exceeds climate model projections, highlighting the importance of considering internal variability in regional SLR projections. Early detection of such accelerations remains challenging, demanding careful consideration of both external and internal climate processes. Future research should focus on refining the understanding of internal variability mechanisms and their interaction with external forcings for more precise regional SLR projections.

The study uses gap-filled tide gauge data and VLM corrections, introducing uncertainties. While nonlinear VLM is considered for the Gulf of Mexico, potential nonlinearities along other parts of the coast might affect the results. The model used to simulate Rossby waves is simplified; a more comprehensive model might offer a more nuanced understanding. The study relies on existing climate model simulations, whose limitations and uncertainties need to be acknowledged.