Rising seas could cross thresholds for initiating coastal wetland drowning within decades across much of the United States

Sea-level rise is expected to be among the most costly and far-reaching consequences of climate change, with coastal cities and ecosystems already affected by increased flooding and saltwater intrusion. Coastal wetlands (marshes, mangrove forests, freshwater swamps) provide critical ecosystem services, including storm protection, fisheries support, and carbon sequestration, yet their future under rapidly rising seas is contested. Some studies project imminent catastrophic wetland loss, while others emphasize wetland resilience and adaptive capacity. This uncertainty complicates management choices between resisting change in place versus facilitating wetland migration and retreat. To inform conservation and restoration decisions, the study aims to clarify the potential timing and locations where coastal wetland drowning could begin under accelerated sea-level rise across the conterminous United States. Building on growing consensus from paleoecological records, instrumental data, and models about submergence limits (drowning thresholds), the authors evaluate where and when relative sea-level rise rates are projected to exceed these thresholds, signaling initiation of wetland drowning.

The paper synthesizes advances indicating that coastal wetland vertical adjustment can keep pace with only limited sea-level rise before reaching submergence limits. Evidence from paleo-stratigraphic records, contemporary measurements, and numerical models shows non-linear biogeomorphic feedbacks that can build elevation (via root growth and sedimentation) but fail beyond critical rates. A global synthesis suggests deficits are likely at 4 mm/yr and highly likely at 7 mm/yr, with likely drowning and conversion to open water when such rates are sustained. Region-specific analyses identify approximate thresholds of 5 mm/yr for eastern U.S. marshes, 6–9 mm/yr for Mississippi River delta marshes, and 7 mm/yr for San Francisco Bay marshes. The literature also notes that once thresholds are crossed, complete submergence is not immediate; persistence can vary depending on tidal range, elevation capital, sediment inputs, landscape position, and local relative sea-level rise rates.

Study area and grid: The analysis spans the coastal conterminous United States (22 coastal states plus Washington, D.C.). A grid of 168 one-degree cells was created to match the spatial resolution and registration of relative sea-level rise (RSLR) data and to include only cells containing estuarine vegetated wetlands, identified using NOAA’s 2016 Coastal Change Analysis Program (C-CAP) 30-m land cover data (estuarine emergent, scrub-shrub, forested classes). Note: tidal freshwater wetlands are not included due to lack of a national classification; results therefore reflect tidal saline vegetated wetlands only. Sea-level rise scenarios and projections: Three RSLR scenarios from the U.S. interagency sea-level rise technical report were used: Intermediate-Low (0.5 m global mean SLR by 2100 relative to 2000), Intermediate (1.0 m), and Intermediate-High (1.5 m). Decadal regional RSLR projections incorporate gravitational/rotational changes, oceanographic shifts, and vertical land motion (subsidence/uplift from sediment compaction, glacial isostatic adjustment, groundwater and fossil fuel withdrawals, etc.). Projections are available from 2005 and at decadal intervals from 2020 to 2150; median projection values were used. Drowning thresholds and timing algorithm: Three drowning thresholds were evaluated to bracket literature-based rates at which wetlands are unlikely to keep pace: 4, 7, and 10 mm/yr. Decadal RSLR rates were computed by dividing decadal sea-level increase by years in the interval (10 years for most, 15 years for 2005–2020). To identify the decade of drowning initiation within each cell, the first decade in which the decadal RSLR rate met or exceeded the threshold and the following decade also met/exceeded the threshold was selected. The onset was assigned to the last year of the first qualifying decade (e.g., 2030 for 2020–2030 if 2030–2040 also ≥ threshold). This approach filters transient exceedances. Supplementary analyses include maps for the 17th and 83rd percentile projections. Extent of potential drowning: For each cell and scenario-threshold combination, the area of wetlands potentially initiating drowning was estimated by overlaying the identified timing with current estuarine vegetated wetland coverage (NOAA 2016 C-CAP). State-level figures show timing, area, and percent of wetlands expected to cross thresholds for 22 coastal states and D.C. Conterminous U.S. figures show national timing and percent. The earliest possible date in analyses is 2020; some areas (e.g., Louisiana’s Mississippi River delta) have already begun drowning. Contextual scenario alignment: The regional trajectories through mid-century broadly align with current observed trends: southeastern Atlantic coast by 2050 falls between Intermediate and Intermediate-High (0.36–0.43 m relative to 2000); western Gulf of Mexico between Intermediate and Intermediate-High (0.57–0.63 m); northwestern Pacific coast between Intermediate-Low and Intermediate (0.15–0.18 m).

- Strong spatial variation in RSLR exposure and wetland abundance drives variability in timing and extent of threshold crossings across the U.S. - Hotspots for early initiation: northwestern Gulf of Mexico, south-Atlantic, and mid-Atlantic coasts, especially subsidence hotspots (e.g., Texas coast, Chesapeake Bay, and particularly Louisiana’s Mississippi River delta) where RSLR rates already exceed critical thresholds. - Louisiana case: From 1932–2016, Louisiana lost 4,833 km² of coastal wetlands due in part to subsidence and high RSLR. Louisiana holds nearly 30% of conterminous U.S. coastal wetlands and is among the first expected to drown. At sustained rates exceeding 6–9 mm/yr, considerable marsh-to-open-water conversion is expected within 50 years in the Mississippi River delta. In a recent period with RSLR >10 mm/yr, 87% of monitored Louisiana wetlands could not keep pace over 13 years. - Florida: Despite historical stability (minimal subsidence and small RSLR), the potential extent of future drowning is large due to low topography and the state containing roughly 25% of U.S. coastal wetlands, including the Greater Everglades mangrove–marsh mosaic. - Pacific coast: Onset is generally later, particularly Washington and Oregon, due to post-glacial rebound lowering relative rates below global means. - Sensitivity to scenarios and thresholds: Under Intermediate-High SLR, drowning begins between 2020–2060 under all thresholds (4, 7, 10 mm/yr). Under the highest SLR scenario and lowest threshold, initiation occurs on all three coasts by 2040. Under Intermediate-Low SLR and the highest threshold, initiation is delayed and varies widely. Results emphasize the importance of both accurate RSLR projections and refined drowning thresholds. - State-specific timing and area vary; large wetland landscapes at risk include the Mississippi River delta, Greater Everglades, Chesapeake Bay, Texas, Georgia, and the Carolinas.

The analysis indicates that areas with high wetland coverage and high relative sea-level rise—often exacerbated by subsidence—will experience initiation of drowning earliest. In Louisiana, Texas, and Chesapeake Bay, evidence suggests that drowning is already underway. Conversely, regions with lower relative rates (e.g., Pacific Northwest) are projected to experience a later onset. The timing and extent of initiation are highly sensitive to both the chosen drowning threshold and the RSLR scenario; thus, improving process understanding that sets thresholds (soil properties, sediment inputs, tidal range, plant communities) and utilizing the best-available regional RSLR projections are crucial. By identifying when and where thresholds are crossed, the study offers a rapid, scalable screening tool to prioritize scientific investigations and management actions, guiding where local, customized models are most needed to evaluate duration and progression of drowning and to plan adaptation strategies (e.g., sediment augmentation, accommodation space for landward migration). The approach may require customization in other global regions where thresholds or controlling factors differ.

This work integrates recent advances on wetland drowning thresholds with regional relative sea-level rise projections to estimate the timing and locations of threshold crossings that initiate coastal wetland drowning across the conterminous United States. Results reveal substantial spatial heterogeneity, with early initiation concentrated in subsiding, wetland-rich regions along the Gulf and Atlantic coasts, and later onset along much of the Pacific coast. The findings underscore two imperatives: (1) minimize acceleration of sea-level rise to avert catastrophic wetland loss and associated transformative coastal change; and (2) anticipate and prepare for social and ecological implications, as initiation under many scenarios is projected within decades. The threshold-crossing framework provides a rapid, large-scale prioritization tool for directing detailed local studies and targeted management, and it highlights research needs to refine thresholds and local RSLR characterization. Future work should incorporate tidal freshwater wetlands, improve representation of vertical land motion and sediment dynamics, and assess duration and progression of drowning beyond initiation.

- Wetland coverage: Analyses include only estuarine vegetated (tidal saline) wetlands from NOAA 2016 C-CAP; tidal freshwater wetlands were excluded due to lack of a national classification, likely underestimating total vulnerable area (notably in wetland-rich states). - Projections: Median RSLR projections were used; although percentile maps are provided in supplements, uncertainties remain, especially regarding low-confidence processes (e.g., ice sheet dynamics) not included in IPCC medium-confidence projections. - Threshold uncertainty: Drowning thresholds (4, 7, 10 mm/yr) are bracketing estimates from literature; true thresholds vary by local conditions (soil, sediment supply, tidal range, plant communities). - Temporal methodology: Decadal rates and the requirement for consecutive decades reduce spurious detections but may miss short-term dynamics; initiation timing represents threshold crossing, not the full duration to complete submergence or loss. - Spatial resolution: One-degree grid may smooth fine-scale variability; local vertical land motion and sediment dynamics can vary within cells. - Generalizability: Approach tailored to conterminous U.S.; thresholds and controls may differ elsewhere, requiring regional customization.