Contamination of 8.2 ka cold climate records by the Storegga tsunami in the Nordic Seas

Tsunamis are known to disturb seabeds and rework offshore sediments, as observed after the 2011 Tohoku tsunami in Japan, which produced redeposited mud and sorted sand layers down to ~100 m water depth and turbidity currents transporting sediments into deeper waters. The Storegga Slide in the Norwegian Sea generated a giant tsunami whose onshore deposits have been widely mapped in Scotland, Norway, Shetland and the Faroe Islands. For long-wave tsunamis, currents are largely depth-uniform and are tied to wave amplitude and water depth, with maximum horizontal velocity Umax ≈ ηmax√(g/d). For example, ηmax = 5 m in 250 m water depth implies Umax ≈ 1 m/s, sufficient to erode a wide range of grain sizes (fine silt to fine pebbles). The Storegga tsunami is tightly dated by radiocarbon on green moss fragments in onshore deposits to 8080–8180 cal yr BP (8140 ± 55), falling within the coldest decades of the ~160-year-long 8.2 ka event and likely occurring in late autumn. This study asks whether marine sediment records of the 8.2 ka cooling in the Norwegian and North Seas are contaminated by sediments reworked and redeposited by the Storegga tsunami. The approach combines tsunami flow simulations with a re-investigation of a key sediment core (MD95-2011) from the Vøring Plateau (~1050 m water depth).

Prior work has extensively documented the Storegga tsunami deposits onshore across the North Atlantic region (e.g., Scotland, Norway, Shetland, Faroe Islands). The 2011 Tohoku-oki tsunami provided modern analogs for tsunami-induced sediment remobilization and turbidity currents reaching deep waters. Marine cores in the Nordic Seas have previously been used to reconstruct a distinct and dramatic 8.2 ka cooling. Specific sediment cores (e.g., MD95-2011, LINK14 east of the Faroe Islands, and cores from the Norwegian Channel and off Iceland) have reported anomalous layers around 8.2 ka characterized by changes in grain size, foraminiferal assemblages, magnetic susceptibility, and isotopes; some prior studies suggested potential sorting/reworking. The present work builds on and reassesses these observations in light of tsunami-induced currents and transport.

• Tsunami and landslide modeling: The Storegga Slide was modeled as a cohesive clay-rich debris flow using the two-layer depth-averaged BingCLAW model to simulate slide dynamics and time-evolving bathymetric changes. The resulting tsunami was simulated with the GloBouss model, driven by the time-dependent seabed deformation from BingCLAW. The computational domain used paleo-bathymetry (accounting for changes since ~8150 yr BP) spanning 12.5°W–16.6°E and 53.3°N–70.0°N on ~2 km grids. Simulations ran for 10 hours to capture wave evolution across regions of interest. Landslide parameters (volume 3200 km³; initial yield strength τ1 = 12 kPa; residual yield strength τr = 3 kPa; remolding coefficient I = 0.0005; added mass Cm = 0.1) were selected to best match observed tsunami run-up heights, following sensitivity guidance in earlier studies.
• Currents and boundary layer: Simulations provided maximum flow velocities largely uniform through the water column with a thin bottom boundary layer (~1–10 m). For site-specific erosion estimates, boundary-layer reductions of near-bed velocity were inferred from numerical boundary-layer simulations, yielding factors of ~0.7 (LINK14) to ~0.95 (MD95-2011) relative to depth-uniform flow.
• Critical erosion thresholds: The largest/smallest erodible grain sizes for the simulated currents were estimated using the Sundborg diagram and equations from Miller et al., applied to site-specific near-bed velocities.
• Core re-investigation (MD95-2011, Vøring Plateau ~1050 m water depth): High-resolution analyses identified an anomalous 2-cm layer (533–535 cm) dated to ~8.1 ka by surrounding stratigraphy. Measurements included grain-size distribution (>63 µm fraction; counts of >150 µm grains), abundances of planktonic foraminifera (Neogloboquadrina pachyderma; Neogloboquadrina incompta) and reworked benthic foraminifera (indicative of shallower, high-current environments), and oxygen isotopes (δ18O) of N. pachyderma and N. incompta. Radiocarbon dating of foraminifera within, above, and below the layer used Marine20 calibration with local reservoir correction (ΔR = −145 ± 35). Age–depth modeling incorporated a hiatus inferred from age jumps across the layer.
• Regional core assessments: Additional cores (LINK14 on the Faroe shelf trough; 28-03 in the Norwegian Channel; Iceland shelf/deeper sites) were evaluated using simulated local tsunami currents and previously published stratigraphic/foraminiferal/isotopic records to assess potential reworking and redeposition linked to the Storegga tsunami.

• Simulated tsunami flow velocities: On shallow shelves (Western Norway, Shetland, Faroe Islands) and parts of the North Sea, maximum currents reached 2–5 m/s; velocities >1 m/s occurred at depths shallower than ~250 m and in many areas down to ~1000 m, especially near the Storegga Slide. Initial drawdown toward the slide backwall produced the strongest currents (5–10 m/s) and sea-surface lowering of 15–30 m.
• At MD95-2011 (Vøring Plateau, paleo-depth ~965–1050 m), simulated peak current speed near the seafloor (within ~1–10 m) was ~0.39 m/s, associated with a wave amplitude of ~3.8 m.
• MD95-2011 8.2 ka layer (533–535 cm) exhibits: a sharp, erosive base with upward fining (turbidite character); 12–14× increases in both cold (N. pachyderma) and warm (N. incompta) planktonic foraminifera; abundant benthic foraminifera characteristic of shallower, high-energy environments; δ18O of N. pachyderma rising from ~2.25 to ~2.95‰ (apparent ~3°C cooling if interpreted as in situ signal) while N. incompta shows no such isotopic spike.
• Radiocarbon ages within the 8.2 ka layer indicate redeposition of much older material: N. pachyderma dated 10,690–11,170 and 10,990–11,880 cal yr BP; N. incompta dated 8,670–9,520 and 8,680–9,190 cal yr BP. These are older than ages above/below the layer, indicating a hiatus of ~400 years at the base, corresponding to ~20 cm of erosion at the site.
• The MD95-2011 anomaly is a fine-grained turbidite derived from tsunami-induced backwash erosion and downslope turbidity currents from the shelf break, not a primary climate signal.
• Additional sites: LINK14 experienced simulated water-column velocities up to ~1.8 m/s (~1.2 m/s at 1 m above bed), sufficient to transport sand from shallower banks into the trough; the 114–116 cm sand layer with fauna indicative of sorting is consistent with tsunami reworking, with modeled layer age 7,930–8,470 cal yr BP (2σ). Norwegian Channel core 28-03 saw simulated near-bed velocities ~0.5 m/s aligned along-channel, capable of stirring/reworking fine sediments and foraminifera in the 3–4 cm clayey silt layer dated near 8.2 ka. Around Iceland, strong simulated shelf currents align with observations of reworked Neogene coccoliths and reworking signals off North Iceland.
• Regional impact: Simulations indicate maximum velocities ≥0.25 m/s (able to mobilize grains up to ~1 mm) down to ~1000 m water depth over broad areas (58°–74° N), implying widespread seabed reworking and potential turbidity current initiation.
• Implication for paleoclimate: Previously inferred large, abrupt marine 8.2 ka cooling signals in affected Nordic Seas cores likely reflect tsunami-induced reworking and should be reconsidered or discarded as climate indicators.

The study directly addresses whether marine sediment records of the 8.2 ka cooling in the Norwegian and North Seas were contaminated by the Storegga tsunami. Numerical simulations demonstrate that tsunami currents were sufficiently strong and widespread to erode, transport, and redeposit sediments across shelves and into deep waters, consistent with modern analogs from the 2011 Tohoku event. Re-investigation of MD95-2011 shows the hallmark features of a tsunami-triggered turbidite—an erosive base, normal grading, spikes in reworked planktonic and benthic foraminifera, and radiocarbon ages significantly older than the stratigraphic context. The δ18O anomaly of N. pachyderma is thus explained by incorporation of older, colder-water foraminifera rather than an in situ surface-ocean cooling of ~3°C at 8.2 ka. Comparable mechanisms are supported at additional sites (LINK14, Norwegian Channel, off Iceland), where modeled currents exceed thresholds for mobilizing the observed grain sizes and faunal assemblages consistent with sorting and reworking. Collectively, these findings imply that a substantial portion of the sea floor in the Nordic Seas was disturbed during the Storegga tsunami and that prominent 8.2 ka layers in several marine cores represent tsunami deposits or reworked horizons rather than primary climate signals, necessitating a reassessment of regional marine paleoclimate reconstructions for this interval.

Simulations of the Storegga tsunami, combined with stratigraphic, micropaleontological, isotopic, and radiocarbon evidence, indicate that the widely reported 8.2 ka anomaly in several Nordic Seas marine sediment cores is largely the product of tsunami-induced reworking and turbidite deposition. Flow velocities of 2–5 m/s on shelves and ≥1 m/s down to ~1000 m depth were capable of mobilizing and redepositing sediments and foraminifera, as demonstrated for core MD95-2011, where the 8.2 ka layer comprises redeposited Early Holocene material with an erosive hiatus. Consequently, prior reconstructions of a large, abrupt 8.2 ka cooling from such marine records are likely erroneous and should be reconsidered as tsunami deposits. Future work should systematically re-evaluate marine cores from the Norwegian, North, and adjacent seas for evidence of tsunami reworking, integrate tsunami sedimentology into paleoclimate interpretations, and refine regional models of tsunami flow–sediment interaction using expanded core networks and improved paleo-bathymetric constraints.

The tsunami modeling depends on assumptions about landslide rheology and volume, which are uncertain; although the authors argue that adjusted material properties can reproduce observed run-up heights even with smaller slide volumes, parameter choices influence simulated amplitudes and currents. Boundary-layer near-bed velocities were inferred from prior numerical studies and simplified factors, and erosion thresholds were estimated using generalized diagrams/equations, introducing uncertainty at specific sites. The direct re-investigation with new radiocarbon dating focuses on one key core (MD95-2011), with additional sites assessed using existing data; broader regional confirmation would benefit from more extensive re-dating and sedimentological analyses across multiple cores.