Higher than present global mean sea level recorded by an Early Pliocene intertidal unit in Patagonia (Argentina)

The study addresses how high global mean sea level (GMSL) was during the Early Pliocene, a warm period considered an analog for near-future climate. Because the only direct proxies for past sea level are elevations of relative sea level (RSL) indicators (e.g., fossil reefs, beach deposits), robust, well-dated, and well-preserved older indicators are crucial but rare. Previous precise Early Pliocene RSL indicators exist from Mallorca (Spain) and South Africa, suggesting high sea levels. The Early Pliocene had CO2 between pre-industrial and ~450 ppm, with global temperatures 2–3 °C above pre-industrial, and sea-level oscillations paced by ~40 kyr cycles. Ice-sheet modeling indicates reduced Greenland and West Antarctic Ice Sheets and possible East Antarctic contributions. This study surveys, dates, and interprets an intertidal unit at Camarones (Patagonia, Argentina) to infer site paleo-RSL and, after correcting for departures from eustasy (GIA, tectonics, dynamic topography), to estimate contemporaneous GMSL.

- Direct Early Pliocene sea-level indicators: Mallorca (Coves d’Artá) phreatic overgrowths at 31.8 ± 0.25 m above MSL with U-Pb age 4.29 ± 0.39 Ma; after GIA/tectonic corrections, GMSL ~25.1 m (16th–84th percentiles 10.6–28.3 m). South Africa (Cliff Point) oyster shells constraining paleo RSL 35.1 ± 2.2 m with SIS ages 4.28–4.87 Ma, with arguments for limited vertical land motion since 5 Ma. - Indirect proxies: Stacked benthic δ18O curves over last 5.3 Ma provide inferred ice volume/sea-level histories but include uncertainties in temperature-ice volume partitioning. One-dimensional ice-sheet models yield differing sea-level amplitudes; some underpredict observed Early Pliocene highstands, while hydraulic modeling at Gibraltar potentially overpredicts. - Geological/tectonic context of Patagonia: Passive Atlantic margin with long-term moderate uplift since Miocene; multiple mapped Pleistocene and older marine terraces; prior Early Pliocene marine deposits identified regionally. - Dynamic topography (DT) influences long-wavelength vertical land motions; recent models (Flament et al., Müller et al.) suggest uplift in Patagonia since MIS 5e, with significant model spread. These studies frame the need for new, well-dated field constraints and careful corrections for GIA and DT to estimate GMSL.

Field sites and stratigraphy: Two nearby outcrops on the Camarones High Terrace (~40 m a.s.l.) were examined: (1) Roadcut (previously described), and (2) Caprock (newly identified). Both display a transgressive sequence over a raised shore platform. A key unit (Unit Cp) comprises well-cemented fine conglomerates with rounded pebbles and a dense shellbed of 15 bivalve and 11 gastropod species, interpreted as foreshore (intertidal) deposition. Elevation measurements: Differential GPS (Trimble ProXRT, Trimble Tornado antenna) with OmniSTAR HP real-time corrections measured orthometric heights relative to GEOIDEAR16 geoid (accuracy ~0.1–0.6 m; geoid accuracy ~0.1 m). Positions were recorded in WGS84/ITRF2008 ellipsoid and converted to heights above geoid. Unit Cp was measured at two points at each site. The average elevation is 36.2 ± 0.9 m (1σ) above the geoid (N=439 filtered positions). Paleo-RSL was derived by interpreting Unit Cp as intertidal with an indicative range (IR) equal to the modern tidal range at Camarones (MLLW to MHHW = 5 m). Using RSL = mean elevation − reference water level (mid-tide) and σRSL = sqrt(σE^2 + (σIR/2)^2), paleo RSL = 36.2 ± 2.7 m (1σ), assuming no post-depositional movement and modern-like tides. Chronology (Strontium Isotope Stratigraphy, SIS): Ostreidae shells from Unit Cp were dated using SIS (LOWESS v5 curve). Sequential leaching targeted minimally altered inner carbonate; preservation was assessed by leach variability, screening techniques, and elemental analyses, with a preservation index (1=unaltered to 3=altered); samples >2.0 excluded. TIMS measurements were made at SUNY Stony Brook and LDEO. Of three oysters (one Caprock, two Roadcut), one Roadcut sample was excluded due to alteration. Accepted inner-leach ages (n=6) yield a mean SIS age of 4.98 Ma with 2σ SEM range 4.69–5.23 Ma. Glacial Isostatic Adjustment (GIA): A gravitationally self-consistent sea-level model (including shoreline migration and rotational feedback) computed RSL effects from (i) incomplete present-day adjustment to late Pleistocene ice loading (ICE-5G) and (ii) Early Pliocene Antarctic 40 kyr ice-volume oscillations inferred from scaled benthic δ18O. A suite of 36 Earth viscosity/lithosphere models was used to derive mean and standard deviation of the local GIA correction. For Camarones, μGIA = −14.6 ± 3.2 m (1σ), comprising −9.5 ± 3 m from ongoing postglacial adjustment plus ~−5 m from Pliocene Antarctic variability. Vertical Land Motions (VLM) from Dynamic Topography: Long-term VLMs were estimated from published DT models (Flament et al. 2015; Müller et al. 2018). Predicted uplift rates near Camarones average 4.5 ± 2.2 m/Ma (cases summarized in Table 2). Multiplying by the deposit age (with 1σ age uncertainty) gives a downward correction to GMSL of 22.4 ± 11.0 m (1σ). GMSL computation and uncertainty: GMSL = RSL − μGIA − μVLM; σGMSL = sqrt(σRSL^2 + σGIA^2 + σVLM^2). Using RSL 36.2 ± 2.7 m, μGIA −14.6 ± 3.2 m, and VLM 22.4 ± 11.0 m yields GMSL = 28.4 ± 11.7 m (1σ) at 4.69–5.23 Ma. Methods also include data processing scripts and GIA code availability.

- The Camarones intertidal (foreshore) Unit Cp is Early Pliocene in age, dated by SIS to 4.69–5.23 Ma (2σ SEM range; mean ~4.98 Ma), based on screened Ostreidae shells. - Elevation of Unit Cp measured by differential GPS at two nearby outcrops (Roadcut and Caprock) averages 36.2 ± 0.9 m above the GEOIDEAR16 geoid; interpreted paleo-RSL is 36.2 ± 2.7 m (1σ) above present. - Site-specific GIA correction (mean of 36 Earth models) is −14.6 ± 3.2 m (1σ), reflecting ongoing postglacial rebound (~−9.5 ± 3 m) and Pliocene Antarctic ice oscillations (~−5 m). The GIA-corrected paleo-RSL is 50.8 ± 4.2 m (1σ). - Dynamic topography-derived vertical land motion implies a downward correction to GMSL of 22.4 ± 11.0 m (1σ), based on an average uplift rate of 4.5 ± 2.2 m/Ma at the site. - Resulting Early Pliocene GMSL at time of deposition is 28.4 ± 11.7 m (1σ) above present. - Cross-site comparison: This estimate aligns with Mallorca (Spain) GMSL ~25.1 m (16th–84th percentiles 10.6–28.3 m) and, after corrections, is consistent with South Africa within large uncertainties. Collectively, Early Pliocene GMSL likely exceeded 20 m above present. - Implications: Achieving ~28 m GMSL requires an ice-free Greenland (~7.4 m), major WAIS loss (~3.3 m), and substantial melt from marine-based sectors of EAIS (~≥8–19 m).

The findings provide a direct, well-dated, intertidal RSL indicator from South America that, after rigorous corrections for GIA and long-term vertical land motion, indicates Early Pliocene GMSL of ~28 m above present. This value addresses the central question of how high sea level rose under Pliocene warmth and supports the interpretation that polar ice sheets were substantially reduced relative to today. Comparison with independent sites in Spain and South Africa shows coherence among geographically disparate records. The field-based GMSL estimates are most consistent with higher-amplitude sea-level reconstructions obtained by scaling the Lisiecki and Raymo benthic δ18O stack, and with certain ice-sheet model peaks, while some other ice-sheet models underpredict or the Gibraltar planktic δ18O-based curve overpredicts relative to observations. The primary uncertainties in the GMSL estimate stem from VLM corrections associated with dynamic topography and potential additional tectonic/flexural effects that are difficult to quantify fully. Improving constraints on mantle structure and rheology, as well as better local uplift histories, would reduce these uncertainties. The magnitude of inferred GMSL implies an ice-free Greenland, major WAIS retreat, and significant EAIS marine-sector loss during the Early Pliocene, informing expectations for ice-sheet sensitivity under sustained warming.

This study documents an Early Pliocene intertidal unit at Camarones (Patagonia, Argentina), precisely surveyed and dated, yielding a paleo-RSL of 36.2 ± 2.7 m and a corrected GMSL of 28.4 ± 11.7 m at 4.69–5.23 Ma. In concert with records from Spain and South Africa, the results indicate Early Pliocene sea level likely exceeded 20 m above present, requiring an ice-free Greenland and substantial Antarctic contributions. These field constraints provide key targets for ice-sheet and sea-level models and underscore the relevance of the Early Pliocene as an analog for future warming scenarios. Future work should refine vertical land motion estimates through improved mantle tomography, better dynamic topography calibration, and local uplift histories derived from multi-period shoreline sequences.

- The dominant source of uncertainty is vertical land motion from dynamic topography; model predictions vary widely and uncertainties are not fully quantifiable. - Potential additional VLM from flexural response to sediment loading/unloading or unmodeled tectonics was not explicitly included and could bias estimates. - Paleo tidal range was assumed equal to modern (5 m); differences in Pliocene shelf bathymetry could change the indicative range and RSL uncertainty. - GPS measurements at the Roadcut site may be affected by reduced satellite visibility; however, multiple measurements and error propagation were applied. - SIS ages rely on sample preservation screening; diagenetic alteration remains a possibility despite rigorous leaching and exclusion criteria. - GIA modeling relies on assumptions about Pliocene ice volume partitioning and δ18O scaling, introducing additional uncertainty.