The validity of sea-level estimates based on sedimentological correlations have been debated for many decades. This is because relative sea-level curves differ between sites due to local tectonics, sediment deposition rates, and changes in elevation due to the Earth’s mantle flow. In deep time, those estimates essentially rely on methods associated with sediment based observations and plate tectonic reconstructions, which assess changes to the depth and the extent of ocean basins. However, plate tectonic methods become less and less reliable with time, and are mainly based on synthetic datasets for times before 150 Million years ago.
Compared to regional stratigraphic observations, global flooding estimates (Fig. 1) represent averages of sea-level changes across large geographic areas, making them less sensitive to modifications by regional alterations. In this study, global sea-level change is calculated directly from flooding maps using what is called the hypsometric method. This method converts continental flooding into changes in sea level using a landmass-specific hypsometric curve, which depicts the proportion of land area at various elevations. This method provides estimates of sea-level fluctuations with coarser time-resolution compared to other methods (e.g., seismic or stratigraphic analyses). However, the hypsometric method holds the significant advantage of being applicable over long periods.
Global sea levels estimated in a new study led by CEED PhD student Chloé M. Marcilly demonstrate high sea levels around 500 Million years (more than 250 meters above present-day sea levels) (Fig. 2, thick blue curve). Minimum sea levels comparable with today are seen at around 200 million years and thereafter an accelerated sea-level rise, culminating at 280 meters at about 80 Ma. Since then sea levels have dropped systematically to present-day values.
It is widely recognized that tectonic processes drive long-term changes in sea levels, principally linked to changes in the volume of the ocean basins, which can be computed from global full-plate models. However, sea level changes observed in this study correlates well with sediment-based reconstructions.
It has previously been assumed that the hypsometric method is limited to first-order approximations. However, the new reconstructions by Marcilly and colleagues demonstrate both first and second-order cycles. Indeed, second-order cycles of ~ 37 Myrs were observed with a recurring strong signal for most of the past 520 million years. Similar cycles have previously been identified and attributed to astronomical processes but a spectral analysis carried out on seafloor spreading proxy datasets suggests that tectonic processes might be the source for these second-order cycles.
It has previously been assumed that the hypsometric method is limited to first-order approximations, sea level cycles happening across timescales of 50-100 million years or more. However, the new reconstructions by Marcilly and colleagues demonstrate both first and second-order cycles, sea level cycles of 5 to 50 Million years. Indeed, second-order cycles of ~ 37 Myrs were observed with a recurring strong signal for most of the past 520 million years. Similar cycles have previously been identified and attributed to astronomical processes but this new study also shows that tectonic processes also exhibit cycles of about 37 Myrs, and might be the source of these second order sea level cycles. The close temporal link between the presented sea levels and tectonic processes further strengthens the case for the reliability of the hypsometric method for reconstructing past sea levels. Moreover, this method may also allow a to identify regional variations in sediment-based observations, linked to uplift and subsidence, through deviation from the resulting global sea level curve.
Contact information: – Chloé M. Marcilly, c.f.m.marcilly@geo.uio.no
Publication details: Marcilly, C. M., Torsvik, T. H., & Conrad, C. P. (2022). Global Phanerozoic Sea levels from paleogeographic flooding maps. Gondwana Research. https://doi.org/10.1016/j.gr.2022.05.011.
Funding for this research came from the Research Council of Norway (RCN), through its Centres of Excellence funding scheme, project 223272 (CEED), and 288449 to CPC (MAGPIE project).
