Pacific Ocean in Northern California. Photo: Dan Meyers, Unsplash.

A new global sea-level curve

A new global Phanerozoic sea-level curve has recently been published, which deviates from the famous Haq curves in places.

Traditional long-term eustatic sea-level reconstructions suffer from uncertainty in stratigraphic interpretations of the rock record, which leads to ambiguity. Sea-level reconstructions from plate tectonic modelling are limited in use due to a lack of preserved oceanic crust further back in time. As a result, Phanerozoic eustasy remains poorly constrained.

From a geological perspective, sea-level is very low at the present day due to the presence of polar icecaps and reduced oceanic crustal spreading.

It is well known that the primary drivers for eustasy are changes in the rate of mid-ocean-ridge spreading and ice cap volume. The alternative method of sea-level reconstruction presented here integrates these two elusive drivers by using a corrected strontium global record to quantify plate tectonics and global average paleotemperatures to estimate continental glaciation.

Global tectonic eustasy

The well-constrained 87Sr/86Sr strontium isotope record derived from marine carbonates provides insight into plate tectonic activity. At first order, the strontium isotopes in carbonates record contributions of volcanism and weathering of radiogenic crust. We developed a novel approach to remove the weathering component using run-off estimates from paleo-climate modelling. This resulted in an overview of only the plate tectonic component of the strontium record, which was subsequently used as a proxy for mid-ocean-ridge spreading rates, and hence average ocean crustal age.

In turn, the oceanic crustal age was used to calculate water depths. A young and shallow ocean floor is found at mid-ocean ridges, whereas old crust leads to a deep ocean floor. From the tectonic component of the strontium isotope record, it follows that ridge spreading rates were twice as high in the mid-Mesozoic compared to the present day. This result is consistent with independent studies from plate tectonic models, and subducted slabs imaged in the mantle. The tectonic eustatic component leads to a sea-level amplitude range of up to 190 m.

Figure 1. Isostatically compensated effects of plate tectonics, glaciations (>1Myr) and the combined Tectono-Glacio-Eustatic curve including RMS uncertainty ranges (shaded blue). Modified from van der Meer et al. (2022).


Water volume variations as a function of perennial ice sheets have a significant contribution to sea-level amplitude changes. To improve the tectonic eustatic reconstruction, we incorporated a glacio-eustatic component by estimating land and grounded ice sheets on the continental shelf through time. Global average paleotemperatures published by Scotese were used to estimate continental ice areas and volumes throughout the Phanerozoic.

Using the T = -10°C isotherm, paleo-latitudes for potential ice sheets were calculated and combined with digital elevation models in a paleogeographic reconstruction. This provided estimates of perennial ice areas through time in 1-million-year increments.

Average ice thicknesses were then calibrated with the well-studied Late Cenozoic icehouse conditions. Using the best-fit average ice sheet thickness for the late Cenozoic (1.4 km), ice volumes were obtained for short-lived Mesozoic and long-lasting Paleozoic icehouse times. This was finally converted to ocean water volume variation, leading to long-term glacio-eustatic amplitude of up to 90 m.

Tectono-Glacio-Eustatic curve

Both components – long-term glacio-eustatic and plate tectonic eustatic reconstructions – were combined and isostatically corrected, resulting in the Tectono-Glacio-Eustatic (TGE) curve shown in Figure 1. The overall amplitude has a range of around 250 m throughout Phanerozoic. From a geological perspective, sea-level is very low at the present day due to the presence of polar icecaps and reduced oceanic crustal spreading. A similar situation occurred during the Late Paleozoic icehouse approximately 300 million years ago.

Global Mean Sea-level was very high (+200 m) during the Late Jurassic-Eocene greenhouse period. This correlates well with published gross-depositional environment (GDE) maps when large parts of continental shelves and present-day land were flooded. Based on the TGE curve and present-day hypsometry, there was approximately 30% less land area during that time.

Figure 2. TGE curve zoomed in to Jurassic-Cenozoic with key sea level cycles at >10Myr time-scales.

Mega-Sequence Stratigraphy and comparison to Haq curves

The TGE curve provides insight into the eustatic driver of passive margin stratigraphy, which acts in addition to other factors such as local tectonics and changes in sediment supply. At >10 Myr time scales, five rising-falling sea-level cycles are interpreted since the Jurassic (Figure 2). In comparison with other published sea-level curves that were based on plate tectonic models, stratigraphic methods or continental flooding mapping, there are differences in trends and amplitudes.

Haq does not specify in detail the methods used behind reconstructing his sea-level curve, which makes it hard to QC. Perfect data for one basin does not mean that it is representative for global sea-level changes, as local tectonics and delta lobe shifting can heavily influence changes in stratigraphy.

The TGE curve and the compilation of three different published curves by Haq are illustrated together over a focused Jurassic to Cenozoic time period (Figure 3) for the purpose of investigating second-order features. Qualitatively, there is a general agreement between the two curves that there are several Maximum Flooding Times (MFT). There is consensus on the Cenomanian-Turonian peak, the Paleo-Eocene peak and a subsequent falling sea-level during the Cenozoic and finally the Late Jurassic peak.

Figure 3. Comparison with the long-term curves of Haq et al. (2005, 2014, 2018, spliced together) for the Jurassic-Cenozoic.

Noteworthy opposing trends are seen in the Berriasian, Aptian, and Campanian. Also, where trends agree, significant differences are noted in second-order amplitudes (i.e., Berriasian-Valanginian). To test the mega-sequences inferred from the TGE curve, examples from the prolific Guyana-Suriname Basin passive margin will be illustrated in a follow-up article that will soon be published by GEO ExPro.

Douwe van der Meer and Kent Wilkinson, CNOOC


Scotese, C. R., Song, H., Mills, B. J. W., and van der Meer, D. G. [2021]. Phanerozoic paleotemperatures: The earth’s changing climate during the last 540 million years. Earth-Science Reviews, 103503.

van der Meer, D. G., Scotese, C. R., Mills, B. J. W., Van den Berg van Saparoea, A.-P. H., Sluijs, A., and van de Weg, Ruben. M. B. [2022]. Long-Term Phanerozoic Global Mean Sea Level: Insights from Strontium Isotope Variations and Estimates of Continental Glaciation. Gondwana Research, 111, 103–121.

Previous article
“Shooting a fish in a barrel”

Related Articles