Main Image

Strombus Bubonius (Photo credit: Irene Thom)

Many of us are obsessed with the Last Interglacial (LIG or Stage 5e MIS). Between c. 130000 and 115000, global mean temperatures appear to have been slightly warmer by 1 to 2 degrees than present.  Global mean sea level appears to have been higher and ice volumes less than today. If we can improve our understanding of these factors, it not just improves our knowledge of the past, but may guide us to how future ice sheets and sea levels are responding to global warming induced by greenhouse gas emissions, not just solar insolation.

My own obsession commenced in the early 1960s as I hypothesised the age of NSW “Inner Barrier” deposits. Degraded morphology, a single date beyond C14 limit coupled with deep (below sea level) ground water podzolization led me to think Interglacial and not Holocene. I equated it in time with Northern Hemisphere interglacial terms: Sangamon in US and Eemian in Europe. During PhD research in South Carolina this thinking continued where the team I was working with recognised the problem of dating the youngest of the Pleistocene terraces in Horry County (Myrtle Beach Barrier—Socastee Formation; DuBar et al., 1974). We noted “whether it was deposited during the Sangamon or a Wisconsin interstadial is uncertain”. Uncertainty still prevails to some extent as discussed in that excellent synthesis of LIG sea levels by the two Colins, Murray-Wallace and Woodroffe, in their book on Quaternary Sea-Level Changes (2014, p.290). Inspired by the early U/Th dating work of Herb Veeh (1966), I was involved LIG dating studies in the 1970s and 80s, in collaboration with John Chappell, John Marshall and Colin Murray-Wallace. We used different dating techniques to document the morphostratigraphy and chronology of the LIG at some east coast sites. John Chappell and I published a short note in Nature that brought together some of our data and speculated on the “Termination of last interglacial episode and the Wilson Antarctic surge hypothesis” (Chappell and Thom, Nature, 1978, 272, 809-810). I will return to this paper later.

In February 1976, I had the good fortune of spending a month visiting Quaternary coastal sites from Spain to Israel. This trip enabled me to meet with several geologists and palaeontologists who generously shared information from their regional studies that documented stratigraphy, elevation, fauna, and age of what they considered at that time to be LIG (Eemian) age. One distinguishing feature was the occurrence of so-named “Senegalensis” warm-water fauna containing the distinctive mollusc then called “Strombus bubonius” (it is now called Persististrombus latus—a real mouthful!). I have a sample from Las Rocas on Mallorca where I was guided by an ebullient palaeontologist, Dr Cuerda, who had a personal affinity with this species saying “goodbye Strombus” as we left one site, and “I meet the Strombus here” at another. Cuerda could recognise two units of LIG age which he linked to other Mediterranean shoreline deposits with similar fauna: Older Eutryrrhenian and Neotryrrhenian. Glacial loess and colluvium partly covered these deposits. Early U/Th dates were seen as problematic because of recrystallization of carbonate matrix, but they were sufficient to dispute an older Holocene interpretation some 15 years before my visit from Rhodes Fairbridge. Elevation of deposits at these sites ranged from 1.5m for younger to 3.5m for older LIG. My notes say, “The last interglacial (if Eutryrrhenian is equivalent) must have had a distinctive SST regime (worldwide?)”. I was yet to work up a similar story from the NSW sites.

Visits to Nice and Monaco, Tarquinia in Italy, and along the Israel coast south of Haifa did not reveal as much on LIG deposits as I was hoping. Those at Nice were quite debateable I discovered. However, there was still a lot to see in the way of older Quaternary and discuss shoreline changes during lower and higher sea level periods. At this time controversy ranged on the conditions under which aeolianites were deposited and cemented including relation to sea levels. There were also debates on extent of tectonism in elevating or depressing palaeo sea-level features. I will leave all that to another blog, except that tectonics is a factor in one of the two recent papers on LIG sea levels that now deserve comment.

The first paper is by an Italian team (F. Pasquetti et al., 2021, “Chronology of the Mediterranean sea-level highstand during the Last Interglacial: a critical review of the U/Th-dated deposits”, J. Quat. Science, August 2021). This is a very thorough and comprehensive review looking at 323 relative sea level (RSL) ages around the Med Basin. Even with all this data on LIG, RSL “is still not fully understood despite a plethora of morphological, stratigraphic and chronological studies carried out on highstand deposits of this area”. Even with this critical appraisal of dates on corals, molluscs, and speleothems they conclude that further reappraisal is needed using “advanced geochronological methods”. One aim was to differentiate dates related to MIS 5e (LIG) from periods MIS 5c and 5a. Of dates they regard as “reliable” they attribute 65% to MIS 5e, none that are unequivocal to 5c and 17% to 5a. Of course they discuss our friend the Strombus (not all that reliable for dating, sad!), and the two-fold split in Tyrrhenian sequences. There is little doubt about age range of Eutryrrhenian (126-116 ka), but still doubt if Neotryrrhenian constitutes different facies or a younger second peak. Mallorca is singled out as having contrasting results. Tectonics was shown to be a factor at some locations (e.g. Nice). These authors were most concerned at improving chronological constraints as “many Mediterranean stratigraphic reconstructions are based on unreliable chronological constraints”.

The second paper takes a more global approach to the LIG (B. Dyer et al., 2021, “Sea-level trends across The Bahamas constrain peak last interglacial ice melt”, PNAS , 118, no.33). These US, German and Canadian authors do not question the reliability of dates in the same way as Pasquetti et al. They are more concerned about interpreting the elevation and ages of deposits and correcting the elevations for the effect of glacio-isostatic adjustment (GIA) associated with the growth and decay of the Laurentide Ice Sheet (LIS). The aim is to use this analysis to improve the precision and accuracy of estimating Global Mean Sea Level (GMSL or eustatic ice-equivalent level). The Bahamas is seen as tectonically stable and subject to peripheral forebulge effects of the LIS. Their work involved an assessment of the GIA model process and an analysis of U/Th ages at the interface of aeolian and beach facies and corals from underlying marine limestone. A “marine limiting” trend was observed from high at the northern end of The Bahamas archipelago to lower towards the southwest, a pattern which is GIA predicted. It becomes interesting to see how GIA-corrected sea level observations provide at 95% confidence a LIG peak elevation of +1.2m with only a 5% probability that the peak would exceed 5.3m. They infer an average 0.5m minor dip in sea level during the LIG agreeing with those who see a double peak. On the basis of this work they are critical of those who have estimated GMSL from heights of LIG deposits between +6 to 9m. Their lower LIG level enables them to conclude that “polar ice sheets may be less sensitive to high-latitude warming than previously thought”. Moreover, as such insolation driven changes affect the two hemispheres at different times. However, with greenhouse gas warming, we may expect both hemispheres to respond more equally. This argument is not followed through except to add the further rider that caution is needed in interpreting GIA models in different areas because differences in the Earth’s structure. Hydro-isostasy is not included in this study.

When I look back at my paper with John Chappell in 1978, I can see we had ideas that made sense although a bit wild with the surge hypothesis. Remarkably the top band of our LIG envelope (Fig 1a) looks very similar to that of Dyer et al. (Fig. 4) where 95% of solutions fall within their outer envelope. I again pay tribute to John’s brilliance in this study. It is so very gratifying to see that some 40 or so years on there are scientists devoting so much energy and resources to this fascinating period of Earth’s history to which we owe a little to our friend the “Strombus”.

Bruce Thom  

Words by Prof Bruce Thom. Please respect the author’s thoughts and reference appropriately: (c) ACS, 2021. For correspondence about this blog post please email