European Glacial Landscapes: The Last Deglaciation

European Glacial Landscapes: Last Deglaciation brings together relevant experts on the history of glaciers and their impact on the landscape of the main European regions. Soon after the Last Glacial Maximum, a rapid process of the glacial retreat began throughout Europe. This was interrupted several times by abrupt climate cooling, which caused rapid, although moderate, re-advance of the glaciers, until the beginning of the Holocene when the climate became relatively stable and warm. These successive glacial advances and retreats during the Last Deglaciation have shaped much of the European landscape, reflecting abrupt climatic fluctuations.

As our knowledge of abrupt climate changes since the Last Glacial Maximum progresses, new uncertainties arise. These are critical for understanding how climate changes disseminate through Europe, such as the lag between climate changes and the expansion or contraction of glaciers as well as the role of the large continental ice sheets on the European climate. All these contributions are included in the book, which is an invaluable resource for geographers, geologists, environmental scientists, paleoclimatologists, as well as researchers in physics and earth sciences.

• Provides a synthesis that highlights the main similarities or differences, through both space and time, during the Last Deglaciation of Europe
• Features research from experts in quaternary, geomorphology, palaeoclimatology, palaeoceanography and palaeoglaciology on the Last Deglaciation in Europe during Termination 1 and the important Late Pleistocene-Holocene transition
• Includes detailed colour figures and maps, providing a comprehensive overview of the glacial landscapes of Europe during the last deglaciation

• Language: English
• Publisher: Elsevier Science
• Release date: Sep 16, 2022
• ISBN: 9780323985116

Book preview

Part I

Introduction

Outline

Chapter 1 Introduction

Chapter 2 The terminations of the glacial cycles

Chapter 3 Previous synthesis of the Last Deglaciation in Europe

Chapter 1

Introduction

David Palacios¹, Nuria Andrés¹, Philip D. Hughes² and José M. García-Ruiz³,    ¹Research Group on Physical Geography of High Mountains, Department of Geography, Complutense University of Madrid, Madrid, Spain,    ²Department of Geography, School of Environment, Education and Development, University of Manchester, Manchester, United Kingdom,    ³Instituto Pirenaico de Ecologıá (IPE-CSIC), Campus de Aula Dei, Zaragoza, Spain

Abstract

Throughout the Quaternary, glaciers tended to extend across the north of the large continents of the Northern Hemisphere. This trend of ice expansion was interrupted by abrupt and rapid melting of the ice, called deglaciation or termination. The Last Deglaciation started between 21 and 18 ka and ended between 12 and 9 ka depending on the regions. In this book, the period from 18.9 to 11.7 ka has been selected to study the effects of deglaciation on the European landscape, which is its main objective. Part I justifies the need for this book and its characteristics, in addition to studying the origin of these terminations and analysing the few previous studies that have tried to synthesise these studies. Part II examines the climate of Europe during the Last Deglaciation, its abrupt changes in trends and the possible origin of these changes. Part III studies the landscapes of Europe shaped during the first period of the deglaciation or main deglaciation (18.9–14.6 ka) with special attention to the cold phase of Heinrich Stadial 1. Part IV looks at the landscapes of Europe shaped during the second period, the Bølling–Allerød Interstadial (14.6–12.9 ka) which was dominated mainly by a warm climate. Part V analyses the last period of the deglaciation interval, the cold Younger Dryas Stadial (12.9–11.7 ka) and its effects on the European landscape. Part VI attempts to correlate the evolution of the climate described in Part II with the extent of the glacial landscapes in Parts III–V. The ages described in the book and other parameters, as well as the maps, are standardised so that the glacial landforms of different regions and periods can be compared.

Keywords

European glacial landscapes; Last Deglaciation book; structure and content of the book; standardised parameters used

Chapter outline

Outline

1.1 The importance of the Last Deglaciation in the European landscape 3

1.2 The intensity of climatic changes during the Last Deglaciation and the delineation of the main subperiods 4

1.3 Objectives of the book 5

1.4 The glaciated European regions 6

1.5 The European landscapes shaped during the different periods of the Last Deglaciation 6

1.6 Standardised ages cited in the book 7

References 8

1.1 The importance of the Last Deglaciation in the European landscape

The first volume of this collection (Palacios et al., 2022) concluded by describing the period when most European glaciers reached their maximum expansion, during the so-called Last Glacial Maximum (29–19 ka according to that book; LGM) and the effect that this expansion had on the European landscape. The ice masses reached their maximum volume since 130 ka and, consequently, the sea level reached its lowest position, around 134 m lower than today (Lambeck et al., 2014). The European Arctic archipelagos, the Barents Sea, the Scandinavian Peninsula, eastern European Russia, northern central Europe, the North Sea, and much of the British Isles were covered by ice in the so-called European Ice Sheet Complex (EISC), which at 20 ka reached its maximum extent of 5,630,800 km² (Hughes et al., 2022a). Iceland was covered also by a large ice sheet that extended more than 40 km from the present coastline. In many of the European mountains, glaciers had reached their maximum extent since at least 130 ka at the same time. Whilst in some areas glaciers reached their overall maximum extent earlier in the Last Glacial Cycle, glaciers were still substantial and either at or close to their maximum extents in all areas across Europe at the LGM. The geomorphological significance of the LGM ice advance is therefore key in understanding the European landscape today, not only because of the intensity of glacial erosion, but also because of the effects on nonglaciated regions of the intense cold and extreme aridity that characterised this period (Hughes, 2022). In fact, the lowest proportion of CO2 in the atmosphere during the last 130 ka was recorded in the LGM period, together with the highest proportion of dust, which facilitated the expansion of periglacial and permafrost conditions, the drastic reduction of vegetation cover and the change in river regimes, as well as the expansion of aeolian deposits (Toucanne et al., 2022). The important consequences of the ice expansion during the LGM and of the climate system that prevailed during that period on the European landscapes was extensively explained in the previous book (Hughes et al., 2022b,c).

Moreover, in many cases, the geomorphological evidence of this maximum expansion is confined to the margin of the ice sheets and mountain glaciers, where terminal moraines were deposited. When the glaciers retreat, they transform the landscape perhaps most effectively and more definitively than in the LGM, filling depressions with glaciofluvial sediment or eroding outcrops and moraines through the high meltwater activity. This greater efficiency in remodelling the landscape is more striking if we take into account that the retreat of the ice occurred in a very short period of time, compared to the time it took to form the glaciers. Indeed, as is often the case throughout the Quaternary, long periods of glacial expansion, of around 100 ka in the last 800 ka, were interrupted by short periods of major melting of only about 10 ka. These periods, in which the huge ice sheets that covered Northern America and Europe completely melt from their maximum expansion, are called terminations or deglaciations. Chapter 2 of this book is devoted to explaining the working hypotheses underlying their origin and, in particular, that of the Last Deglaciation.

The Last Deglaciation, which includes Termination I, is generally considered to span the time period between the LGM (which in our definition ended at 19 ka) and the onset of the Holocene (11.7 ka), although it can be extended until the sea level stops rising and stabilises, 9 ka ago (Cheng et al., 2009; Denton et al., 2010). As we have seen in the previous book, the extent of the LGM was not synchronous in Europe, nor did it occur in all cases within the period delimited in that book (29–19 ka): some glaciers or sectors of the Ice Sheets may have expanded earlier, and in other exceptional cases later (Hughes et al., 2022a,b,c). Therefore, after the LGM had ended at 19 ka many glacier margins were already retreating, with some having commenced retreat several thousand years before. In any case, by 18.9 ka (the start of the Last Deglaciation defined here) a large majority of European glaciers were already in full retreat, the sea level had changed to a rising trend, which justifies the selection of this date for the start of the main period of the Last Deglaciation in Europe. However, the end of the Last Deglaciation is neither clear nor synchronous. With the arrival of the Holocene, many European glaciers had already disappeared beforehand, others still survived the beginning of the Holocene and did not disappear until the arrival of a warmer climate, from 9 ka onwards. In any case, the arrival of the Holocene marked the end of a period of intense cold and meant an intensive glacial retreat, even the disappearance of many glaciers, so it is also justified to select this date (11.7 ka) as the end of deglaciation.

The concept of Termination I or Last Deglaciation is complicated and does not allow a simple definition. Glacier behaviour was diachronous, that is glacier changes were occurring at different rates and times in different places. Furthermore, it cannot really be defined as a period when glaciers simply retreated, because the ice melting was not constant and the trend was often interrupted or even reversed with glacier advances observed in each hemisphere. Nor can it be considered a single type of climate, since it is characterised by abrupt climate changes and, moreover, with climatic inversions between the hemispheres. In reality, the only parameters that are more or less constant are the tendency for sea levels to rise and thus an overall negative glacial balance, and a constant tendency for CO2 to increase in the atmosphere (Chapter 2). In any case, from the beginning of the period selected as the Last Deglaciation to its end more than 85% of the glaciated surface of Europe had been left ice-free, with an impressive succession of limited advances and large retreats, which shaped the current landscape of northern Europe and most European mountains. On the other hand, the sea level had risen by 80 m at the onset of the Holocene (Lambeck et al., 2014), changing the configuration of the European coastlines. Studies on the pattern and timing of these deglaciation processes has increased enormously over the last decades. More and more sectors of the EISC, including those now covered by the sea, are becoming better known. The same is happening in many European mountains, with an extraordinary advance in knowledge in the last decades. Interestingly, whilst regional knowledge is advancing, studies that attempt to synthesise and interrelate this knowledge are not as well developed. Chapter 3 is devoted to the few previous studies on European landscapes resulting from deglaciation.

1.2 The intensity of climatic changes during the Last Deglaciation and the delineation of the main subperiods

The general warming trend during the Last Deglaciation that started around 20 ka in the North Atlantic, Europe and in Greenland, was caused most likely by an increase of summer insolation in the Northern Hemisphere. It was interrupted by abrupt changes, with sudden cooling of sea temperatures in the North Atlantic, accompanied by intense changes in global atmospheric circulation and oceanic currents. These cooling episodes arrived so suddenly that they are called events, but their consequences on the climate last for hundreds of years, so that longer cold periods can be delimited, which are called Stadials. Chapters 4 to 7 presents the succession and chronology of these stadial: pre-Heinrich Stadial 1 (~20–19 ka); Heinrich Stadial 1 (HS 1; between ~18 and 14.6 ka; also equivalent to the Oldest Dryas) and Younger Dryas Stadial (YD; between 12.9 and 11.7 ka). Between these two cold periods there was an abrupt warm period, the Bølling–Allerød Interstadial (B-A; between 14.6 and 12.9 ka). In reality, the chronological boundaries between these phases are not clear and depend rather on the climatic approach used. The temperature of the sea is deduced from the proportion in the ocean sediments of foraminifera shells, between the heavy oxygen isotope (δ¹⁸O), indicative of low evaporation in the sea, and the light isotope (δ¹⁶O), indicative of high evaporation, therefore, high values of δ¹⁸O are indicative of cold sea temperature. This climate variability has also been detected in air bubbles trapped in the deep Greenland ice cores for the last half a million years. In this case, the climatic cycles, called Dansgaard–Oeschger (D-O) cycles, were also identified by the δ¹⁸O/δ¹⁶O ratio, although in the opposite sense that in the marine sediments, therefore, low ratios δ¹⁸O are indicative of cold periods. Cold events are also deduced from the marine sediments, when some layers of the North Atlantic seabed revealed the existence of large discharges of icebergs, mainly from the Hudson Strait Ice Stream of the Laurentide Ice Sheet (LIS), but also from other sectors of the LIS and the EISC, which reached up to the latitude 43°N and deposited coarse-grained iceberg-rafted debris (IRD).

Therefore, alternating cold and warm periods during the Last Deglaciation were deduced from values of δ¹⁸O in the ocean sediments of foraminifera shells and in air bubbles trapped in the deep Greenland ice cores (in inverse sense) and in the IRD layers in the ocean sediment. In addition, changes in climate are revealed in changes in vegetation on the continent, as analysed in pollen records from lake and marine sediments. Finally, changes in climate are well represented, of course, in stillstands or advances of the glaciers during the cold periods and retreats during the warmer ones. The problem is that each of these parameters has its own methods of measuring the timing of changes and the results do not always match each other. This lack of complete synchrony has made it difficult to accept subdivisions of the Last Deglaciation or the names given to each of the proposed subdivisions (Hoek, 2009; Andrews and Voelker, 2018; Rasmussen et al., 2014; Mangerud, 2020). The correct timing of the changes between the different parameters, as well as the synchronisation of the individual parameters across the two hemispheres, is one of the main challenges of today’s science to discover the patterns and origin of the abrupt changes in climate during Last Deglaciation.

The timing synchronisation challenge is not easy to resolve, especially if environmental changes are diachronous across regions such as Europe, although the key to understanding this complexity is through improved geochronology and application of suitable dating techniques. Radiocarbon, the most commonly used dating method in marine and lacustrine sediments, is not used for dating glacial sediments, as they rarely contain organic matter. It has usually only been applied to sediments indirectly related to glaciers, such as glaciofluvial or glaciolacustrine sediments. Organic matter is also extremely scarce in Greenland ice cores, so dating must be complemented by other means, such as tephra or the content of certain gases. Optically Stimulated Luminescence methods, which constrain the time at which a mineral, normally quartz, contained in a sediment was last exposed to light, is problematic in moraines, and normally is applied only to date glaciofluvial or glaciolacustrine sediments. Other techniques occasionally used have been U-series to date sediments and landforms cemented by secondary carbonates and speleothems in cave systems under glaciated or paraglacial terrains. ⁴⁰Ar/³⁹Ar method has been applied to volcanic tephra overlying deposits associated with glaciation. However, during the first two decades of the 21st century, the most relevant progress to date glacial landforms and events has been the widespread use of the method for dating the exposure of rock surfaces to cosmogenic radiation. This exposition generates specific isotopes of certain elements in the rocks, with ¹⁰Be and ³⁶Cl being the most used cosmogenic isotopes. This method makes it possible to directly determine the time of exposition to this radiation (age) in erosional and sedimentary glacial landforms, with a huge range of ages, from a few hundred to millions of years. The problem with cosmogenic methods is that their degree of uncertainty is much greater than that obtained by radiocarbon, and often exceeds the duration of the short climatic periods that follow, one after the other, throughout the Last Deglaciation.

Despite these difficulties, the succession of cold and warm periods is evident in glacial records during the Last Deglaciation (chapter 4). Therefore, in order to compare the evolution of glaciers in different regions, it is possible to delimit different glacier phases from geomorphological evidence. In this book, we have decided to follow the most widely accepted chronological subdivisions that exist today established in different fields of Quaternary science. The first period delimited is the main deglaciation, from 18.9 to 14.6 ka, which includes warm phases in the first centuries, followed by the cold HS 1 in the second half. The climatic succession during this period is shown in Chapter 5. The second period is characterised by a predominantly warm climate, interrupted by short cold phases of varying magnitude and distribution, the B-A (14.6–12.9 ka), whose climatic characteristics are described in Chapter 6. The last period is the YD (12.9–11.7 ka), whose cold climate is discussed in Chapter 7. In addition, an introduction to the characteristics of these three periods and the background knowledge about glacial landscapes are presented in the Chapters 8, 27 and 44, respectively.

1.3 Objectives of the book

The Last Deglaciation involves a succession of complex processes in which the cryosphere, atmospheric and oceanic systems are interlinked, which lead to a series of abrupt changes in the Earth’s climate. Understanding these complex interrelationships will be a great help in understanding the evolution of the planet’s climate, including current and future changes. Over the past decades, excellent information has been accumulated in the different European regions about the evolution of the palaeoglaciers during Last Deglaciation. The key issues that have shaped many of these European landscapes during this deglaciation are being resolved. It is therefore time to synthesise, correlate and expose all this accumulated knowledge, to help decipher the dynamics of deglaciation. For these reasons, the aims of this book are (1) to synthesise the state of knowledge on the morphology and origin of European glacial landscapes derived from the Last Deglaciation in most of their regions; and (2) to place this glacial evolution in the context of the evolution of the atmosphere and ocean circulation and characteristics of global climate change during this period.

The main focus of the book is on glacial landforms in the different environments and regions, although attention is also devoted to the landforms and deposits indirectly related to glaciers (glaciofluvial terraces, glaciolacustrine sediments, etc.) or to paraglacial activity during the retreat of glaciers (e.g. debris flows, avalanches, landslides, collapses), as well as landforms that have expanded as the glaciers receded, such as periglacial and aeolian landforms. The book also gives great importance to submarine glacial landforms and deposits, developed under the sea or, sometimes, in subaerial conditions and subsequently flooded by the sea-level rise during deglaciation.

1.4 The glaciated European regions

The different European regions studied in this book are similar to that of the first volume of this collection (Palacios et al., 2022) and are presented in Part II of this book. These regions are divided into two groups. The first group includes the regions that have been covered by the EISC. These regions are treated first globally in each studied period (Chapters 9, 28, 45) and then individually in each of the regions that comprise them: Fennoscandia (i.e. Scandinavia, Denmark and Finland) (Chapters 10, 29, 46 and 47); Northern Central Europe (here northern Germany, northern Poland, northern Belarus and the Baltic countries; Chapters 11, 30 and 48); European Russia (here Pskov, Novgorod, Leningrad, Murmansk, Arkhangelsk, Moscow, Smolensk, Tver, Yaroslavl, and Vologda Russian Districts, as well as in the republics of Komi and Karelia) (Chapters 12, 31 and 49); Eurasian Arctic (Barents Sea, Kara Sea, associated High Arctic islands and archipelagos, and the adjacent mainland of Arctic Russia) (Chapters 13, 32 and 50); North Sea and Mid Norwegian Continental Margin (Chapter 14); Britain and Ireland (Chapters 15, 33 and 51). The second group includes mountainous regions not directly affected by the EISC: The Urals (Chapter 16); Iceland (Chapters 17, 34 and 52), which was covered by the Icelandic Ice Sheet, although it always remained separated from the EISC; the Tatra Mountains (Chapters 18, 35 and 53); the Romanian Carpathians (Chapters 19, 36 and 54); the Alps (Chapter 20, 37 and 55); the Pyrenees (Chapter 21, 38 and 56); the Iberian Peninsula ranges (Chapters 22, 39 and 57); the Italian Mountains, mainly the Apennines (Chapters 23, 40 and 58); the Balkans (Chapters 24, 41 and 59); and the mountains of the Anatolian peninsula (Chapters 25, 42 and 60), which despite technically being in Asia have been also included because of their fundamental contribution to the understanding of glaciers in southeast Europe and the eastern Mediterranean (Fig. 1.1).

Figure 1.1 The studied European regions. The regions affected by the EICS are highlighted in red: (1) Fennoscandia; (2) Northern Central Europe; (3) European Russia; (4) Eurasian Arctic; (5) North Sea (A) and Mid Norwegian Continental Margin (B); and (6); Britain and Ireland. The regions not directly affected by the EISC are highlighted in black: (7) Urals; (8) Iceland; (9) the Tatra Mountains; (10) the Romanian Carpathians; (11) the Alps; (12) the Pyrenees; (13) the Iberian Peninsula ranges; (14) the Italian Mountains; (15) the Balkans; and (16) the mountains of the Anatolian peninsula.

1.5 The European landscapes shaped during the different periods of the Last Deglaciation

Part III of the book is dedicated to presenting the state of knowledge on glacial evolution and on European landscapes derived from this evolution during the main deglaciation (18.9–14.6 ka). Chapter 8 discusses prior glacial knowledge in general for this subperiod and Chapter 9 that of the EISC. The EISC-dominated regions are presented in Chapters 9–15 and the mountain regions in Chapters 16–25. Chapter 26 synthesises and interrelates the knowledge from all regions and contrasts it with the climate information in Chapter 5.

Part IV of the book studies the state of knowledge on glacial evolution and on European landscapes derived from the B-A Interstadial (14.6–12.9 ka). Chapter 27 synthesises prior glacial knowledge in general for this subperiod and Chapter 28 that of the EISC. The EISC-dominated regions are presented in Chapters 29–33 and the mountain regions in Chapters 34–42. Chapter 43 synthesises and interrelates the knowledge from all regions and contrasts it with the climate information from this period in Chapter 6.

Part V of the book presents the state of knowledge on glacial evolution and on European landscapes derived from the YD Stadial (12.9–11.7 ka). Chapter 44 discusses prior glacial knowledge in general for this subperiod and Chapter 45 that of the EISC. The EISC-dominated regions are presented in Chapters 46–51 and the mountain regions in Chapters 52–60. Chapter 61 synthesises and interrelates the knowledge from all regions and contrasts it with the climate information from this period in Chapter 7.

The book also seeks to synthesise the glacial evolution during the Last Deglaciation across Europe in Part VI, Chapter 62, trying to underline the regional trends and the interrelationships between climate and glacial fluctuations.

1.6 Standardised ages cited in the book

The variety of dating methods used to date glacial sediments and landforms has already been shown above. Each method has improved over the last decades, so that ages calculated a few years ago may be somewhat different from those obtained today, even if the same dating method is used. This is especially so for radiocarbon and cosmogenic exposure dating. It is therefore necessary to update some previous age calculations, so that they are comparable across regions.

The terrestrial ¹⁴C dates have been calibrated using the CALIB 7.1 or OXCAL programs. The marine Accelerator Mass Spectrometry (AMS) ¹⁴C dates younger than 21.786 kyr BP have been calibrated: (1) using the CALIB 7.10 program and the ‘global’ marine calibration dataset (Marine13) (Hughen et al., 2004). To accommodate local effects, it was necessary to introduce the Delta R of the place where the core was retrieved following the suggestion of Reimer et al. (2013) (http://calib.org/marine/google/). (2) As an option, it was constructed with the age model using the open-source software Clam 2.2 (Blaauw, 2010) (http://www.chrono.qub.ac.uk/blaauw/clam.html) implemented in the R environment (R Development Core Team, 2013), or using Oxcal program version 4.2 (Ramsey et al., 2010) (https://c14.arch.ox.ac.uk/oxcal.html). Both programs have a choice for using the Marine13 calibration dataset and introduce the Delta R for each study place (http://calib.org/marine/google/). (3) IntCal20 was available during the writing of this book and was also used (Reimer et al., 2020).

Efforts were also made to unify cosmogenic exposure ages. This was mostly important for legacy exposure ages published more than around 5 years ago, because of newly established production rates and developments/improvements in the age calculators. CRONUScalc (Marrero et al., 2016a) was used on recalculations for the whole-rock cosmogenic ³⁶Cl ages on moraine boulders and glacially polished rock surfaces (http://cronus.cosmogenicnuclides.rocks/2.0/html/cl/). ¹⁰Be cosmogenic ages have been recalculated by using CRONUScalc (Marrero et al., 2016b) (http://cronus.cosmogenicnuclides.rocks/2.0/html/al-be/). The ‘CREp’ (Cosmic Ray Exposure Program) online calculator (Martin et al., 2017; http://crep.crpg.cnrs-nancy.fr/#/) has been also used, as the result difference is less than 5% with the forementioned calculators. With the aim of unifying the exposure age calculations for all samples, we have applied the Lifton–Dunai–Sato (LDS) scaling scheme (Lifton et al., 2014), the ERA40 atmospheric model (Uppala et al., 2005) and the geomagnetic database based on the LDS Framework (Lifton et al., 2014). In some cases, where topographic shielding has a major influence on exposure ages, the topographic shielding factor of each ¹⁰Be/³⁶Cl sampling site has been recalculated through the Topographic Shielding Factor through the ‘Topographic Shielding Calculator v.2’ belonging to the ‘CRONUSCalc’ Program (Marrero et al., 2016b): http://cronus.cosmogenicnuclides.rocks/2.0/html/topo/. However, for those samples whose field measurements for topographic shielding factor calculation are unreliable or unavailable, it has been obtained from the ‘Point-based Shielding Model’ GIS-tool devised by Li (2018) (http://web.utk.edu/~yli32/pointshielding.zip), which implements the method proposed by Balco et al. (2008). Where ages have not been recalculated, sometimes because of insufficient data available to perform a recalculation or where ages are from the most recent literature, this is indicated in the text and the published ages and the original sources are provided.

To facilitate the location of the sites mentioned in the text, a kml file is included for each of the study regions. It can be viewed on Google Earth or in a GIS.

In order to be able to compare the extent of glaciers in the different periods and regions, tables have been drawn up for each region and period, with synthesised glacier parameters. For mountain areas the minimum altitude of the glacier fronts and the value of the Equilibrium Lines Altitude (ELA) have been indicated for each period. Most of the ELAs have been calculated by the toe-to-headwall method (toe-to-headwall altitude ratio (THAR)), with a ratio of 0.4, except where another method is indicated in the chapter alongside the original source reference.

References

Andrews and Voelker, 2018 Andrews JT, Voelker AHL. Heinrich events (& sediments): a history of terminology and recommendations for future usage. Quaternary Science Reviews. 2018;187:31–40 https://doi.org/10.1016/j.quascirev.2018.03.017.

Balco et al., 2008 Balco G, Stone JO, Lifton NA, Dunai TJ. A complete and easily accessible means of calculating surface exposure ages or erosion rates from 10Be and 26Al measurements. Quaternary Geochronology. 2008;3(3):174–195 https://doi.org/10.1016/j.quageo.2007.12.001.

Blaauw, 2010 Blaauw M. Methods and code for ‘classical’ age-modelling of radiocarbon sequences. Quaternary Geochronology. 2010;5(5):512–518 https://doi.org/10.1016/j.quageo.2010.01.002.

Cheng et al., 2009 Cheng H, Edwards RL, Broecker WS, et al. Ice age terminations. Science. 2009;326(5950):248–252 https://doi.org/10.1126/science.1177840.

Denton et al., 2010 Denton GH, Anderson RF, Toggweiler JR, Edwards RL, Schaefer JM, Putnam AE. The last glacial termination. Science. 2010;328(5986):1652–1656 https://doi.org/10.1126/science.1184119.

Hoek, 2009 Hoek ZW. Bølling-Allerød Interstadial. In: Gornitz V, ed. Encyclopedia of Paleoclimatology and Ancient Environments Encyclopedia of Earth Sciences Series. Springer 2009;100–103. https://doi.org/10.1007/978-1-4020-4411-3_26 ISBN 978-1-4020-4551-6.

Hughen et al., 2004 Hughen KA, Baillie MG, Bard E, et al. Marine04 marine radiocarbon age calibration, 0–26 cal kyr BP. Radiocarbon. 2004;46(3):1059–1086.

Hughes, 2022 Hughes PD. Concept and global context of the glacial landforms from the Last Glacial Maximum. In: Palacios D, Hughes P, García-Ruiz JM, Andrés N, eds. European Glacial Landscapes: Maximum Extent of Glaciations. Elsevier 2022;355–358. https://doi.org/10.1016/B978-0-12-823498-3.00039-X.

Hughes et al., 2022a Hughes ALC, Winsborrow MCM, Greenwood SL. European ice sheet complex evolution during the Last Glacial Maximum (29–19 ka). In: Palacios D, Hughes PD, García-Ruiz JM, Andrés N, eds. European Glacial Landscapes: Maximum Extent of Glaciations. Elsevier 2022a;361–372. https://doi.org/10.1016/B978-0-12-823498-3.00038-8.

Hughes et al., 2022b Hughes PD, Palacios D, García-Ruiz JM, Andrés N. The European glacial landscapes from the Last Glacial Maximum Synthesis. In: Palacios D, Hughes P, García-Ruiz JM, Andrés N, eds. European Glacial Landscapes: Maximum Extent of Glaciations. Elsevier 2022b;507–516. https://doi.org/10.1016/B978-0-12-823498-3.00005-4.

Hughes et al., 2022c Hughes PD, Palacios D, García-Ruiz JM, Andrés N. The importance of European glacial landscapes in a context of great climatic variability. In: Palacios D, Hughes PD, García-Ruiz JM, Andrés N, eds. European Glacial Landscapes: Maximum Extent of Glaciations. Elsevier 2022c;519–522. https://doi.org/10.1016/B978-0-12-823498-3.00047-9.

Lambeck et al., 2014 Lambeck K, Rouby H, Purcell A, Sun Y, Sambridge M. Sea level and global ice volumes from the Last Glacial Maximum to the Holocene. PNAS. 2014;111:15296–15303 https://doi.org/10.1073/pnas.1411762111.

Li, 2018 Li YK. Determining topographic shielding from digital elevation models for cosmogenic nuclide analysis: a GIS model for discrete sample sites. Journal of Mountain Science. 2018;15(5):939–947 https://doi.org/10.1007/s11629-018-4895-4.

Lifton et al., 2014 Lifton N, Sato T, Dunai TJ. Scaling in situ cosmogenic nuclide production rates using analytical approximations to atmospheric cosmic-ray fluxes. Earth and Planetary Science Letters. 2014;386:149–160 https://doi.org/10.1016/j.epsl.2013.10.052.

Mangerud, 2020 Mangerud J. The discovery of the Younger Dryas, and comments on the current meaning and usage of the term. Boreas. 2020;50(1):1–5 https://doi.org/10.1111/bor.12481.

Marrero et al., 2016a Marrero SM, Phillips FM, Caffee MW, Gosse JC. CRONUS-Earth cosmogenic ³⁶Cl calibration. Quaternary Geochronology. 2016a;31:199–219 https://doi.org/10.1016/j.quageo.2015.10.002.

Marrero et al., 2016b Marrero SM, Phillips FM, Borchers B, Lifton N, Aumer R, Balco G. Cosmogenic nuclide systematics and the CRONUScalc program. Quaternary Geochronology. 2016b;31:160–187 https://doi.org/10.1016/j.quageo.2015.09.005.

Martin et al., 2017 Martin LCP, Blard PH, Balco G, et al. The CREp program and the ICE-D production rate calibration database: a fully parameterizable and updated online tool to compute cosmic-ray exposure ages. Quaternary Geochronology. 2017;38:25–49 https://doi.org/10.1016/j.quageo.2016.11.006.

Palacios et al., 2022 Palacios D, Hughes PD, García-Ruiz JM, Andrés N. European Glacial Landscapes: Maximum Extent of Glaciations Elsevier 2022; Elsevier. ISBN978-0-12-823498-3.

R Development Core Team, 2013 R Development Core Team, 2013. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, http://www.r-project.org.

Ramsey et al., 2010 Ramsey CB, Dee M, Lee S, Nakagawa T, Staff RA. Developments in the calibration and modeling of radiocarbon dates. Radiocarbon. 2010;52(3):953–961 https://doi.org/10.1017/S0033822200046063.

Rasmussen et al., 2014 Rasmussen SO, Bigler M, Blockley SP, et al. A stratigraphic framework for abrupt climatic changes during the Last Glacial period based on three synchronized Greenland ice-core records: refining and extending the INTIMATE event stratigraphy. Quaternary Science Reviews. 2014;106:14–28 https://doi.org/10.1016/j.quascirev.2014.09.007.

Reimer et al., 2013 Reimer PJ, Bard E, Bayliss A, et al. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon. 2013;55(4):1869–1887 https://doi.org/10.2458/azu_js_rc.55.16947.

Reimer et al., 2020 Reimer PJ, Austin WE, Bard E, et al. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0–55 cal kBP). Radiocarbon. 2020;62(4):725–757 https://doi.org/10.1017/RDC.2020.41.

Toucanne et al., 2022 Toucanne S, Landais A, Naughton F, Rodrigues T, Vázquez Riveiros N, Sánchez Goñi MF. The Global Last Glacial Maximum: the Eastern North Atlantic (marine sediments) and the Greenland Ice Sheet climatic signal. In: Palacios D, Hughes PD, García-Ruiz JM, Andrés N, eds. European Glacial Landscapes: Maximum Extent of Glaciations. Elsevier 2022;189–194. https://doi.org/10.1016/B978-0-12-823498-3.00052-2.

Uppala et al., 2005 Uppala SM, Kållberg PW, Simmons AJ, et al. The ERA-40 re-analysis. Quarterly Journal of the Royal Meteorological Society. 2005;131(612):2961–3012 https://doi.org/10.1256/qj.04.176.

Chapter 2

The terminations of the glacial cycles

David Palacios¹, Philip D. Hughes², María F. Sánchez-Goñi³, José M. García-Ruiz⁴ and Nuria Andrés¹,    ¹Research Group on Physical Geography of High Mountains, Department of Geography, Complutense University of Madrid, Madrid, Spain,    ²Department of Geography, School of Environment, Education and Development, University of Manchester, Manchester, United Kingdom,    ³Environnements, Paléoenvironnements Océaniques et Continentaux, Ecole Pratique des Hautes Etudes (EPHE, PSL), Pessac, France,    ⁴Instituto Pirenaico de Ecologı´a (IPE-CSIC), Campus de Aula Dei, Zaragoza, Spain

Abstract

The geography of the Earth at the end of the Tertiary, with the new arrangement of continents, oceans and the distribution of mountain ranges, especially since the opening of the Drake Strait and the closing of the Isthmus of Panama, favoured a global cooling trend that culminated in the Quaternary. In addition to the ice sheets of Antarctica and Greenland, glaciers during the Quaternary tended to expand, especially on the Northern Hemisphere continents and in the mountains. The expansion ended abruptly for short periods, of about 10 ka, during which these glaciers largely disappeared. These periods are called terminations and mark the end of different glacial cycles. In the first half of the Quaternary, terminations occurred every 41 ka, but in the last 800 ka, terminations have been delayed, whilst glaciers could extend over larger areas, occurring every 100 ka. The onset of the terminations and their dynamics remains a mystery, but it coincides with a series of processes, where it is difficult to know what the cause is and what is the effect. It appears that when the glaciers in the Northern Hemisphere reach their maximum extent, an increase in insolation in the mid-latitudes of the Northern Hemisphere causes the onset of global termination. Once termination has begun, a series of temperature changes take place, intense in the Northern Hemisphere and milder in the Southern Hemisphere, but with inverse trends, and in direct relation to changes in the intensity of the Atlantic Meridional Overturning Circulation (AMOC) and to latitudinal changes in atmospheric circulation. Despite these changes in temperature, CO2 in the atmosphere increases throughout the termination, albeit with varying intensity. Once the balance between the AMOC and the proportion of CO2 in the atmosphere is in equilibrium, the temperature stabilises and the termination ends, leading to the onset of an interglacial optimum. This occurs when the northern continental ice sheets have disappeared or have reduced their extension.

Keywords

Quaternary; glacial cycles; glacial termination; deglaciation; interglacial

Chapter outline

Outline

2.1 The origin of glacial terminations 11

2.2 The trigger and pattern of the glacial terminations 12

2.3 The onset of the Last Deglaciation (Termination I) 15

2.4 The dynamics of glacial terminations 17

2.5 Conclusions 18

References 19

2.1 The origin of glacial terminations

The intense geological evolution of the Earth over the last 60 million years changed the distribution of seas and continents, as well as its atmospheric circulation and ocean currents (Lisiecki and Raymo, 2005, Lisiecki, 2014; Ehlers et al., 2018; Lagemaat et al., 2021; Starr et al., 2021; Palacios et al., 2022): (1) large continental landmasses were distributed along the Northern Hemisphere, with axis at the 65°N parallel; (2) Antarctica was isolated at the South Pole with the opening of the Drake Strait, surrounded by the circumpolar Antarctic current; and (3) continents and large mountain and plateaus systems were arranged in the meridian direction, leading to a change in the general atmospheric circulation and ocean current systems, reinforcing stable pressure centres, strong polar fronts and the formation of large gyres in the oceans. This process culminates mainly from the closure of the Isthmus of Panama around 13–15 million years ago (Montes et al., 2015), which led to the formation of the Atlantic Meridional Overturning Circulation (AMOC) transporting warm and salty waters from the Caribbean to the North Atlantic and sinking carbon and nutrients into the deep ocean towards the South Atlantic (Lynch-Stieglitz, 2017; Lozier et al., 2019; Kirillova et al., 2019).

As a consequence of all these geological and geomorphological changes, the Earth’s climate started a cooling tendency. For the first time since more than 200 million years ago, glaciers became established in continents: in Antarctica 44.5 million years ago (Ingólfsson, 2004; Barker et al., 2007) and in Greenland 35 million years ago (Bierman et al., 2014). These glacial ice sheets underwent several oscillations, but as the climate continued to cool, they became stable 14 million years ago, at about a size similar to the present (Shevenell et al., 2004; Bierman et al., 2014).

The cooling trend continued on Earth and since 2.58 million years ago (with the beginning of the Quaternary) the ice sheets not only remained stable in Antarctica and Greenland, but tended to spread over large parts of the northern continents and over the great mountains of the planet (Gibbard and Cohen, 2008). In fact, for most of the Quaternary, these sectors of the northern continents and high mountains have been occupied respectively by ice sheets and mountain glaciers (Ehlers et al., 2018).

However, the expansion of the northern ice sheets has been interrupted by periods when they suddenly and completely melted in a few thousands of years, until the glaciers slowly expanded again across northern Europe and America (Fig. 2.1). This glacial expansion has been interrupted up to 41 times during the Quaternary, but 14 of them after the glaciers expanded reaching the oceans surrounding these continents (Ehlers et al., 2018). These periods of massive meltdowns have been called ‘terminations’ or ‘deglaciations’ and are considered the end of a glacial cycle (Raymo, 1997; Cheng et al., 2009, 2016; Denton et al., 2010) (Fig. 2.2). The concept of ‘terminations’ was originally defined as early as the 1970s in the deep-sea sedimentary sequences as the time interval between the highest and the lowest values of the benthic foraminifera oxygen isotope ratio (δ¹⁸Ob), following seminal work by Broecker and van Donk (1970) and Broecker (1984). Their duration is generally of ~10,000 years, excluding Termination V, at ~430 ka, that lasted almost 30,000 years (Sarnthein and Tiedemann, 1990). In the first half of the Quaternary, these terminations occurred every 41 ka, but since about 800 ka ago, glaciers have expanded over larger areas and for longer periods, and terminations have been delayed until about 100 ka (between 80 and 120 ka in the last 10 cycles; Hughes and Gibbard, 2018), although the intensity of melting over only a few thousand years has been maintained (Barker et al., 2011; Spratt and Lisiecki, 2016) (Figs 2.2 and 2.3). Therefore one of the enigmas that research has been trying to unveil is the cause of these terminations.

Figure 2.1 Evolution of the δ¹⁸O record in marine sediments, showing the evolution of the Earth’s temperature and the proportion of water in glaciers over the last 5.5 million years. Note the terminations are being increasingly delayed and start and finish at more extreme temperatures over the last 800 ka. Data from Lisiecki, L., Raymo, M.E., 2005. A Pliocene–Pleistocene stack of 57 globally distributed benthic δ18 O records. Paleoceanography, 20, PA1003. https://doi.org/10.1029/2004PA001071

Figure 2.2 Evolution of parameters indicating the climate evolution over the last 800 ka and the duration of the Terminations (grey band). (A) Stable isotope (δDeuterium) record from the EPICA Dome C (EDC) ice core showing temperature variations in Antarctica ( Jouzel et al., 2007). (B) Composite CO2 record (0–800 ka) for Antarctica (Lüthi, et al., 2008). (C) Dust record from the EDC ice core covering 0–800 ka. Measurements of dust concentration using both laser and Coulter Counter methods ( Lambert et al., 2008). (D) Global sea level evolution (Spratt and Lisiecki, 2016). Marine Isotope Stages (MIS) are given in italics, Arabic numerals.

Figure 2.3 Sea level evolution over the last 250 ka and the duration of the Terminations (grey band), from the time the sea level starts to rise until it stabilises. Note the slow growth of the Northern Hemisphere Ice Sheets, whose vast changes in size between glacials and interglacials dominate the marine isotopic record (Hughes et al., 2018, 2020). The effect of abrupt melting of the Northern Ice Sheets on the marine isotopic record is clear at the end of each cycle (Terminations). Data from Spratt, R.M., Lisiecki, L.E., 2016. A Late Pleistocene sea level stack. Climate of the Past 12 (4), 1079–1092. https://doi.org/10.5194/cp-12-1079-2016

2.2 The trigger and pattern of the glacial terminations

Milankovitch (1941) related the onset of terminations with an increase of solar radiation at 65°N caused by cyclical variations in the Earth’s orbital movements (precession cycle of 21 ka each, obliquity cycle of 40 ka each, and eccentricity cycle of 100 ka each; the combination of all is called ‘Incident solar radiation’). Hays et al. (1976) discovered that the temperature evolution of the oceans, obtained from δ¹⁸O in the sediment cores, directly reflected the orbital effect. As eccentricity (cycles of 100 ka) does not have a large effect on overall global solar radiation, the increase of solar radiation at 65°N can be a consequence of four or five precessional cycles (Hays et al., 1976; Cheng et al., 2016) or of two or three obliquity cycles (Huybers and Wunsch, 2005; Maslin and Brierley, 2015), or a combination of these two orbital movements (Imbrie and Imbrie, 1986; Tziperman et al., 2006; Huybers, 2011). In fact, the distribution of the huge continents in the Northern Hemisphere with the associated albedo and the existence of the AMOC (Denton et al., 2010) amplify the response of the planet’s climate to the original small changes in the seasonal and latitudinal distribution of insolation (Lisiecki, 2014; Ehlers et al., 2018; Lagemaat et al., 2021; Starr et al., 2021). In fact, precessional cycles and the seasonal patterns of solar receipt exert strong controls on global ice volume, with deglaciations triggered every fourth or fifth precessional cycle (Raymo 1997; Ridgwell et al., 1999; Hughes and Gibbard, 2018). Although orbital effects are still considered a key factor in triggering terminations (Maslin, 2016), not all orbital phases that increase radiation in the middle latitudes cause terminations. Therefore orbital is a necessary but not sufficient cause (Cheng et al., 2009; Denton et al., 2010; Lisiecki, 2014).

Terminations onsets coincide with an increase in radiation in the mid-latitudes of the Northern Hemisphere but only when the Northern Hemisphere’s ice sheets have reached their maximum critical extent (Birchfield and Broecker, 1990; Imbrie et al., 1993; Raymo, 1997; Paillard, 1998; Denton et al., 2010; Abe-Ouchi et al., 2013; Deaney et al., 2017). This maximum critical growth occurred in logical coincidence with: (1) the lowest sea level (Clark et al., 2009; Yokoyama et al., 2018); (2) the maximum isostatic depression over the continents (Denton et al., 2010); consequently, most of these ices sheets become marine-based at their maximum (Gildor et al., 2014); (3) the maximum isostatic depression in the continents leads to an elevation of the Equilibrium Line Altitude (ELA) of the continental ice sheets (Denton et al., 2010); (4) the minimum CO2 in the atmosphere and, of course, the minimum air temperature in the planet (Waelbroeck et al., 2009; Shakun et al., 2012; Toucanne et al., 2022); and (5) the minimum precipitation and the minimum plant cover density at a global scale and the maximum dust in the atmosphere (Ellis and Palmer, 2016). For this reason, northern ice sheets reach their maximum expansion just before the onset of termination, in a moment called the Last Glacial Maximum (LGM) of each glacial cycle (Clark et al., 2009; Hughes et al., 2013; Yokoyama et al., 2018; Hughes, 2022; Toucanne et al., 2022; Vázquez-Riveiros et al., 2022).

When the ice sheets covering the Northern Hemisphere continents are at their maximum and critical growth, an increase by orbital effects in summer radiation at 65°N initiates a chain reaction of self-feeding processes, culminating in the massive melting of ice sheets at the end of terminations (Brook and Buizert, 2018). These processes are very complex and interact and feedback on each other (Fig. 2.4), including: (i) changes in atmospheric circulation and composition, especially in CO2 and CH4 (Deaney et al., 2017; Monnin et al., 2001; Shakun et al., 2012; Sigman et al., 2010; Sigman and Boyle, 2000); (ii) changes in ocean composition and circulation, (iii) changes in vegetation distribution and their effects on atmospheric CO2 (e.g. Hes et al., 2021 in review); (iv) changes in sea ice extent in response to changing global ocean temperatures, which in turn affects interhemispheric ocean circulation (Bereiter et al., 2018; Ferrari et al., 2014); (iv) changes in the interplay between glacial melting, atmosphere and oceans (Fogwill et al., 2017; Schmittner and Galbraith, 2008); and (v) changes in the dynamics of the ice sheet margin. The last point is an important consideration in glacial periods because ice margins are much more sensitive to the changes of the relative sea-level, under significant glacio-isostatic depression (Scourse et al., 2019) and particularly to the significant amplification of tides during this maximum glacial expansion (Griffiths and Peltier, 2009; Scourse et al., 2018).

Figure 2.4 Evolution of the main parameters related to the evolution of the climate and the cryosphere during Last Deglaciation (Termination I) YD: YD (Younger Dryas, 12.9–11.7 ka); B-A (Bølling–Allerød Interstadial 14.6–12.9 ka); MD (Main Deglaciation 18.9–14.6 ka). (A) Summer insolation at 65°N ( Berger and Loutre, 1991). (B) Changes in past atmospheric carbon dioxide concentrations preserved in the air trapped of the EPICA Dome C ice core along the Last Deglaciation according to Lüthi et al. (2008). (C) Evolution of Surface Sea Temperature (SST) through the Last Deglaciation. (C1) SST (southwestern Iberian margin, off Lisbon) during the Last Deglaciation has been inferred from the analysis of alkenones (UK’37) in the framework of the International Marine Global Change Studies (IMAGES) programme of MD95–2042 core (37°45′N, 10°10′W) according to Pailler and Bard (2002) with ages modelled in Sánchez-Goñi et al. (2008). (C2) SST in the Alboran Sea (western Mediterranean) during the Last Deglaciation has been inferred from the analysis of alkenones (UK’37) in the framework of the International Marine Global Change Studies (IMAGES) programme in MD95–2043 core (36°8′N, 2°37′W) according to Cacho et al. (1999). (D) Global Relative Sea Level evolution during the Last Deglaciation modelled by Lambeck et al. (2014).

The sequence of processes involved can be summarised (Broecker, 1998; Cheng et al., 2009; Denton et al., 2010; Zhisheng et al., 2011; Deaney et al., 2017) as follows: (1) the increase in radiation at the latitude where the continental ice sheets extend cause the rise of their ELAs, which were already high because they are on isostatically depressed continents under the weight of the ice; (2) the increase in ELAs entailed the massive loss of huge accumulation areas of the ice sheets; (3) in addition, when ice sheets reach the continental shelves the ice sheet margins become very sensitive to sea level change and therefore very unstable; (4) the instability of the marine-base ice-sheet and the imbalance of the glacier masses causes the collapse and subsequent melting of a large part of the marine-based ice sheet sectors; (5) the large glacial melt causes the sea level to rise, which increases the ease of ice-sheet to collapse; (6) the large amount of dust in the atmosphere, the rising sea levels, the increased precipitation and expansion of vegetation in the ice-free landmasses causes the reduction in albedo, which makes the orbital increased radiation at these latitudes more effective in increasing global temperature; (7) rising summer temperatures in the Northern Hemisphere and rising sea levels allow the Asian monsoon to recover, introducing moisture into the interior of the greater Asian continent; (8) the Intertropical Convergence Zone (ITCZ) moves northward; (9) the AMOC strengthens considerably and westerly winds intensify in the mid-latitudes of the Southern Hemisphere; and (10) this change in the general atmospheric circulation and in the intensity of ocean currents allows the release of large amounts of CO2 stored deep in the Southern Hemisphere oceans, so that the temperature of the planet will tend to rise over the course of the termination (Fig. 2.4).

Thus the northern ice sheets that had been growing, with slight pauses and retreats, for about 100 ka, disappear completely in about 10 ka. Once they have disappeared, the temperature of the planet stabilises in the interglacials, during about 10–30 ka, when they begin to expand again towards a new maximum (Ehlers et al., 2018).

The most important terminations occurred after major glacial expansions (the LGM of each cycle; Spratt and Lisiecki, 2016), especially at the end of MIS 16 (c.621 ka), MIS 12 (c.424 ka), MIS 6 (c.130 ka) and MIS 2 (c.19 ka), and were much less intense in more limited glacial expansions, such as at the end of MIS 14 (524 ka), MIS 10 (337 ka) and MIS 8 (243 ka) (Hughes et al., 2020). Obviously, the best known are those that ended the last two great glacial cycles, that is, MIS 6 (Termination II) and MIS 2 (Termination I or Last Deglaciation), when the most extreme climates and the greatest contrast between the maximum expansion of glaciers and their rapid melting during terminations were reached during the Quaternary (Hughes and Gibbard, 2018).

2.3 The onset of the Last Deglaciation (Termination I)

Terminations affected the two main northern continental ice sheets. The most important was the Laurentide Ice Sheet (LIS), which stretched across northern North America, in terms of its contribution to the general sea level (Lambeck et al., 2014) and its impact on the coupled ocean-atmosphere system during terminations (Broccoli and Manabe, 1987; Heinrich, 1988; Clark, 1994; Barber et al., 1999; Hemming, 2004; Stokes et al., 2016; Stokes, 2017). The European Ice Sheet Complex (EISC) was located in northeastern Europe as a result of the confluence of three smaller ice sheets, the Barents Sea Ice Sheet, Fennoscandian Ice Sheet and British–Irish Ice Sheet, at the peak of glacial expansion in each glacial cycle (Hughes et al., 2016, 2022b; Patton et al., 2016). The EISC is the second largest source of contribution to sea level fluctuations during the terminations (Lambeck et al., 2014). The LIS and the EISC reached their LGM in the last glacial cycle approximately between 29 and 19 ka, with various asynchronies within the expansion of both ice sheets (Stokes et al., 2016; Stokes, 2017; Palacios et al., 2020; Hughes et al., 2022a,b), just before the onset of Termination I or Last Deglaciation. The lowest sea level was reached at 26.5–19 ka, mainly driven by the expansion of the LIS (Clark et al., 2009). Most of the LIS and EISC fronts began to retreat from 21 ka onwards and glacier retreat was widespread by 19 ka, with the melting of major marine-based ice streams (Margold et al., 2018), except in those sectors where precipitation increased due to ice sheet deglaciation (Heath et al., 2018).

Most glaciers of the rest of the planet reached their LGM in the last glacial cycle also in the 31–19 ka period. The Antarctica Ice Sheet expanded across its continental shelf and, with some asynchrony within its various ice margins, began its retreat, after reaching its maximum expansion between 21 and 19 ka (Mackintosh et al., 2014; Bentley et al., 2014; Siegert et al., 2022). In this case, deglaciation mainly affected only the continental shelf and its contribution to sea level rise was very limited, compared to the effects of the northern ice sheets (Siegert et al., 2022). The permanent ice sheet of Greenland reached its maximum volume also during ~26.5–21 ka (Clark et al., 2009; Vasskog et al., 2015) and expanded to most of its continental shelf. Although with some asynchrony in its different sectors, this ice sheet as a whole had been receding since 21 ka, and the retreat accelerated considerably since 19 ka (Vasskog et al., 2015). In any case, its contribution to sea level rise during deglaciation did not exceed ~4% (Vasskog et al., 2015). The maximum extent of the Iceland Ice Sheet also coincided with the onset of deglaciation, causing a major collapse from 19 ka onwards (Benediktsson et al., 2022).

Many of the world’s great mountain ranges were covered by glaciers during the last glacial cycle and, in general, their maximum expansion coincided with the maximum extent of the two main northern continental ice sheets (Shakun et al., 2015). For example, in the great European mountain ranges, such as the Alps, the beginning of deglaciation occurred around 21–19 ka (Ivy-Ochs et al., 2022). The western mountains of North America and the volcanoes of Central America, reached their LGM at 26.5–19 ka, coinciding with the LIS and EISC (Palacios et al., 2020). Circumstances were different in the Himalayas, the most glaciated region outside the polar regions, where monsoon weakening during the LGM periods provoked intense aridity in some areas and prevented a major glacier growth. In any case, glacier retreat was widespread from 21 to 19 ka, during the Last Deglaciation (Owen et al., 2002; Owen, 2020; Yan et al., 2021). Changes in atmospheric circulation, for example, in the location of the ITCZ and the Southern Hemisphere westerlies during the LGM northern ice sheets, caused major changes in precipitation distribution in many Southern Hemisphere mountains, where in many cases the maximum glacier extent occurred before the LGM (Shakun et al., 2015). The existence of an aborted termination around 65–45 ka has even been hypothesised in the Southern Hemisphere, after glaciers had achieved their maximum extents (Schaefer et al., 2015). This pattern is evident in mountains of East Africa (Shanahan and Zreda, 2000; Mahaney, 2011), New Zealand (Schaefer et al., 2015; Darvill et al., 2016) and Kerguelen (Jomelli et al., 2018), as well as in the Andes. Here, an increase in precipitation facilitated the maximum extension of some tropical Andean glaciers after 19 ka (Mark et al., 2017; Martin et al., 2018), whilst temperate and subpolar Andean glaciers started their deglaciation from 21 to 19 ka (Palacios et al., 2020). Therefore although not all mountain regions reached their maximum glacier expansion in coincidence with the northern ice sheets, the onset of the main deglaciation is synchronous, around 19 ka (Schaefer et al., 2006; Shakun et al., 2015). In any case, the contribution of mountain glaciers to sea level rise during terminations was very limited compared to northern ice sheets (Shakun et al., 2015).

According to the previous sequence, it can be deduced that the distribution of Quaternary oceans, continents and mountains facilitates the expansion of glaciers in the polar areas and in the high mountain regions of the planet. Thus over long periods, glaciers tend to expand, unless this expansion itself causes intense aridity in some regions. During the gradual expansion of terrestrial glaciers, the formation of the two large northern ice sheets, the EISC and particularly the LIS, is the main cause of a significant drop in the sea level. A certain asynchrony in reaching the maximum extension of the glaciers occurs between the different regions of the planet, although most of them coincide with the LGM of the northern ice sheets. When these ice sheets reach their critical size and the oceans record their lower level, an increase in radiation at their latitudes, due to orbital effects, triggers a series of chain reactions that lead to (1) the release of large amounts of CO2 stored in the oceans, (2) a decrease in albedo, and (3) a general increase in global temperature. This facilitates a general retreat of the glaciers across the whole planet during terminations (He et al., 2013). The opposite occurs during glacial inceptions when low CO2 and low insolation are potential triggers for global ice buildup (Ganopolski et al., 2016), and this highlights the importance of the combination of CO2 and orbital forcing in driving glacial cycles. At the end of glacial cycles, when the northern ice sheets disappear completely or they are significantly reduced in size, then the termination ends, the climate and sea level stabilise and a warm interglacial period begins (Shakun and Carlson, 2010). This book analyses precisely the consequences of Termination I or the Last Deglaciation to the onset of the Holocene (18.9–11.7 ka) on the European landscapes. Moreover, the Termination I itself is considered to extend up to 9 ka, when the minima in benthic foraminifera d¹⁸O values are reached, in coincidence with the decrease in the sea-level rise (Alley et al., 2005).

2.4 The dynamics of glacial terminations

The rapid melting of northern continental ice sheets and, almost simultaneously, many of the world’s mountain glaciers during terminations, marks the end of glacial cycles. This culminates with the beginning of the interglacials when glaciers either disappear entirely or stabilise to much-reduced positions in equilibrium with a much warmer global climate. However, this progressive melting is not a constant and steady process. Instead, the terminations are characterised by abrupt climatic oscillations that punctuate the deglaciation trend, causing periods of glacier stabilisation and even glacial readvances (Cheng et al., 2009). The melting of glaciers depends primarily on rising temperatures, which in turn depend primarily on the increase in insolation and the atmospheric CO2 content (Shakun et al., 2012). The CO2 content of the atmosphere increases, with some fluctuations, over the whole termination periods (Monnin et al., 2001), although temperature oscillates, sometimes sharply, and often inversely in the two hemispheres (Barker et al., 2009, 2010) (Fig. 2.5). In particular, the Last Deglaciation (Termination I), was punctuated by abrupt temperature variations in Greenland related to the Dansgaard–Oeschger (D-O) cycles 2 and 1, and associated with more limited temperature changes in Antarctica, termed Antarctic Isotope Maxima (Dansgaard et al., 1993; EPICA community members, 2006). The contrast of the two records clearly shows that when Greenland is in a Greenland stadial (very cold period), Antarctica gradually warms and during the Greenland interstadial (warm period) it gradually cools (Stocker and Johnsen, 2003; Pedro et al., 2011, 2018) (Fig. 2.5).

Figure 2.5 Evidence of opposite climate trends during the different periods of the Last Deglaciation (Termination I) between Antarctica and Greenland according to the temperature records in the respective ice cores YD: YD (Younger Dryas, 12.9–11.7 ka); B-A (Bølling–Allerød Interstadial 14.6–12.9 ka); MD (Main Deglaciation 18.9–14.6 ka). NGRIP curve: δ¹⁸O records of North Greenland Ice Core Project (NGRIP) in the centre of Greenland (75°6″N, 42°19″W, 2917 m, 3085 m a.s.l.) according to Rasmussen et al. (2014) and Seierstad et al. (2014). The temperature axis on the right side indicates approximate surface temperatures at NGRIP modelled according to Kindler et al. (2014). Epica Dronning Maud Land (EDML) curve: δ¹⁸O records of European Project for Ice Coring in Antarctica (EPICA) in the ice core from EDML 75°S, 123°E, 2892 m a.s.l. (EPICA community members, 2006). The temperature axis on the right side indicates approximate surface temperatures at EDML as derived from the spatial δ¹⁸O/temperature gradient (EPICA community members, 2006).

The rapid melting of the northern continental ice sheets cools the North Atlantic Ocean and slows down the AMOC, limiting the transfer of heat from the tropics to the temperate and high latitudes of the Northern Hemisphere (Barker et al., 2009; Deaney et al., 2017; Muschitiello et al., 2019). The rapid melting caused by an initial warming of the Northern Hemisphere mid-latitudes due to orbital effects, reverses the thermal trend of the North Atlantic, and expands the winter sea ice and the northern polar front, resulting in very cold winters in the northern continents (Cheng et al., 2009; Denton et al., 2010). The cooling of the Northern Hemisphere pushes the northern westerlies southward (Naughton et al., 2019; Chen et al., 2019; Hudson et al., 2019), and weakens the Asian monsoon (Cheng et al., 2016, 2019). Consequently, the ITCZ, the southern trade winds, and the southern westerlies move to the south, which ultimately leads to ventilation and upwelling in the southern oceans. This causes the release of CO2 into the atmosphere, amplifying global warming (Broecker, 1998; Stephens and Keeling, 2000; Stocker and Johnsen, 2003; Anderson et al., 2009; Skinner et al., 2010; Brook and Buizert, 2018; Clementi and Sikes, 2019). In fact, the rise in CO2 is synchronous with the increase in Antarctic temperatures (Ahn et al., 2012; Beeman et al., 2019) in Termination I, as this increase in CO2 along the terminations overrides the cooling effect from orbital variations in the Southern Hemisphere (He et al., 2013). Moreover, this evolution was complex and not always evident during Termination II (Landais et al., 2013).

The cooling of the Northern Hemisphere leads to a stabilisation and even a readvance of the northern ice sheets, with a reduction in meltwater input that restores the intensity of the AMOC, the northward migration of the ITCZ, the intensification of Asian monsoon and the cooling of the southern oceans (Cheng et al., 2009; Denton et al., 2010). An interplay between the two hemispheres, called ‘bipolar seesaw hypothesis’, involving the effects of meltwater input from the northern continental ice sheets and the upwelling and CO2 release in the southern oceans govern the terminations, where the AMOC is the conveyor belt (Broecker, 1998; Barker et al., 2009, 2010; Wang et al., 2015; Lynch-Stieglitz, 2017, Brook and Buizert, 2018; Pedro et al., 2018), although the exact mechanism that causes this seesaw is not entirely clear (Pedro et al., 2018). To achieve an effective correlation to demonstrate the ‘bipolar seesaw hypothesis’ and that the D-O cooling drives the corresponding Antarctic warming requires a precise synchronisation of the Greenland and Antarctic ice core records. The results obtained so far seem to demonstrate the delay of the Antarctic temperature response with respect to the D-O event, with a very short time lag (208±96 years: WAIS Divide Project Members, 2015; 122±24 years: Svensson et al., 2020). However, other observations cast doubt on the direct link between northern cooling and the increase in atmospheric CO2 and Antarctic temperature on a more detailed time scale (Zheng et al., 2021). In any case, whilst glaciers in one hemisphere stabilise or readvance, those in the other retreat drastically (Shakun et al., 2015; Pedro et al., 2015; Palacios et al., 2020). When, finally, the northern ice sheets disappear or are