Periglacial Geomorphology E-Book

Colin K. Ballantyne

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 1392

Veröffentlichungsjahr: 2017

Beschreibung

A FASCINATING AND INFORMATIVE EXPLORATION OF PERIGLACIAL PROCESSES, PAST AND PRESENT, AND THEIR ROLE IN LANDSCAPE EVOLUTION

Periglacial Geomorphology presents a comprehensive introduction to the processes that operate in present periglacial environments and discusses the inferences that can be drawn about former periglacial environments from those processes. Organized into six parts, the book opens with the historical and scientific context of periglacial geomorphology and the nature of periglacial environments. Following chapters provide systematic coverage of the full range of topics germane to a thorough understanding of periglacial geomorphology, including:

• The physics of ground freezing and thawing, characteristics of permafrost, and the nature and origin of underground ice
• Characteristics, formation and significance of landforms, sediments, and structures associated with permafrost, permafrost degradation, and seasonal ground freezing and thawing
• Rock weathering in periglacial environments, periglacial processes operating on hillslopes, and the characteristic landforms produced by rock breakdown and slope processes in cold environments
• The operation of fluvial, aeolian and coastal processes in cold environments, and the resulting distinctive landforms and sediments
• The use of relict periglacial features to reconstruct past cold environments in midlatitude regions and the responses of periglacial environments to recent and predicted climate change

Periglacial Geomorphology is an important resource for undergraduate and graduate students studying geomorphology or Quaternary science within the context of geography and geology degree programs. It will be of use to all scientists whose research involves an understanding of cold environments, whether from a geographical, geological, ecological, climatological, pedological, hydrological, or engineering perspective.

Leseprobe

Table of Contents

Cover

Title Page

Preface

Acknowledgements

1 Introduction

1.1 The Periglacial Concept: Definitions and Scope

1.2 The Periglacial Realm

1.3 The Development of Periglacial Geomorphology

1.4 Periglacial Geomorphology: The Quaternary Context

1.5 The Aims and Organization of this Book

2 Periglacial Environments

2.1 Introduction

2.2 Periglacial Climates

2.3 Soils in Periglacial Environments

2.4 Vegetation Cover in Periglacial Environments

2.5 Synthesis

3 Ground Freezing and Thawing

3.1 Introduction

3.2 Ground Heating and Cooling

3.3 Soil Freezing

3.4 Ice Segregation in Freezing Soils

3.5 Thaw Consolidation

3.6 Synthesis

4 Permafrost

4.1 Introduction

4.2 Permafrost Thermal Regime

4.3 Classification of Permafrost

4.4 Detection, Mapping and Modelling of Permafrost

4.5 Permafrost Distribution

4.6 Permafrost–glacier Interactions

4.7 The Geomorphic Importance of Permafrost

5 Ground Ice and Cryostratigraphy

5.1 Introduction

5.2 Genetic Classification of Ground Ice

5.3 Description of Ground Ice

5.4 Ice Contacts

5.5 Cryostratigraphy

5.6 The Transition Zone

5.7 Massive Ground Ice

5.8 Yedoma

6 Thermal Contraction Cracking: Ice Wedges and Related Landforms

6.1 Introduction

6.2 Thermal Contraction Cracking and Polygon Evolution

6.3 Ice Veins and Ice Wedges

6.4 Ice‐wedge Polygons

6.5 Sand Veins and Sand Wedges

6.6 Composite Veins and Composite Wedges

6.7 Sand‐wedge Polygons

6.8 Frost Cracking of Seasonally Frozen Ground

6.9 Thaw Modification of Frost Wedges

6.10 Frost‐Wedge Pseudomorphs and Frost Polygons in Areas of Past Permafrost

7 Pingos, Palsas and other Frost Mounds

7.1 Introduction

7.2 Characteristics of Pingos

7.3 Hydrostatic Pingos

7.4 Hydraulic Pingos

7.5 Pingo Problems and Problem Pingos

7.6 Segregation Ice Mounds: Palsas, Lithalsas and Related Landforms

7.7 Palsas

7.8 Peat Plateaus

7.9 Lithalsas

7.10 Permafrost Plateaus

7.11 Other Permafrost Mounds

7.12 Ephemeral Frost Mounds

7.13 Relict Permafrost Mounds

8 Thermokarst

8.1 Introduction

8.2 Thermokarst Lakes and Drained Lake Basins

8.3 Thermokarst Pits, Bogs and Fens

8.4 Retrogressive Thaw Slumps

8.5 Small‐scale Thermokarst Features: Beaded Streams, Sinkholes and Thermokarst Gullies

8.6 Sediment Structures associated with Thermokarst

8.7 Relict Thermokarst Phenomena

9 Seasonally Frozen Ground Phenomena

9.1 Introduction

9.2 Upfreezing of Clasts

9.3 Frost Heave of Bedrock

9.4 Patterned Ground: The Embroidery on the Landscape

9.5 Patterned Ground Processes

9.6 Sorted Patterned Ground

9.7 Nonsorted Patterned Ground

9.8 Cryoturbations

9.9 Pedogenic Effects of Freezing and Thawing

9.10 Fragipans

9.11 Synthesis

10 Rock Weathering and Associated Landforms

10.1 Introduction

10.2 Physical Weathering Processes

10.3 Chemical Weathering Processes

10.4 Biotic Weathering Processes

10.5 Weathering Processes in Periglacial Environments

10.6 Cold‐climate Karst

10.7 Tors

10.8 Blockfields and Related Periglacial Regolith Covers

10.9 Brecciated Bedrocks

11 Periglacial Mass Movement and Hillslope Evolution

11.1 Introduction

11.2 Solifluction Processes

11.3 Solifluction Landforms

11.4 Pleistocene Solifluction Landforms and Slope Deposits

11.5 Active‐layer Failures

11.6 Permafrost Creep

11.7 Nivation

11.8 Cryoplanation

11.9 Slope Form and Slope Evolution

12 Talus Slopes and Related Landforms

12.1 Introduction

12.2 Rockfall Talus

12.3 The Geomorphic Role of Snow Avalanches

12.4 Debris‐flow Activity

12.5 Rock Glaciers

12.6 Pronival (Protalus) Ramparts

12.7 Synthesis

13 Fluvial Processes and Landforms

13.1 Introduction

13.2 Periglacial Hydrology

13.3 Slopewash

13.4 Slushflows

13.5 Sediment Transport in Periglacial Rivers

13.6 Bank and Channel Erosion

13.7 River Channels

13.8 Alluvial Landforms in Periglacial Environments

13.9 Valley Form

13.10 Pleistocene Periglacial Rivers

13.11 Synthesis

14 Wind Action

14.1 Introduction

14.2 Aeolian Processes

14.3 Wind Erosion in Present Periglacial Environments

14.4 Aeolian Deposits in Present Periglacial Environments

14.5 Quaternary Aeolian Deposits

14.6 Synthesis

15 Periglacial Coasts

15.1 Introduction

15.2 The Nature of Periglacial Coasts

15.3 The Role of Ice in Shoreline Evolution

15.4 Ice‐rich Permafrost Coasts

15.5 Thermokarst Coasts

15.6 Barrier Coasts

15.7 Salt Marshes and Tidal Flats

15.8 Rock Coasts

15.9 Raised and Inherited Shorelines

15.10 Lake Shorelines

15.11 Synthesis

16 Past Periglacial Environments

16.1 Introduction

16.2 Palaeoenvironmental Reconstruction Based on Periglacial Features

16.3 Past Periglacial Environments of the British Isles

16.4 Pre‐Late Devensian Periglacial Features in the British Isles

16.5 The Dimlington Stade in the British Isles

16.6 The Younger Dryas (Loch Lomond) Stade in the British Isles

16.7 Past Periglacial Environments of the British Isles: Commentary

16.8 Late Weichselian Periglacial Environments in Continental Europe

16.9 Late Wisconsinan Periglacial Environments in North America

16.10 Permafrost Extent in the Northern Hemisphere During the Last Glacial Stage

16.11 Concluding Comments

17 Climate Change and Periglacial Environments

17.1 Introduction

17.2 Permafrost Degradation

17.3 Geomorphological Implications of Climate Change in the Circumpolar North

17.4 Geomorphological Implications of Climate Change in High Mountain Environments

17.5 Climate Change, Permafrost Degradation and Greenhouse Gas Emissions

17.6 Conclusion

Appendix Text Abbreviations, Units and Symbols Employed in Equations

Text Abbreviations

Units

Symbols

References

Index

End User License Agreement

List of Tables

Chapter 03

Table 3.1 Thermal properties of soils and constituent materials.

Chapter 05

Table 5.1 Cryostructures in frozen ground.

Table 5.2 Cryofacies of frozen sediments. A diamicton is a poorly sorted sediment body, usually involving clasts embedded in finer sediment. Cryostructure coding follows that in column one of Table 5.1.

Chapter 06

Table 6.1 Temperature conditions leading to thermal contraction cracking.

Table 6.2 Typical characteristics of cryogenic wedge pseudomorphs.

Chapter 08

Table 8.1 Factors responsible for initiating or retarding thermokarst.

Table 8.2 Sedimentary facies of deep thermokarst lake basins, Tuktoyaktuk Coastlands.

Chapter 11

Table 11.1 Measured rates of solifluction.

Table 11.2 Mean morphometric characteristics of solifluction lobes.

Table 11.3 Measured rates of ploughing boulder movement.

Table 11.4 Rates of mass transfer (10

3

 t m km

−2

 a

−1

) by active‐layer failures in the Mackenzie Valley (65° N), northwest Canada, and on the Fosheim Peninsula (80° N), Ellesmere Island.

Chapter 12

Table 12.1 Measured rockwall retreat rates in periglacial environments.

Chapter 13

Table 13.1 Annual water balance of McMaster River Basin, arctic Canada, 1975–81. Water balance calculated from September to September.

Table 13.2 Specific sediment yields and implied bedrock denudation rates for some periglacial rivers in arctic North America.

Table 13.3 Specific sediment yields and implied bedrock denudation rates for some low‐arctic and subarctic catchments.

Table 13.4 Estimated total annual sediment yields for some periglacial catchments.

Chapter 14

Table 14.1 Arctic and subarctic dunefields and sand sheets, Scandinavia and North America.

Table 14.2 Lithofacies characteristics of riparian aeolian deposits, west Greenland. Facies are ordered downwind from the active floodplain.

Table 14.3 Facies types and characteristics of Pleistocene aeolian sands in Northern Europe.

Chapter 15

Table 15.1 Estimates of sediment concentration in sea ice.

Table 15.2 Sediment input into the Arctic Ocean through coastal erosion.

Chapter 16

Table 16.1 Palaeonvironmental implications of thermal contraction crack features, relict ground ice landforms and structures, and patterned ground. Evidence for continuous permafrost (CFP) usually implies MAAT less than −5 °C to −8 °C. Evidence for discontinuous or sporadic permafrost (DPF) usually implies MAAT in the range −1 °C to −6 °C.

Table 16.2 Palaeonvironmental implications of landforms, deposits and structures formed by periglacial weathering and mass movement.

Table 16.3 Palaeonvironmental implications of periglacial alluvial, littoral and aeolian landforms and deposits.

Table 16.4 Area estimates (10

3

 km

2

) of the northern circumpolar permafrost region during the last glacial maximum (LGM). Estimates include zones underlain by discontinuous and sporadic permafrost.

List of Illustrations

Chapter 01

Figure 1.1 Periglacial landscape of the North Slope of Alaska (69° N), looking towards the Brooks Range. Though this area is underlain by permafrost and subject to severe winter freezing, the broad outlines of the landscape differ little from those in other environments.

Figure 1.2 Wedge‐shaped structure in sand and gravel deposits near Lincoln, eastern England. This structure represents infill of the void left by thaw of an ice wedge that formed in permafrost during the last glacial period.

Figure 1.3 Landscape of northern Yukon, Canada, an area that escaped glaciation during the Pleistocene epoch and has evolved under periglacial conditions since the beginning of the Quaternary period.

Figure 1.4 Subsidence of buildings due to thaw of underlying permafrost in (a) Alaska and (b) Yakutsk.

Figure 1.5 Relationship between periglacial geomorphology and cognate sciences.

Figure 1.6 Marine oxygen isotope stage (MIS) record of glacial–interglacial oscillations during the past 700 ka. Changes in the δ

18

O:δ

16

O ratio in benthic foraminifera in deep ocean cores record the expansion and contraction of global ice sheets, and thus changes in global temperatures. Dashed horizontal lines mark the main boundaries between glacial stages (even numbers) and interglacial stages (odd numbers).

Chapter 02

Figure 2.1 Duration of continuous daylight and continuous darkness at latitudes 60–90° N, defined by whether the sun is above or below the horizon.

Figure 2.2 Isohyets of mean annual precipitation (MAP) in the Arctic. Isohyet intervals are 100 mm for precipitation <600 mm and 200 mm for precipitation >600 mm. Extensive areas of the high Arctic and arctic or subarctic continental interiors have MAP <400 mm.

Figure 2.3 Mean monthly temperature and precipitation at high‐arctic, low‐arctic and continental interior locations near sea level. Horizontal dashes represent mean monthly temperature and vertical lines the range between maximum and minimum mean monthly temperatures. The histograms represent mean monthly precipitation. The figures in the centre of each graph are the mean annual air temperature (MAAT) and mean annual precipitation (MAP).

Figure 2.4 Mean monthly temperature and precipitation at Antarctic and sub‐Antarctic locations near sea level. Horizontal dashes represent mean monthly temperature, vertical lines the range between maximum and minimum mean monthly temperatures and histograms the mean monthly precipitation.

Figure 2.5 Mean monthly temperature and precipitation on the Niwot Ridge, Colorado, USA, a mid‐continental location with a large annual temperature range, and in Fanåraken, Norway. Horizontal dashes represent mean monthly temperature, vertical lines represent the range between maximum and minimum mean monthly temperatures and histograms represent the mean monthly precipitation. Note that the precipitation scale is compressed compared with Figures 2.3 and 2.4.

Figure 2.6 Vegetation cover in periglacial environments. (a) Polar desert, Svalbard. (b) Polar desert, Ellesmere Island, Canadian Arctic Archipelago. (c) Tundra, Tuktoyaktuk Peninsula. (d) Tundra, Alaska. (e) Wetlands in taiga forest near Churchill, Manitoba. (f) Alpine vegetation: coniferous forest in the valley, alpine grasslands and bare ground above ~2500 m, Savoie, French Alps.

Chapter 03

Figure 3.1 Changes in the properties of a moist silty‐clay soil and a moist sandy loam within the temperature range +1 to −4 °C.

Figure 3.2 Relationship between geothermal gradient (d

T

/d

z

) and depth of permafrost (

z

p

). The upper line is associated with ground exhibiting relatively low thermal conductivity, the lower line with ground exhibiting relatively high thermal conductivity.

Figure 3.3 Annual temperature variations in unfrozen soil, illustrating diminishing amplitude and increased phase lag with depth (Z).

Figure 3.4 (a) Decline in annual temperature range with depth. For an explanation of terms, see text. Note that this diagram ignores the effect of geothermal gradient. (b) Definition of phase lag (

t*

), which is the difference in timing between peak ground surface temperature and peak temperature at depth

z

. For clarity,

A

s

and

A

z

are shown as semi‐amplitudes (0.5 × amplitude).

Figure 3.5 Cooling curve for soil water and ice. Initial supercooling of unfrozen water (to point S) is interrupted by ice nucleation. The resultant release of latent heat raises the temperature to the equilibrium freezing point (E), which may be depressed slightly below 0 °C by the presence of dissolved salts. The remaining ‘free’ or unbound water freezes at the equilibrium freezing point. Thereafter, ‘bound’ water held in capillary attraction or adsorbed on to mineral surfaces freezes at progressively lower temperatures.

Figure 3.6 Diagrammatic representation of the free energy of ice, pure water and soil water plotted against temperature. Pure water and ice have equal free energies at 0 °C, but the free energy of soil water is reduced by the presence of dissolved salts and other effects, resulting in ‘freezing point depression’ (Δ

T

) below 0 °C. In reality, there is no single freezing point of soil water, as diminishing quantities of soil water coexist with ice as the temperature is lowered below 0 °C.

Figure 3.7 Phase diagram for water. The solid lines separating different phases (ice, liquid water and water vapour) are ‘coexistence lines’ where Δ

G

 = 0; in other words, where the Gibbs energy of ice equals that of liquid water, or where that of liquid water equals that of water vapour. The two coexistence lines meet at the ‘triple point’ of water where

P

 = 6.03 × 10

−3

atmospheres (0.603 kPa) and

T

 = 0.0098 °C. The reverse slope of the ice–water coexistence line implies that an increase in pressure causes freezing point depression.

Figure 3.8 Segregated ice lenses in lacustrine silty clay, western Canadian Arctic. The lenses near the top of the permafrost are thickest, and become progressively thinner with increasing depth. The section is about 2 m high.

Figure 3.9 Differentiation of frost‐susceptible soil (in which ice lenses form during freezing) and non‐frost‐susceptible soil on the basis of particle size distribution, as proposed by Beskow (1935).

Figure 3.10 Linear relationships between cryosuction and temperature predicted by the Clapeyron equation for Δ

P

i

 = 0, Δ

P

i

 = Δ

P

w

and Δ

P

i

 = 5Δ

P

w

.

Figure 3.11 (a) Microscopic view of soil particles at the ice–soil interface. The convex‐outward radius of the ice between soil particles at the base of the ice lens exceeds the pore radius, so ice cannot penetrate the pore spaces under the ice lens (capillary model). (b) Microscopic view of ice extending down from an ice lens into a frozen fringe.

Figure 3.12 (a) Ice distribution in the frozen fringe beneath a growing ice lens. (b) Predicted load supported by particle contacts. Above the base of the frozen fringe, the load supported by particle contacts decreases to a minimum because of the large gradients in liquid pressure needed to drive flow through the lower permeabilities caused by increasing upwards ice saturation. At the depth where the minimum load tends towards zero, a new ice lens will form.

Figure 3.13 Regime diagram displaying the behaviour of a freezing soil as a function of freezing rate and overburden pressure. Dashed lines describe the limits of hysteretic behaviour (see text).

Figure 3.14 Ice lens formation through lateral crack extension. (a)–(c) Crack extension and lens formation from an initial ice‐filled flaw (void) in freezing soil. (d) Crack extension from an initial flaw located in the wall of an ice‐filled shrinkage crack.

Chapter 04

Figure 4.1 Ice wedge in permafrost, Svalbard. The soil above the ice wedge represents the active layer, which freezes and thaws annually.

Figure 4.2 Seasonal patterns of ground freezing and thawing: (a) above permafrost; (b) seasonally frozen ground in non‐permafrost terrain. In (a), two‐sided freezing above ‘cold’ permafrost is depicted; above ‘warm’ permafrost, freezing is mainly from the surface downwards only.

Figure 4.3 Three‐layer model of the active layer, transition zone and permafrost. The curve represents the relative probability of annual thaw depth (a) directly following very deep thaw at subdecadal to centennial timescales and (b) with ice enrichment of the transition zone several centuries after the deep thaw event. Ice enrichment of the transition zone in (b) involves formation of ice lenses and upwards growth (rejuvenation) of secondary and tertiary ice wedges.

Figure 4.4 Schematic illustration of taliks. Open taliks penetrate the permafrost body completely, whereas closed taliks occupy a depression in the permafrost table, usually below a lake or river; their temperature remains above 0 °C because of heat storage in the surface water body. Isolated taliks are surrounded by frozen ground, and take the form of cryopegs or transient noncryotic taliks. Lateral taliks are overlain and underlain by permafrost. Hydrothermal and hydrochemical taliks are those that are fed by groundwater; the former are noncryotic, and resist freezing because of the heat supplied by flowing groundwater, whereas the latter are cryotic, but freezing is prevented by groundwater salinity.

Figure 4.5 Simplified thermal regime of permafrost, showing temperature change with depth. (a) Full depth of permafrost, showing how the permafrost base is defined by the point where the geothermal gradient crosses 0 °C. (b) Expansion of the top part of (a) to show the amplitude of annual temperature change in the active layer and underlying permafrost, the depth of zero annual amplitude and a definition of the active layer as the point where

T

max

crosses 0 °C. (c) Idealized profile of the phase lag in temperature at 2‐monthly time intervals, illustrating how temperature at depth lags behind temperature change at the surface. MAGT, mean annual ground temperature;

T

min

and

T

max

, minimum and maximum annual temperature; AL, active layer; PFT, permafrost table;

T

TOP

, mean annual temperature at the top of the permafrost.

Figure 4.6 Schematic mean annual temperature profile through the active layer and overlying surface boundary layer, showing the relation between air temperature and permafrost temperature. MAAT, mean annual air temperature; MAGST, mean annual ground surface temperature;

T

TOP

, temperature at the top of the permafrost.

Figure 4.7 Relationship between the

n

f

factor, snow depth and mean annual air temperature (MAAT).

Figure 4.8 Circum‐arctic map of permafrost.

Figure 4.9 Schematic north–south transect of permafrost and active‐layer thickness in arctic and subarctic Canada, based on data for Hay River (61° N), Norman Wells (64° N) and Resolute (74° N). Active layer (stippled) is not to scale.

z

a

, active‐layer thickness;

z

p

, permafrost thickness.

Figure 4.10 Modelled two‐dimensional steady‐state isotherms of ground temperature distribution in an idealized alpine summit ridge based on borehole temperature data measured on the summit of Stockhorn, Switzerland. Permafrost is absent from the south face but underlies the north face.

Figure 4.11 Monthly temperature regime measured near the surface and at depth on Plateau Mountain, Alberta, Canada. The blockslope temperatures are not only much lower than those for fine‐grained loessic soil, but also display much greater annual amplitude, especially at depth.

Figure 4.12 Models for undercooling of openwork bouldery debris. (a) Free exchange of air with atmosphere during winter under snow‐free conditions. (b) Downslope advection of cold air and compensatory upslope advection of warm air under snowcover. (c) The ‘chimney effect’: upslope advection of relatively warm air draws in cold air through overlying snowcover. (d) Cooling by conductive heat transfer through boulders protruding through shallow snowcover.

Figure 4.13 Cave ice in a limestone cave, Picos de Europa, northern Spain.

Figure 4.14 (a) Velocity profile of a cold‐based glacier moving over a rigid (bedrock) bed; movement occurs through internal deformation of the ice. (b) Velocity profile of the same glacier over ice‐rich permafrozen sediments; the velocity profile extends into the substrate, causing deformation of the upper part of the permafrost.

Figure 4.15 Multi‐crested push moraines at Bergmesterbreen, a polythermal glacier on Svalbard. The rigidity of the frozen sediments has caused stacking of successive rafts of frozen sediment at the glacier margin during periods of ice‐margin advance.

Chapter 05

Figure 5.1 Ground ice: ice wedges at the top of the photograph penetrate downwards into ice‐rich frozen silts that overlie a body of stratified massive ice.

Figure 5.2 Genetic classification of ground ice.

Figure 5.3 Classification of types of pore ice (‘ice cement’) in frozen soils. Ice separates grain‐to‐grain contacts in ‘basal’ ice, but not in ‘pore’, ‘film’ or ‘contact’ ice.

Figure 5.4 Needle ice crystals, 15 cm long. Segmentation within the crystals indicates more than one period of growth. At the top of the crystals is soil that has been pushed up by crystal growth. The soil at the bottom of the crystals represents the ground in which crystal growth was rooted.

Figure 5.5 Schematic illustration of cryostructures, following the scheme in Table 5.1. Ice is shown in black, and the grey objects in the crustal cryostructure are frost‐susceptible clasts.

Figure 5.6 Cryostructures in cores extracted from ice‐rich permafrost: (a) lenticular; (b) banded; (c) reticulate; (d) ataxitic.

Figure 5.7 Thaw unconformity truncating banded massive ice, Crumbling Point, NWT, Canada. The zone above the uncorformity represents the palaeo‐active layer.

Figure 5.8 Cryostratigraphic relationships within a stable substrate. C2 is a palaeo‐thaw layer separated from C1 by a thaw unconformity representing the base of a former, deeper active layer. The truncated ice wedge (1) within C1 is younger than the host sediments in C1 but older than the ice in C2. The truncated massive ice body (2) is also younger than the host sediments in C2 if it represents intrasedimental ice, but may be roughly coeval with these sediments if it is buried glacier ice. The upper active ice wedge (3) and intrusive sill ice (4) are younger than the host sediment.

Figure 5.9 Cryostratigraphic record of the wall of part of the CRREL permafrost tunnel: undisturbed syngenetic permafrost is to the left of the erosional contact, and frozen gully infill is right of the contact.

Figure 5.10 Conceptual model of landscape development from the Late Pleistocene to the present in west‐central Alaska, based on the cryostratigraphy of frozen sediments and peat plateaus.

Figure 5.11 Development of the transition zone during epigenetic and syngenetic permafrost aggradation. The latter depicts pulsed sediment accumulation, with concomitant rises in the average position of the permafrost table, producing several transient ice‐rich transition zones, as in the accumulation of yedoma deposits.

Figure 5.12 Massive stratified ground ice, Hooper Island, NWT, Canada. The upper parts of the ice body have been deformed as a result of overrunning by glacier ice.

Figure 5.13 Proposed explanations for the growth of massive intrasedimental ice in glaciated lowlands. (a) An ice sheet advances across permafrost, creating a subglacial thawed zone up‐ice from the ice‐sheet margin. (b) Proglacial downwards permafrost aggradation follows ice margin retreat, and subglacial meltwater under hydraulic pressure moves through a confined aquifer to feed growth of massive intrasedimental ice. (c) Ice‐sheet thinning causes permafrost aggradation under cold‐based ice, and subglacial meltwater under hydraulic pressure moves through a confined aquifer to feed the growth of subglacial massive intrasedimental ice.

Figure 5.14 Distribution of yedoma deposits in Siberia and Alaska, showing how they occur only outside the limits of the last (MIS 2) Pleistocene ice sheets.

Figure 5.15 Generalized cross‐sections of yedoma exposures. Top: Kurungnakh Sise Island, Lena River Delta. Bottom: Oyogos Yar coast, East Siberian Sea.

Figure 5.16 Exposures in yedoma sediments. (a) Syngenetic ice wedges in the yedoma deposits at Duvanny Yar, northeast Siberia. (b) Yedoma deposits on the Itkillik River, northern Alaska. Syngenetic ice wedges separate stratified silts.

Figure 5.17 Cryostratigraphic units of the Itkillik yedoma exposure (ice wedge widths not to scale) and radiocarbon ages (cal

14

C ka) obtained on organic material within the yedoma deposits. Wavy horizontal and dashed lines represent ice lenses or microlenses, which tend to be concentrated in bands within the silts. Shaded areas represent ice wedges.

Chapter 06

Figure 6.1 High‐centre frost polygons and thermokarst lakes, central Banks Island, Canadian Arctic Archipelago. Individual polygons form subsets of a larger irregular polygonal net.

Figure 6.2 (a) High‐centre polygons defined by water‐filled troughs overlying ice wedges. (b) Flooded low‐centre polygons bordered by parallel ridges of sediment pushed up by expansion of underlying ice wedges.

Figure 6.3 Development of ice wedges and frost polygons due to repeated thermal contraction cracking of the ground and infill of cracks by vein ice. (a) Initial cracking (orthogonal), first winter. (b) Formation of ice veins in initial cracks in permafrost. (c) Later cracking of established wedges. Ridges pushed up by volume increase define the margins of low‐centred polygons. (d) Incorporation of a new ice vein in the wedge. Wedges progressively widen through time as new ice veins are added.

Figure 6.4 (a) Two ice wedges of different widths in the western Canadian Arctic. The wedges are ~2 m long. (b) Conjoint wedges in coarse deltaic deposits, Svalbard. Coalescence of these wedges suggests that they occur near a frost polygon junction. (c) Single ice wedge at Crumbling Point, western Canadian Arctic. (d) Top of an ice wedge exposed in a permafrost tunnel, showing the subvertical laminae that represent progressive addition of ice veins during wedge widening.

Figure 6.5 Growth of epigenetic, syngenetic and anti‐syngenetic ice wedges. Epigenetic ice wedges develop in stable sediment bodies, syngenetic wedges in aggrading sediments and anti‐syngenetic wedges under conditions of surface lowering. 1, 2 and 3 represent successive levels of the ground surface.

Figure 6.6 Rejuvenated ice wedge, western Canadian Arctic. The top of the wedge has been truncated by active‐layer deepening, but renewed cracking and ice‐vein formation following later active‐layer thinning have allowed the central part of the wedge to form at a higher level, forming the raised finger of ice near the centre of the wedge.

Figure 6.7 Topographic classification of ice‐wedge polygons.

Figure 6.8 Large epigenetic composite wedge with a predominantly sand infill, Crumbling Point, western Canadian Arctic. Bundles of sand veins can be distinguished within the wedge, which extends downward through a thaw unconformity that truncates massive ice and ice‐rich sediments.

Figure 6.9 Vertical section across a single sand‐wedge polygon, showing hypothetical long‐term circulation of sediment. Aggradation of sediment between the wedges is offset by loss of surface sediment into thermal contraction cracks, so there is long‐term circulation of sediment associated with lateral wedge expansion.

Figure 6.10 (a) Late Wisconsinan ice‐wedge pseudomorph in sandy sediments, southwest Ontario, Canada. Note faulting of host sediments along the margin of the wedge. (b) Ice‐wedge pseudomorph in coarse gravels, Lincolnshire, eastern England. The infill of clay and sand is derived from a sediment layer that has subsequently been removed by erosion.

Figure 6.11 (a) Relict sand‐wedge polygon 2.5 m wide exposed by excavation at Newney Green near Chelmsford, England. (b) Small sand wedge in Middle Pleistocene sand and gravel deposits at Broomfield, near Chelmsford, England. Both features are overlain by glacigenic deposits of MIS 12 age, and therefore over 430 000 years old, but the wedge still preserves downwards‐tapering laminae representing individual sand veins.

Chapter 07

Figure 7.1 Hydrostatic pingos, Tuktoyaktuk Peninsula, western Canadian Arctic.

Figure 7.2 Fløtspingo, Reindalen, Svalbard, a hydraulic pingo formed where groundwater flowing under pressure has fed the growth of an ice core within permafrost, updoming the overlying frozen sediments.

Figure 7.3 Schematic evolution of a hydrostatic pingo. (a) Development of a suprapermafrost talik of unfrozen saturated sandy sediments beneath a lake. (b) Lake drainage results in downwards and lateral permafrost aggradation into the talik; expulsion of pore water from the aggrading permafrost results in the development of high hydrostatic pressures under a residual shallow pond where permafrost is thinnest. (c) Pore water expulsion into the residual talik leads to the formation of a sub‐pingo water lens; progressive downwards freezing results in formation of a thickening ice core and updoming of the overlying frozen sediments, forming a hydrostatic pingo. (d) Exposure of the ice core at the pingo summit causes melt of the core and subsidence of the central part of the pingo.

Figure 7.4 (a,b) Contoured maps of a growing hydrostatic pingo surveyed in 1973 and 1993, showing a height increase of about 4 m in 20 years, but without significant change in diameter. (c) Cross‐section of the same pingo based on surveys in 1973 and 1996 and drilling in 1977. BM, benchmark.

Figure 7.5 Collapsing pingo on the Parry Peninsula, NWT, Canada.

Figure 7.6 Surface icing (upslope from snowmobiles), formed where a spring issuing from an open‐system pingo has frozen on reaching the surface, Adventdalen, Svalbard.

Figure 7.7 Schematic evolution of a hydraulic pingo. (a) Groundwater moving under hydraulic pressure through a subpermafrost aquifer feeds growth of a core of injection ice under shallow permafrost. (b) Growth of the ice core results in updoming and cracking of the overlying frozen ground. (c) Exposure of the ice core at the surface causes melt of the core and subsidence of the pingo.

Figure 7.8 Suggested evolution of aligned ‘summit pingos’ on Prince Patrick Island. As permafrost aggrades downwards, expelled pore water accumulating in subpermafrost taliks is forced upwards and freezes at shallow depth, forming pingos.

Figure 7.9 Palsa in Finnish Lapland.

Figure 7.10 A complex of palsas and lithalsas growing amid thermokarst ponds in northern Quebec. The rounded mounds in the centre of the photograph lack peat cover (lithalsas), but nearby mounds are peat‐covered (palsas). The circular thermokarst ponds are the sites of collapsed palsas and lithalsas.

Figure 7.11 Evolution of a palsa through buoyant uplift and frost heave. (a) Deeper winter frost penetration in peat under an area of thinner snowcover. (b) Buoyant uplift of frozen peat during the first summer, leading to development of an underlying water lens. (c) Freezing of the water lens during the next winter, forming a lens of aggradation ice. (d) Renewed buoyant uplift in the following summer, with formation of a second water lens. (e) Freezing of second water lens during the next winter, adding a further layer of aggradation ice. (f) After several years the freezing plane penetrates underlying silts, leading to ice‐lens formation and frost heave.

Figure 7.12 Schematic evolution of a lithalsa. (a) Proto‐lithalsa, formed by permafrost aggradation and frost heave under an initial peat cover. (b) Mature lithalsa, with permafrost extending several metres into the ground and peat cover removed by erosion. (c) Permafrost degradation, initiation of lithalsa collapse and surface ponding. (d) Collapsed (relict) lithalsa, comprising ramparts of sediment encircling a shallow thermokarst pond.

Figure 7.13 Cryostratigraphy of a permafrost plateau developed on silty clay marine sediments on the shore of Hudson Bay. Layer 1 is the active layer; layers 2–4 represent permafrost. Layer 2 contains mainly aggradational ice (50–80% by volume). Layer 3 contains reticulate ice (10–30% by volume). Layer 4 consists of alternating frozen sediment and ice lenses (50–80% by volume), which increase in thickness near the base of the permafrost.

Figure 7.14 Development of a frost blister. (a) Groundwater flow through the active layer during freezeback. (b) Damming of flow as freezing extends to the permafrost table on the valley floor, causing development of high hydraulic pressure and updoming of the overlying frozen ground. (c) Inwards freezing of the water lens, producing further uplift, and escape of water, creating a surface icing. (d) Melt of the ice core during summer, resulting in collapse.

Figure 7.15 Icing blister with summit tension crack, Svalbard.

Chapter 08

Figure 8.1 (a) Thermokarst lakes and thermokarst ponds, central Banks Island, arctic Canada. The outlines of submerged frost polygons can be seen on the floor of the central lake, suggesting that it formed recently. (b) Thermokarst lakes flanking a river channel, Mackenzie Delta area, western Canadian Arctic.

Figure 8.2 Schematic evolution of thermokarst lakes and drained lake basins (a) in syngenetic permafrost, where excess ice is tens of metres deep and (b) in epigenetic permafrost, where excess ice extends only a few metres below the surface.

Figure 8.3 Small thermokarst lake, central Banks Island, arctic Canada. Degradation of ice‐wedge polygons near the lake margins has formed thermokarst mounds. Fens have colonized the lake margins.

Figure 8.4 Simulated talik development (light grey) and lake deepening (dark grey), based on a nonexpanding lake with an initial depth 2 m, 30% excess ice content in the underlying sediments and an assumed lake‐bottom temperature of 3 °C. The model on the left represents an area of deep ground ice (syngenetic permafrost), and that on the right an area of shallow ground ice (epigenetic permafrost). Isotherms are °C. Both models exhibit near‐equilibrium conditions after 8000 years.

Figure 8.5 Schematic representation of the processes leading to alternation of fine sand and silty gyttja in Lake El’gene‐Kyuele, northeast Siberia. Top: Climate‐driven increased (a‐1) or decreased (a‐2) thermoerosion. Bottom: slope destabilization (b‐1) and stabilization (b‐2) phases driven by recurrent exposure of ice wedges. Both scenarios potentially lead to alternating input of fine sand (a‐1 and b‐1) and silty gyttja (a‐2 and b‐2).

Figure 8.6 Pingo in a partially drained alas lake basin, central Yakutia.

Figure 8.7 Conceptual model for the evolution of lakes and drained basins on the coastal plain of northern Alaska. (a) Flooding of surface depressions. (b) Lake enlargement and talik formation. (c) Lake drainage, leaving silty sediments in the middle of the basin and sandy sediments on marginal shelves. (d) Preferential ground ice formation, elevating the silty sediments. (e) Development of a thermokarst lake on the ice‐rich silty sediments.

Figure 8.8 Number of water bodies >0.1 ha in area on the northern Seward Peninsula, Alaska, in four size classes (detected on remotely‐sensed images) for 1950–51, 1978 and 2006–07. The increase in smaller water bodies is largely attributable to partial drainage of larger lakes, leaving residual small lakes and ponds.

Figure 8.9 Thermokarst bog, Alaska. Toppling of trees at the bog margins is due to permafrost degradation.

Figure 8.10 Schematic cross‐section depicting thermokarst pits, bogs and fens associated with boreal birch forest in the discontinuous permafrost zone (50× vertical exaggeration).

Figure 8.11 (a) Large retrogressive thaw slump (megaslump) along the Selawik River, northwest Alaska. Over 700 000 m

3

of sediment was evacuated from this slump between its formation in 2004 and its stabilization in 2016. (b) The same slump after stabilization.

Figure 8.12 Schematic section of the morphological components of a retrogressive thaw slump. Progressive retreat of the headscarp results in extension of the slump floor.

Figure 8.13 Maximum and minimum ground temperatures (2006–07) for undisturbed tundra and a stable vegetated slump surface, Richards Island, NWT, Canada.

Figure 8.14 Incipient development of a beaded stream along a watercourse in central Banks Island, arctic Canada. Thaw at ice‐wedge intersections has created small thermokarst ponds linked by flow along the ice‐wedge network.

Figure 8.15 Thermo‐erosional gully, Ellesmere Island, Canadian Arctic Archipelago, which formed within 2 days following overspill of a small glacier‐dammed pond.

Figure 8.16 Thermokarst involutions within a refrozen palaeo‐active layer, Summer Island, western arctic Canada. (a) Pseudo‐nodules of sand within diamicton. (b) Load casts of sand between rounded diapirs of diamicton. (c) Ball‐and‐pillow structures of sand within diamicton. Load casts of sand and diapirs of diamicton occur towards the top of the image.

Figure 8.17 Thermokarst involutions of last glacial age developed under a stripe pattern, southeast England. The underlying substrate is brecciated chalk.

Chapter 09

Figure 9.1

Top:

The frost‐push mechanism of clast heave. (a) During freezing, ice forms under a clast, pushing it up. (b) Ice‐filled sub‐clast void when the soil is completely frozen. (c) During thaw, soil partly infills the void, preventing the clast from regaining its original position.

Bottom:

The frost‐pull mechanism of clast heave. (a) During freezing, a clast freezes to surrounding soil; further frost penetration heaves the upper frozen soil and the attached clast. (b) Air‐filled void under uplifted clast when the soil is completely frozen. (c) During thaw, soil partly infills the void, preventing the clast from regaining its original position, so there is net upwards movement of the clast over a single freeze–thaw cycle.

Figure 9.2 Patterned ground. (a) Active sorted net, northeast Iceland. (b) Active sorted circles (debris islands), arctic Norway. (c) Stone pits, Kalfafjell, Iceland. (d) Sorted stripes, Tinto Hill, Scotland. (e) Earth hummocks (thúfur), Holar Valley, Northern Iceland. (f) Vegetation‐defined nonsorted circles (frost boils) 1–2 m in diameter, Howe Island, Alaska.

Figure 9.3 Sorted net on a seasonally flooded lake margin, Abisko, northern Sweden.

Figure 9.4 Erected clasts at the margin of a large sorted circle, Mount Bitterness, New Zealand. Clast erection results from lateral pressures exerted on clasts due to expansion of adjacent fine soil during winter freezing.

Figure 9.5 Large annular sorted circles, Vardeborgsletta, Svalbard.

Figure 9.6 Lateral sorting of clasts and fines initiated by inclined freezing planes caused by more rapid freezing under open cracks. Clasts tend to be heaved upwards and towards the zones of more rapid frost penetration, whereas fine soil is driven towards the unfrozen zone under the areas of slower frost penetration.

Figure 9.7 Patterned ground initiation by differential frost heave as proposed by Kessler

et al

. (2001). During freezing of the active layer, frost heave drives material away from the freezing front (dashed line) at a small positive perturbation between a layer of soil and an overlying layer of stones. Soil movement is both outwards towards the ground surface and inwards towards unfrozen soil. As soil displacements are not reversed during thaw, this mechanism promotes upwards growth of a plug of soil over recurrent freeze–thaw cycles.

Figure 9.8 Simulation of the evolution of sorted circles from a layer of stones (light grey) overlying fine soil (dark grey) over 2000 freeze–thaw cycles.

Figure 9.9 Velocity vectors within a sorted circle modelled by Kessler

et al

. (2001). Soil movement is driven by differential frost heave.

Figure 9.10 Miniature sorted stripes on basalt regolith, Faroe Islands.

Figure 9.11 Sorted pattern formation by differential needle‐ice growth. (a) Clasts embedded in frost‐susceptible soil. (b) Differential needle‐ice growth causing updoming of surface clasts. (c) Thaw and bending of needle ice, with movement of clasts to dome margins. (d) Complete thaw and incipient lateral sorting. (e) Renewed freezing, with needle‐ice growth above, and ice‐lens formation within, cells of fine soil; clast movement towards the margins of updomed soil. (f) Miniature sorted net, with clast borders surrounding predominantly fine cells.

Figure 9.12 (a) Model of sediment displacements within sorted circles proposed by Hallet and Prestrud (1986), reproduced with the permission of Elsevier. (b) Model of free convection of pore water during active layer thawing, with resultant development of corrugations in the permafrost table.

Figure 9.13 Large, relict vegetated sorted circles at 1050 m in the Grampian Highlands, Scotland. Such sorted circles were formed in the active layer above permafrost during the Late Pleistocene, and have survived intact throughout the Holocene.

Figure 9.14 Structure of frost boils in the Hudson Bay lowlands.

Figure 9.15 Physical processes taking place within a nonsorted circle (frost boil) during freezing.

Figure 9.16 (a) Earth hummocks developed over cold permafrost, Ellesmere Island, arctic Canada. (b) Earth hummocks (thúfur) developed in volcanogenic loess, northern Iceland. Permafrost is absent. (c) Earth hummocks at 400 m altitude on Dartmoor, southwest England. (d) Sinuous relief stripes on gentle slopes at 900 m altitude in northwest Scotland.

Figure 9.17 Four models of earth‐hummock formation. (a) Cryostatic pressure: squeezing upwards of a pocket of saturated unfrozen soil by advancing freezing planes. (b) Soil‐circulation model (Mackay, 1980): gradual soil movement down a depression in the permafrost table and buoyancy‐driven rise of mud towards hummock crests during thaw. (c) Permafrost aggradation model (Kokelj

et al

., 2007): ice‐lens growth near the top of aggrading permafrost, pushing up the overlying soil. (d) Differential frost heave (Killingbeck and Ballantyne, 2012): migration of silty soil ahead of inclined freezing planes, causing hummock growth over multiple shallow freeze–thaw cycles. PT, permafrost table.

Figure 9.18 Section excavated through an earth hummock on a slope of 8°, Ellesmere Island, Canadian Arctic Archipelago. Burial of organic soil suggests slow downslope movement of the hummock.

Figure 9.19 Cryoturbations in the active layer, Ellesmere Island, Canadian Arctic Archipelago. The active layer is ~40 cm deep.

Figure 9.20 Typology of cryoturbation structures proposed by Vandenberghe.

Figure 9.21 Three‐dimensional morphology of shallow cryoturbation structures developed in a volcanic ash soil overlying pumice in northeast Japan.

Figure 9.22 Photomicrographs of structures associated with freezing and thawing of frost‐susceptible soil. (a) Lenticular aggregates separated by segregation ice (white). (b) Vesicles (large smooth‐walled soil pores). (c) Silt cappings on the upper surfaces of coarse sand grains.

Figure 9.23 Schematic representation of a fragipan in an imperfectly drained silty soil.

Chapter 10

Figure 10.1 Blockfield composed of angular ‘frost‐shattered’ schist debris at 1900 m on Mount Bitterness, South Island, New Zealand. Appearances can be deceptive: some clasts are edge‐rounded by granular disaggregation, and fine sediment is present just below the surface layer of debris.

Figure 10.2 Four effects of frost action on rock. (a) Frost wedging, the opening of joints due to expansion of joint water on freezing. (b) Frost splitting of a boulder, Hornsund, Svalbard. (c) Scaling (exfoliation) of thin flakes of rock, Svalbard; the ice‐filled polygonal cracks may represent cracking due to thermal stress. (d) Rounding of the surfaces of a granite tor due to granular disaggregation.

Figure 10.3 Measurements of (a) crack widening and (b) temperature at the crack surface at a site in the Japanese Alps, showing the association between freezing events and crack‐widening events during seasonal freezeback.

Figure 10.4 Experimental fracture of a chalk block by two‐sided freezing and ice segregation. The block is 45 cm high.

Figure 10.5 (a) Tafoni in sandstone, Ellesmere Island, Canadian Arctic Archipelago. (b) Weathering pits on a granite tor, Cairngorm Mountains, Scotland. (c) Advanced cavernous weathering of a boulder, Dry Valleys, Antarctica.

Figure 10.6 Detachment of weathering rinds in basalt. (a) Development of a weathering rind ~2 mm thick. (b) Frost weathering removes the outer part of the rind. Surface roughening increases and microcracks form within the rind. (c) Microcrack enlargement by ice‐lens growth, fungi, endolithic lichens or algae leads to rock flaking.

Figure 10.7 Model of karst morphology in rugged terrain underlain by discontinuous permafrost.

Figure 10.8 Alpine karst features at 2470 m in the French Alps. (a) Small dolines (sinkholes) formed by roof collapse of an underlying cave system. (b) Solution runnels (karren) formed along a rock joint.

Figure 10.9 (a) Granite tor, Cairngorm Mountains, Scotland. (b) Angular schist tor, Old Man Range, South Island, New Zealand. (c, d) Tors in interior Alaska.

Figure 10.10 Model of progressive glacial modification of granite tors. (a) Unmodified tor, with loose and toppled blocks, and weathering pits on the top and flanks. (b) Tor exhibiting slight glacial modification, with loss of fragile superstructure and removal of exposed blocks and regolith (oblique shading) from the stoss side. (c) More advanced glacial modification, characterized by a residual monolith, displacement of blocks and erratics (black) resting on tor surfaces. (d) Residual tor, comprising a tor plinth rising above glacially‐abraded bedrock, with evidence of glacial plucking on the lee side.

Figure 10.11 (a) Pit excavated in an openwork quartzite blockfield on An Teallach, a mountain in northwest Scotland. Exposed boulder surfaces are slightly rounded by granular disaggregation, and fine sediment underlies the openwork layer. (b) Pit excavated in a sandstone blockfield on the same mountain.

Figure 10.12 Sections excavated through blockfields on mountains in northwest Scotland. Sections 1–6 represent openwork blockfields on quartzite (1,2), granulite (3), microgranite (4) and gneiss (5,6). Sections 7 and 8 are sandy diamictons on arkosic sandstones. Sections 9–12 represent frost‐susceptible silt‐rich diamictons developed on granulite (9,10) and mica‐schist (11,12).

Figure 10.13 Relationships between different types of blockfield on mountains in northwest Scotland, showing the range of regolith types characteristic of fine‐grained quartzite, coarse‐grained quartzite and schist.

Figure 10.14 Model of blockfield evolution through progressive lowering of plateau surfaces during the Quaternary. WF, weathering front, at which bedrock weathering is focused. (1) Pre‐Quaternary weathering profile, comprising saprolite (chemically‐weathered regolith) and rounded corestones. (2) Stripping of regolith, leaving a lag of corestones with residual saprolite. (3) Initiation of Quaternary frost weathering: modification of remnant pre‐Quaternary regolith and frost heave of surviving corestones and frost‐weathered boulders. (4) Mature periglacial blockfield, formed by frost weathering at the weathering front and frost heave of boulders during successive Quaternary cold stages.

Figure 10.15 Periglacial trimline on Mount Hoffman, Yosemite, USA. The trimline marks the boundary between a lower zone of glacially‐scoured bedrock and the summit blockfield, and represents the maximum altitude of the Tioga Glaciation icefield (last glacial stage). The blockfield zone remained above the icefield as a palaeonunatak.

Figure 10.16 Glacially‐transported erratic boulder resting on a granite blockfield on Grytøya, arctic Norway.

Figure 10.17 Quartzite blockstream on the Falkland Islands. The boulders are up to 2 m long.

Figure 10.18 Brecciated chalk bedrock partly overlain by loessic deposits, Sussex, southern England. The brecciation becomes progressively coarser with depth, probably reflecting a former decrease in ground ice with depth during conditions of Pleistocene permafrost.

Chapter 11

Figure 11.1 The four components of solifluction, defined by freezing regime. The left‐hand diagram in each case illustrates ice distribution during diurnal (a,b) and annual (c,d) freezing, and the right‐hand diagram illustrates the form of the resultant velocity profile. One‐sided seasonal freezing (c) also causes annual frost creep and gelifluction even when permafrost is absent.

Figure 11.2 Needle ice crystals 12 cm long that have pushed up a layer of soil and small stones. Segmentation in the crystals indicates two periods of crystal growth uninterrupted by thaw. When the crystals melt, they deposit the uplifted soil and stones downslope from their original positions.

Figure 11.3 Schematic frost‐creep trajectories of soil particles associated with five annual cycles of (a) one‐sided and (b) two‐sided ground freezing (heave) and thawing (settling). The amount of heave and resettling has been exaggerated for clarity. The dashed line represents the resultant vertical velocity profile, and the columns indicate concentration of ice lenses in the frozen soil. In (a), the amount of heave diminishes downwards, reflecting decline in excess ice volume with increasing depth. In (b), most heave results from ice segregation near the base of the active layer, producing an S‐shaped velocity profile. One‐sided freezing in non‐permafrost terrain produces a velocity profile identical to (a).

Figure 11.4 Deformation of strain probes by frost creep over 6‐year periods on alpine slope crests in the Japanese Alps (left) and Swiss Alps (right), showing the characteristic concave‐downward reduction in creep velocity.

Figure 11.5 Downslope movement of soil plotted against mean annual air temperature (MAAT). (a) Surface velocity; (b) volumetric velocity; (c) maximum depth of movement.

Figure 11.6 (a) Turf‐banked lobe on Beinn a’ Bha’ach Ard, Scottish Highlands. (b) Stone‐banked lobe advancing over bedrock, southern French Alps.

Figure 11.7 (a) Solifluction sheet advancing over bedrock, southern French Alps. (b) Terminus of solifluction sheet in western Spitsbergen.

Figure 11.8 Longitudinal sections excavated through (a) a turf‐banked lobe and (b) a stone‐banked lobe, Niwot Ridge, Colorado Front Range.

Figure 11.9 Stratified slope deposits formed through successive burial of the stony treads of shallow stone‐banked lobes under lobes advancing downslope at 5000 m altitude in the Bolivian Andes.

Figure 11.10 Small ploughing boulder in the Fannich Mountains, northwest Scotland, showing the deep furrow formed by downslope movement of the boulder through the soil.

Figure 11.11 Relict stone‐banked lobe composed of granite boulders, Mourne Mountains, Northern Ireland. Such boulder lobes were probably last active under permafrost conditions during the Younger Dryas Stade (12.9–11.7 ka).

Figure 11.12 Gelifluctate (‘head’) deposits on the Gower Peninsula, south Wales. (a) Massive, crudely‐stratified gelifluctate overlying stratified slope deposits. (b) Fabric of gelifluctate at the same site, consisting of angular clasts embedded in a matrix of fines. The ruler is 30 cm long.

Figure 11.13 Stratified slope deposits at Verteuil, western France, consisting of alternating matrix‐rich and openwork fine gravel beds 5–15 cm thick. Compare with Figure 11.9.

Figure 11.14 Two recent active‐layer failures on the Fosheim Peninsula, Ellesmere Island, Canadian Arctic Archipelago. An older vegetated failure separates the two. The tops of ice wedges are visible in the polygons exposed by the recent failure on the left.

Figure 11.15 Tilted wooden piles representing the foundations of a building in Longyearbyen, Svalbard, that was destroyed during the Second World War. Since then, the piles have been slowly titled downslope as a result of permafrost creep and/or solifluction.

Figure 11.16 Displacement profile representing creep within massive ground ice over seven years; strain is largely confined to the lowermost third of the ice.

Figure 11.17 Stages in the development of cambering and valley bulging. (a) Valley incision into clay bedrock results in the formation of stress‐release joints in the caprock and in the development of a proto‐bulge on the valley floor. (b) Frost heave increases the size of the bulge and favours joint opening in the caprock. (c) Permafrost creep within ice‐rich clay produces cambering of valley sides, detachment of caprock blocks and increased valley bulging. (d) Thaw of excess ice within ice‐rich clays causes viscous flow of clay strata towards the valley axis, enhanced cambering and settling, and rotation of caprock blocks.

Figure 11.18 High arctic nivation model proposed by Christiansen (1998a) on the basis of observations in northeast Greenland. The model depicts the situation in late summer, with a perennial snowpatch occupying a nivation hollow that is being progressively enlarged by retrogressive failure of the backwall.

Figure 11.19 Small nivation hollow at 1120 m altitude on Ben Alder, Scottish Highlands. A late‐lying snowpatch feeds meltwater to a flooded, partly vegetated wash zone.

Figure 11.20 Idealized model of long‐term landscape evolution by cryoplanation. Though some landscapes in cold continental interiors exhibit stepped surfaces, the extent of pre‐Quaternary land‐surface inheritance is uncertain. It is unlikely that the full evolutionary sequence depicted has ever reached completion.

Figure 11.21 Slope forms in periglacial environments: (a) cliff and talus accumulation; (b) rectilinear debris‐mantled slope; (c) convexo‐concave slope characteristic of lowlands underlain by low‐resistance bedrock; (d) stepped slope profile due to the development of cryoplanation terraces.

Figure 11.22 Rectilinear debris‐mantled slopes, north Yukon, Canada.

Figure 11.23 Models of periglacial slope evolution: (a) with basal removal of debris; (b) with basal sediment accumulation.

Chapter 12

Figure 12.1 Talus accumulations below a cliff of sedimentary rocks, Svalbard. (a) Coalescing talus cones below rock gullies. The lower‐gradient cone is a debris cone fed by debris flows and snow avalanches. (b) Coalescing talus cones. The bench at the foot of the talus on the left is a talus rock glacier, formed by deformation of ice‐rich permafrost within talus debris.

Figure 12.2 Types of talus slopes and related landforms.

Figure 12.3 Model of talus development proposed by Francou and Manté (1990). The upper rectilinear slope experiences both rockfall input and downslope movement of debris. The basal concavity is a zone of rockfall input without significant downslope movement of debris.

Figure 12.4 Mechanisms of talus stratification. (a) Dry grainflow. (b) Downslope advance of stone‐banked solifluction lobes. (c) Frost‐coated clast flows (logged section). (d) Alternation of sediment movement and niveo‐aeolian deposition (logged section).

Figure 12.5 (a) Snow avalanche debris on a talus slope, Longyeardalen, Svalbard. (b) Snow avalanche debris after deposition. Note the loose, angular, ‘perched’ debris emplaced by ablating snow. (c) Snow‐filled avalanche chutes feeding snow avalanches that have descended over talus, Larsbreen, Svalbard. (d) Avalanche impact rampart, Jostedalen, Norway. Wet‐snow avalanches occur annually down the slope right of the river, and the force of impact has ejected rounded boulders from the riverbed on to the opposite bank, forming the rampart.

Figure 12.6 Small avalanche boulder tongue (roadbank tongue) flanked by debris‐flow tracks, Southern Alps, New Zealand.

Figure 12.7 Downslope transport by debris flows of quartzite debris across darker sandstones, Beinn Eighe, Scotland. The levées of some individual flow tracks are clearly evident.

Figure 12.8 Magnitude–frequency relationship for debris flows on European mountains. Arve, Zell and Zillertal are located in the central Alps, Ariège is in the Pyrenees and Bachelard is in the French Alps.

Figure 12.9 (a) Muragl rock glacier, Engadin, Switzerland, showing conspicuous tranverse ridges in its lower part. (b) Talus rock glaciers, Vardeborgsletta, Svalbard. (c) Glacigenic rock glaciers, Holar Valleys, Iceland. Note residual glacier ice at the headwall. (d) Glacigenic rock glacier, Reindalen, Svalbard.

Figure 12.10 Genetic classification of rock glaciers: (a) talus rock glacier; (b) moraine rock glacier; (c) glacigenic rock glacier. Some authors prefer to refer to glacigenic rock glaciers as ‘debris‐covered glaciers’.

Figure 12.11 Glacier ice core exposed in a glacigenic rock glacier, Holar Valleys, Iceland.

Figure 12.12 Vectors (a) and rates (b) of surface movement of the Macun rock glacier, Swiss Alps.

Figure 12.13 Haeberli’s model of climatic constraints on talus rock‐glacier formation. The stippled area represents possible combinations of mean annual air temperature (MAAT) and mean annual precipitation that permit the development of talus rock glaciers.

Figure 12.14 Relict rock glacier on Mt Olympus, South Island, New Zealand. Subdued longitudinal and transverse ridges reflect melt‐out of internal ice.

Figure 12.15 (a) Arcuate protalus rampart, arctic Norway. The crest is 115 m long, and the distal slope is 15–20 m high. (b) Proximal slope and crest of the same rampart. The coarse angular debrison the proximal slope represents rockfall debris arrested at the rampart. (c) Miniature pronival ramparts formed by snow‐push, Romsdalsalpane, Norway.

Figure 12.16 Model of rampart development at the foot of a progressively thickening firn field.

Figure 12.17 The talus debris‐transfer system: processes and geomorphological response. Relationships between processes are simplified. Many talus landforms are composite.

Chapter 13

Figure 13.1 Fluvial landforms on Svalbard. A tributary stream draining a V‐shaped valley runs over a large alluvial fan to join a braided river on the valley‐floor floodplain.

Figure 13.2 Water and heat fluxes in the active layer above cold permafrost during (a) early winter, (b) spring and (c) summer.

Zf

, depth of freezing fronts during early winter freezeback;

Zpf

, permafrost table;

Z

th

, depth of summer thaw;

Z

w

, depth of summer water table.

Figure 13.3 Slope runoff. (a) In spring, snowmelt‐fed surface runoff over the frozen active layer predominates. (b) By early summer, surface runoff from residual snowbeds sinks to join subsurface groundwater flow sourced by thaw of ice in the active layer in a saturated zone below a shallow water table. (c) By late summer, groundwater flow is confined to a thin saturated zone above the permafrost table, unless reinvigorated by rainfall.

Figure 13.4 Hillslope surface flow, hillslope subsurface flow and river discharge for 1982, McMaster River catchment, Cornwallis Island, arctic Canada (74° N). The onset of flow on slopes precedes that in the channel due to accumulation of water behind snow dams in the channel.

Figure 13.5 (a) Runoff regime of McMaster River, which drains a 33 km

2

catchment on Cornwallis Island (74° N) in the Canadian Arctic Archipelago, 1978–81, showing interannual variations in the timing and magnitude of the annual snowmelt flood. The two discharge spikes after the end of the snowmelt flood in 1978 represent rainfall events. (b) Runoff regime of the Liard River, a tributary of the Mackenzie River, February–December 1974. The Liard River drains a 275 000 km

2

subarctic catchment (57° 30′–61° 20′ N) with its headwaters in the northern Rocky Mountains. The abrupt start of the snowmelt and ice break‐up period is evident, but high discharge is maintained throughout the summer by melt of snow and ice in the higher parts of the catchment, and by summer rainfall. The maximum discharge represents exceptionally high rainfall in the middle and lower catchment.

Figure 13.6 (a) Schei River, Ellesmere Island (78° N) near the end of the snowmelt flood. (b) A tributary of the Schei River in early August. Almost all snow in the catchment has melted and the streambed has dried up.

Figure 13.7 Rillwash over high‐arctic slopes, central Banks Island, Canadian Arctic Archipelago.

Figure 13.8 Regression lines summarizing the relationships between concentration of total dissolved solids (TDS) and concentration of suspended sediment with river discharge (Q) for a 29 km

2

catchment partly underlain by carbonate rocks, Ellesmere Island, Canadian Arctic Archipelago (78° N).

Figure 13.9 (a) Meandering river, north Yukon. (b) Anastomosing river channels, northern Yakutia. (c) Braided river channels on a periglacial sandur, central Banks Island.

Figure 13.10 Schematic representation of periglacial alluvial river types as a function of vegetation cover, ground freezing conditions and sediment supply.

Figure 13.11 Oblique aerial photograph of alluvial fans, Svalbard. The parallel levées on fan surfaces indicate debris‐flow deposition on fan surfaces.

Figure 13.12 Gorge excavated in sandstone, Ellesmere Island, NWT, arctic Canada.

Figure 13.13 Model of floodplain and terrace development during a single glacial–interglacial–glacial cycle.

Chapter 14

Figure 14.1 Approximate fluid and impact thresholds for quartz grains of different sizes, plotted against wind velocity two metres above the ground surface.

Figure 14.2 Wind erosion in periglacial environments. (a) Deflation surface carpeted by lag gravels, central Iceland. The boulders were deposited by the last ice sheet. (b) Residual ‘island’ of windblown sand ~0.4 m thick on a plateau deflation surface, Vestidalur, Faroe Islands.

Figure 14.3 Ventifact (windpolished boulder) resting on a deflation surface, central Iceland. The boulder is ~0.4 m high and exhibits pitting, fluting and polishing by aeolian abrasion.

Figure 14.4 Dune slipface at the margin of the Great Kobuk dunefield, northwest Alaska.

Figure 14.5 Typical cumulative grain‐size curves for loess and aeolian sand‐sheet deposits. The absence of overlap is because loess is produced by accumulation of airborne dust, whereas aeolian sand is a tractional deposit. Most aeolian sand deposits are better sorted (steeper curves) than loess.

Figure 14.6 Niveo‐aeolian deposits of interbedded sand and snow layers overlying dune slipface deposits, Great Kobuk dunefield, northwest Alaska.

Figure 14.7 Distribution of loess deposits in Alaska.

Figure 14.8 (a) Deflation surface and remnant deposit of windblown sand, Ward Hill, Orkney Islands, Scotland. (b) Wind stripes developed normal to dominant wind direction on a plateau in northern Scotland. The pole is 1 m long. (c) Pit excavated in poorly sorted, massive plateau‐top aeolian sediments on the summit of The Storr, Isle of Skye, Scotland. The pit is 2.9 m deep.

Figure 14.9 (a) Approximate distribution of major cold‐climate dunefields and sand sheets (shaded zones) in Europe, from the British Isles to Poland; their continuation beyond eastern Poland is not shown. The dashed line is the approximate southern limit of the last (Late Weichselian) ice sheet. (b) Distribution of major cold‐climate dunefields and sand sheets in the coterminous United States.

Figure 14.10 Late Weichselian aeolian sand facies in northwestern Europe. (a) Facies 1, of Late Pleniglacial age. Massive bedding and alternating sand and loamy sand beds with synsedimentary deformations. The black band separates the deformed OCS I from the undeformed OCS II. (b) Facies 2, of Late Pleniglacial age (OCS II). Alternating bedding of fine sand (lighter) and loamy fine sand to sandy loam (darker), with crinkly lamination and small‐scale injection features caused by deposition on a wet surface. (c) Facies 3, of Younger Dryas age (YCS II); horizontal and low‐angle cross‐bedding in fine to medium sand deposited on a dry surface in source‐proximal dunes. (d) Facies 4, of Younger Dryas age (YCS II), showing large‐scale dune slipface cross‐bedding in fine to medium sand in source‐proximal dunes.

Figure 14.11 Late Pleistocene loess deposits, Port Hills, South Island, New Zealand. The exposure is 3.5 m high.

Figure 14.12 Extent of loess deposits in the coterminous United States.

Figure 14.13 Extent of primary and reworked loess deposits in continental Europe.

Figure 14.14 Exponential decline in the thickness of Malan (last glacial stage) Loess along a WNW–ESE transect across the southern Loess Plateau in China.

Figure 14.15 Model of the effect of glacial–interglacial climatic changes on loess accumulation rates and soil development in the loess region of central China. A strong summer monsoon generates warm, moist conditions that promote soil development, whilst a weak summer monsoon is associated with cold, dry conditions, increased dust influx and loess accumulation.

Figure 14.16 The loess–palaeosol stratigraphy of the Chinese Loess Plateau over the last 2.5 Ma, correlated with the marine oxygen isotope stages and the timing of palaeomagnetic reversals. In the marine isotope record, glacial stages have even numbers and interglacials odd ones. Loess units are prefaced by the letter L and palaeosols by the letter S. Both are numbered from the top of the sequence downward.

Figure 14.17 Stratigraphy of the last glacial age loess sequence at Nussloch in Germany, showing millennial‐scale variations in grain size. Coarser grain sizes are inferred to represent more arid conditions and increased wind strength, whereas finer grain sizes correspond to periods of gleyed tundra soil formation, suggesting wetter conditions.

Chapter 15

Figure 15.1 Ice‐affected coasts of the northern hemisphere, showing the approximate limit of winter sea ice and the limit of summer (multi‐year) sea ice in 2009. The main zone of ice‐rich permafrost coastline extends from Banks Island via the Beaufort Sea, Chukchi Sea and East Siberian Sea to the Laptev Sea.

Figure 15.2 Boulder ridge formation by ice push. (a) Boulders freeze on to the base of pack ice. (b) Lateral thrusting stacks boulder‐bearing ice on the shore. (c) Boulder‐rich ice ridge. (d) Boulder ridge after melt of ice. Most boulder ridges develop as a result of repeated ridge‐building events.

Figure 15.3 Ice‐rich yedoma sediments exposed by coastal erosion, Muostakh Island, southern Laptev Sea. The large syngenetic ice wedges are thought to be of Late Pleistocene age. Coastal retreat is ~50 m a

−1

in this area.

Figure 15.4 Retrogressive thaw slumps on the southeast coast of Herschel Island, Beaufort Sea Coast, arctic Canada. The large slump extends ~500 m inland and is ~400 m wide; the headwall reaches a height of ~25 m.

Figure 15.5 Coastline evolution in thermokarst terrain. (a) Embayments and headlands formed from transgressive coastal erosion of thermokarst lakes. (b) Formation of spits composed of sediment derived from headlands. (c) Isolation of offshore barrier islands by coastal retreat. (d) Migration or drowning of barriers. All of these stages occur along the Beaufort Sea coast.

Figure 15.6 Calton Point, a gravel spit extending from the Yukon coast south of Hershel Island, arctic Canada. Similar gravel spits and barrier islands extend westwards along the Beaufort Sea coast to Point Barrow, Alaska.

Figure 15.7 Intertidal rock platform, northwest Hornsund, Svalbard.

Figure 15.8 Raised rock platform on the west coast of Scotland. Formation of this platform has been attributed to a combination of frost weathering of bedrock and removal of debris by storm waves or sea ice during the Younger Dryas Stade (~12.9–11.7 ka), when permafrost extended to sea level.

Figure 15.9 Evidence for frost weathering of gneissic bedrock and formation of an incipient shore platform, Böverbrevatnet, Jotunheimen, Norway. Partial drainage of the lake has exposed platform fragments above present lake level.

Figure 15.10 Model of lacustrine rock‐platform development due to frost weathering. (a) Lake‐ice thickening, frost penetration into bedrock and movement of unfrozen water towards the freezing plane. (b) The same processes superimposed on a developing rock platform.

Chapter 16

Figure 16.1 The marine oxygen isotope (marine isotope stage, MIS) record for the last 0.9 Ma, based on a composite of deep‐ocean cores and the equivalent terrestrial Middle and Late Quaternary stratigraphy of the northern hemisphere. LP, Late Pleistocene. The even numbers in the MIS record represent glacial stages, and interglacial stages are shown in italics.

Figure 16.2 The NGRIP δ

18

O Greenland ice core record for the past 32 ka, showing the major subdivisions of the last (Late Devensian/Weichselian/Wisconsinan) glacial substage. YD, Younger Dryas Stade; LI, Lateglacial Interstade (Bølling–Allerød Interstade in Europe); GB, Great Britain.

Figure 16.3 Schematic reconstruction of the ages and relative ages of periglacial landforms, sediments and structures. For details, see text.