• No results found

The recent history of the Antarctic ice sheet : constraints from sea-level change

N/A
N/A
Protected

Academic year: 2019

Share "The recent history of the Antarctic ice sheet : constraints from sea-level change"

Copied!
301
0
0

Loading.... (view fulltext now)

Full text

(1)

The recent history of the

Antarctic Ice Sheet:

constraints from sea-level change.

Daniel Peter Zw artz Septem ber 1995

(2)

The w ork described in this thesis was carried out w hile I was a full-tim e student at the Research School of Earth Sciences, at The A u s tra lia n N a tio n a l U n iv e rs ity , b e tw e e n Ju n e 1991 an d Septem ber 1995. Except w here m entioned in the text, the research d escrib ed here is m y ow n. No p a rt of this thesis has been subm itted to any other university or sim ilar institution.

(3)

Acknowledgments

N ow th at I am finally near the end of this thesis, I feel deeply grateful to everyone w ho has helped me along the way, both at w ork and at play. I w as lucky to have an excellent supervisory panel. Kurt Lambeck's su pervision contained suitable p roportions of encouragem ent and persuasion, and his p erseverence in constructively review ing my final drafts u n d e r difficult conditions is greatly appreciated. John Stone and Herb M cQ ueen alw ays provided good advice and their assistance is gratefully acknow ledged. M any people assisted me in m y data collection, and w ithout them the w ork w ould n o t have been possible. John Stone and Michael Bird let m e join th eir V estfold H ills expedition, and w ere excellent com panions. They also g en ero u sly let me use the resu lts of analyses they co n d u cted on th e A ntarctic samples. David H opley was always helpful and his expertise w as m uch appreciated d u rin g field w ork on the Great Barrier Reef. C laudine Stirling, Jerem y Taylor, and M ark and Geoff on the R.V. Kirby w ere also in v alu ab le at this tim e. Perm ission for field w ork in Q u een slan d w as granted by the Great Barrier Reef M arine Park A uthority, and the N ational T idal Facility at F lin d ers U niversity and the Q ueensland M in istry of T ransport provided tidal data. U npublished reef core data from the G reat Barrier Reef were generously provided by David Hopley and John M arshall. Dick Jenssen kindly provided me w ith ice and bedrock m aps of A ntarctica in digital form , th u s saving me a great deal of w ork. My colleagues in G eodynam ics have h elp ed m ake m y tim e here enjoyable, an d I w o u ld p articularly like to th an k Jean Braun, Paul Johnston, Tony Purcell, Vivien Gleeson and Janine D olton for their assistance in various fields w henever it w as needed. Sharing an office w ith Deb Scott was a stim ulating experience w hich I enjoyed imm ensely, and learned a great deal from. It has taken tw o people to replace her, b u t Tony Purcell and Geoff Batt have done a good job of filling the gap. I have enjoyed the com pany of my fellow stu d en ts at RSES, particularly A dam Kent, M elita Keywood, Dave John Brown, Steve Pell, A lfredo Camacho, Sue Keay, Claudine Stirling and Geoff Fraser, am ong others. Finally, for their com m ents on drafts of this thesis, I w o u ld like to thank John Stone, Michael Bird, Paul Johnston and Herb McQueen.

(4)

Abstract

The rheology of the earth and the history of ice sheets, which have a major contributing role in climate change, are both subjects of considerable interest in earth sciences, and the stu d y of sea-level change provides insight into both. Sea-level change since the last glacial m axim um (LGM), about 18,000 years ago, can be explained as the sum of three contributions: the sea-level rise due to m eltin g of the Pleistocene ice sheets; the iso static and gravitational response to the m elting of these ice sheets; and the isostatic and gravitational response to the w ater added to the oceans. Thus, sea-level ch an g e v aries w ith location, and is d e p e n d e n t on the v o lu m e and distribution of the ice rem oved, the shape of the oceans, and the rheological structure of the earth. Using sea-level records from appropriate locations, a n d num erical m odels of the response of the e arth to surface loads, constraints can be placed on some of these param eters. In this thesis, I use new sea-level observations from Antarctica and Q ueensland to estim ate the form er distribution of ice at several Antarctic sites, the total am ount of extra ice which was stored in Antarctica at the LGM com pared to the present, and the eustatic sea-level change which has occurred in the last 6,000 years.

I p resen t a new high-resolution sea-level record from the V estfold Hills, A ntarctica, obtained by dating the lacustrine-m arine and m arine-lacustrine transitions in sedim ent cores from lakes which were form erly connected to the sea. A sea-level m axim um ~9 m above present sea-level 6,000 years ago is docum ented. Sea-level observations from other Antarctic sites have been com piled and com pared w ith predictions derived from sim plified m odels of m elting at the ice sheet m argin. The results indicate that at the LGM the East Antarctic ice sheet m argin was 25 - 100 km beyond its present position, resulting in ice thicknesses of 500 - 1000 m at sites now on the coast.

Eustatic sea-level change in the last 6,000 years can be estim ated from the difference betw een sea-level predictions and observations at sites unaffected by details of the ice sheet reconstructions. Using new sea-level observations and a com pilation of published data from north Q ueensland, a eustatic sea- level rise of 3 - 6 m in the last 6,000 years is inferred, w ith the rate of rise decreasing tow ards the present.

(5)

C o n ten ts

1 Introduction: Sea-level change, ice sheets and climate... 1

1.1 What can sea-level tell us about climate and climate history? 1 1.1.1 What forces climatic cyclicity?... 2

1.1.2 How much ice was on Antarctica at the LGM?... 3

1.1.3 How was it distributed?... 3

1.1.4 When did it melt?...4

1.2 Modelling sea-level change... 5

1.2.1 Glacio-hydro-isostasy... 5

1.2.2 Previous results from glacio-hydro-isostasy... 5

1.3 Content of this thesis... 7

1.3.1 Some definitions...8

Part I : Sea-level and ice sheet observations... 9

2 Sea-level Observations from Far North Queensland... 11

2.1 Introduction... 11

2.2 Methods ... 13

2.2.1 Microatolls... 13

2.2.2 Reef core data...16

2.2.3 Beachrock... 17

2.2.4 Chenier plains... 17

2.2.5 Oyster shells... 18

2.2.6 Other m ethods...19

2.3 Previous work... 20

2.3.1 LGM sea-level... 20

2.3.2 The postglacial transgression... 21

2.3.1 Sea-level change since ~6 k a ...27

Microatolls... 28

Oyster shell deposits...31

Drill core data... 33

Beach ridges... 33

2.4 Sea-level change at Orpheus Island...35

2.4.1 Introduction... 35

2.4.2 Study sites... 35

Pioneer Bay... 35

Pioneer Bay South... 35

Little Pioneer Bay... 36

Iris Point...36

Hazard Bay... 37

2.4.3 Sampling procedure... 37

Microatolls...37

Clam shells...37

(6)

Reduction of levelling data...37

2.4.4 Radiocarbon dates... 38

Preparation of samples... 38

Results... 38

2.4.5 Interpretation... 40

Pioneer Bay South...40

Discussion...42

2.5 Northern GBR microatolls... 44

2.5.1 Introduction... 44

2.5.2 Sample collection...46

2.5.3 Results...46

Microatolls...46

Oysters... 49

2.5.4 Discussion...50

2.6 Drillcore data... 52

2.6.1 Introduction... 52

2.6.2 Data sources... 52

2.6.3 Sites studied... 53

2.6.4 Results... 54

3 Sea-level Observations from Antarctica...57

3.1 Introduction... 57

3.2 Methods ...59

3.2.1 Raised Beaches... 59

3.2.2 Isolated Lakes... 59

3.2.3 Marine Lim it... 62

3.2.4 Limiting Observations...62

3.2.5 Radiocarbon dating... 62

3.3 Vestfold H ills... 65

3.3.1 Previous w ork... 65

3.3.2 New observations... 70

Collection of cores... 70

Core stratigraphy and sediment characteristics...71

Interpretation... 76

Dating the marine-lacustrine transitions... 77

Sea-level history of the Vestfold Hills...80

3.4 Victoria Land... 83

3.4.1 McMurdo Sound region...84

3.4.2 Terra Nova Bay...87

3.4.3 Marine limits... 89

3.4.4 Miscellaneous sea-level observations... 91

3.5 Windmill Islands...92

3.5.1 Marine limits... 93

3.5.2 Lake sediment cores... 94

3.6 Bunger H ills... 96

3.7 Larsemann Hills...99

3.8 Soya Coast... 100

3.9 Antarctic Peninsula... 103

(7)

4 The Antarctic Ice Sheet...107

4.1 Introduction...107

4.2 Glacial Geology... 109

4.2.1 Vestfold Hills... 109

Lake sediment ages... 109

Carbon analyses... I l l Cosmogenic isotope measurements...114

Discussion...116

4.2.2 Other coastal oases... 117

Larsemann H ills...117

Bunger Hills...117

Windmill Islands...118

Soya Coast...118

4.2.3 Ross Sea Region... 120

4.2.4 Lambert Glacier... 121

4.2.5 Sor Rondane M ountains...124

4.2.6 Antarctic Peninsula... 124

4.3 Marine Geology... 125

4.3.1 Weddell Sea...125

4.3.2 Prydz Bay... 125

4.3.3 Ross Sea... 125

4.3.4 Wilkes Coast...126

4.3.5 Antarctic Peninsula... 126

4.4 Ice Cores ... 127

4.4.1 Total gas content... 127

4.4.2 Isotopic records... 129

4.4.3 Cores from coastal regions... 131

4.4.4 Ice temperature profile... 131

4.4.5 Discussion... 131

4.5 Theoretical M odels...133

4.6 Estimates of Modern Mass Balance...135

4.7 Sum m ary... 136

Part I I : Modelling sea-level change...137

5 Basics of glacio-hydro-isostasy... 139

5.1 Introduction...139

5.2 The sea-level equation... 142

5.3 Axisymmetric models...143

5.3.1 Axisymmetric ice sheet models... 143

5.3.2 Ocean m odels... 144

5.3.3 Earth rheological model... 145

5.4 Sea-level predictions for axisymmetric m odels... 147

5.4.1 Near field sea-level change... 147

Ice-load component... 147

Water-load component...149

Total sea-level change...151

5.4.2 Far field sea-level change...152

(8)

5.4.3 Moving coastlines in the far field...155

5.5 Realistic models... 158

6 Modelling sea-level change in North Queensland...159

6.1 Introduction...159

6.2 Sea-level at the Last Glacial Maximuum (LGM)... 161

6.3 The post-glacial transgression...165

6.3.1 Introduction... 165

6.3.2 Time at which present sea-level first attained... 165

Model predictions...165

Comparison with observations... 171

6.3.3 Sea-level Stillstands and regressions...176

6.4 Mid-late Holocene sea-level...178

6.4.1 Introduction... 178

6.4.2 Late Holocene eustatic sea-level change...179

Spatial variation... 183

7 The Antarctic ice sheet and sea-level change...185

7.1 Objectives of this chapter... 185

7.2 Previous work...186

7.2.1 The former Antarctic Ice Sheet... 186

7.2.2 Rheological models of the Earth... 189

7.3 Methods 191 7.3.1 Ice sheet configuration...191

7.3.2 Regional models... 192

Ice sheet retreat... 194

Distribution of ice removed...194

Timing of melting... 195

7.3.3 Continental-scale m odels... 197

Distribution of ice removed...197

Timing of melting... 198

7.3.4 Ocean m odels...199

Static coastline m odels...199

Moving coastline m odels... 201

Iterations of the sea-level equation... 203

7.3.5 Earth m odel... 203

7.3.6 Estimation of total relative sea-level change...203

7.4 Results - Regional models...204

7.4.1 Pattern of deformation...204

7.4.2 Shape of the RSL curve...204

7.4.3 ESL m odel... 207

7.4.4 Ice load distribution... 209

7.4.5 Ice volum e... 211

7.4.6 Lithospheric thickness...212

7.4.7 Upper mantle viscosity...214

7.4.8 Lower Mantle Viscosity...216

7.4.9 Time extent of models...217

7.4.10 Summary... 217

(9)

7.5.1 Elevation change inland... 219

7.5.2 Distant effects...223

7.6 Regional models - Comparison with Antarctic observations...226

7.6.1 Calculation of variance...226

7.6.2 Optimum ice model... 227

Ice thickness and retreat distance...227

Distribution of removed ice... 231

Timing of melting... 233

Late Holocene melting...234

7.6.3 Optimum Earth model... 235

Lithospheric thickness...235

Upper mantle viscosity...235

Lower mantle viscosity...236

7.6.4 Local spatial gradients in sea-level change... 236

7.6.5 Discussion...239

7.7 Continental-scale models - Comparison with observations...241

7.7.1 Elevation change inland...241

7.8 Comparison with Australian Holocene sea-levels...242

7.9 Realistic models...248

7.9.1 Estimating the total volume change of the Antarctic Ice Sheet...248

7.10 Sum m ary...252

8 Conclusion... 253

8.1 New Antarctic sea-level observations...253

8.2 Regional ice sheet reconstructions... 254

8.3 Total Antarctic ice sheet volume change... 255

8.4 Late Holocene melting...256

Appendix : Site Reports...259

(10)

List of Figures

2.1 A guide to tidal planes... 13

2.2 Living and fossil microatolls at Iris Point, Orpheus Island...14

2.3 The response of microatolls to different patterns of sea-level change...15

2.4 Dated mangrove peat and shell material from sediment cores in Princess Charlotte Bay...24

2.5 Dated mangrove and shell material from cores on the northern Great Barrier Reef near Cooktown... 24

2.6 Eustatic sea-level record for the last glacial cycle...27

2.7 Location map of the central to north Great Barrier Reef... 30

2.8 Height-age relationship for dated microatolls from the central to north Great Barrier Reef... 31

2.9 Relationship between microatolls and oyster shell beds at Magnetic Island...32

2.10 Proposed 5000 yr sea-level isobase... 34

2.11 Location map of Orpheus Island... 36

2.12 All coral and clam shell dates from Orpheus Island...41

2.13 Dated coral and clam shell samples from Pioneer Bay South... 41

2.14 Dated microatolls from Orpheus Island... 42

2.15 Location map of the Lizard Island - Princess Charlotte Bay region, Great Barrier Reef... 45

2.16 All dated microatolls from the Lizard Island - Princess Charlotte Bay region... 48

2.17 Dated microatolls from the Lizard Island - Princess Charlotte Bay region, by location... 49

2.18 Reef core location m ap ... 53

2.19 All dated coral samples from reef cores in the Great Barrier Reef... 55

2.20 Possible sources of error when inferring sea-level from a dated core sam ple... 56

3.1 Location map of Antarctica...58

3.2 Schematic representation of lake sedimentation in a changing sea-level environment... 60

3.3 Reconstruction of sea-level history from lake sediment records... 61

3.4 Published radiocarbon ages from the Vestfold Hills...66

3.5 Total emergence and inferred average emergence rate in the Vestfold H ills... 67

3.6 Published constraints on the sea-level curve for the Vestfold H ills... 67

3.7 Map of W atts/Nicholson/Anderson lake system...68

3.8 Aerial view of Anderson Lake, Watts Lake and Ellis Fjord... 69

3.9 Locations of lakes cored in the Vestfold H ills... 72

(11)

3.11 Location map for Victoria Land... 83

3.12 Sea-level curve for McMurdo Sound... 86

3.13 Sea-level curve for Terra Nova Bay...89

3.14 Elevation of the marine limit in Victoria Land... 90

3.15 Location map for Windmill Islands... 92

3.16 Sea-level curve for Windmill Islands...95

3.17 Sea-level curve for the Bunger Hills... 98

3.18 Location map for Lützow-Holm Bay... 100

3.19 Sea-level curve for Skarvsnes, in Lützow-Holm Bay...101

3.20 Sea-level observations from King George Island ... 104

3.21 Summary of sea-level observations for Antarctic sites... 105

4.1 Location map of Antarctica... 108

4.2 TOC and 513C profiles from Scale Lake and Lake McCallum... 113

4.3 Cosmogenic isotope exposure ages from the Vestfold Hills... 115

4.4 Location map for Lützow-Holm Bay... 119

4.5 Profiles of drift sheets in the Transantarctic Mountains...122

4.6 Reconstructions of the Ross Ice Sheet...123

4.7 Total gas volume measurements in the Vostok ice core... 129

4.8 Maximum reconstruction of the LGM Antarctic Ice Sheet...134

5.1 Schematic illustration of glacio-hydro-isostatic sea-level changes... 140

5.2 Profiles of axisymmetric ice sheet models... 144

5.3 The ice-load component in the near field... 148

5.4 The water-load component in the near field... 150

5.5 Components of predicted sea-level change for one and two iterations of the water load... 151

5.6 The ice-load component in the far-field... 153

5.7 The second iteration of the water load in the far field...154

5.8 Total predicted sea-level change at three far field sites...155

5.9 Bathymetric profiles of the far field coastline... 157

5.10 The second iteration of the water-load component at the far field coastline... 157

6.1 Predicted sea-level at the LGM, 18 000 years ago... 162

6.2 Predicted sea-level at the LGM in the Australian region... 164

6.3 Predicted sea-level curves at sites on a transect of the GBR... 166

6.4 Predicted sea-level histories on a transect of the north Queensland coast,... 167

6.5 Predicted time at which sea-level first reached its present level in the Australian region...168

6.6 Predicted first attainment time in the study area on the GBR... 169

6.7 Same as Figure 6.6, but for models with a 50 km lithosphere...170

6.8 Predicted and observed spatial gradients in the first attainment time... 171

6.9 Same as Figure 6.8, for models with a 50 km lithosphere... 172

(12)

6.11 Same as Figure 6.9, but including 1.7 m of late Holocene

melting...174

6.12 Same as Figure 6.8, but including 3 m of late Holocene melting... 175

6.13 Same as Figure 6.9, but including 3 m of late Holocene melting... 175

6.14 Predicted relative sea-level curves at Barbados and Halifax Bay...177

6.15 Predicted relative sea-level curves for inner and outer margins of continental shelves...178

6.16 Observed-predicted sea-level residuals for mid-late Holocene observations in north Queensland... 180

6.17 Same as for Figure 6.16, but for an earth model with 100 km lithosphere and a uniform mantle viscosity of 1021 Pa.s... 181

6.18 Comparison of sea-level residuals from north Queensland with the result of Lambeck (1993b)...182

6.19 Spatial variation in observed-predicted sea-level residuals... 183

7.1 Published models for the Antarctic meltwater contribution...187

7.2 Example predicted and observed far-field sea-level curves...187

7.3 Distribution of ice load for axisymmetric models... 192

7.4 Transects of the Antarctic Ice Sheet at study sites... 193

7.5 Distribution of ice removed for constant-height and constant-b m odels... 195

7.6 Three melting curves used for the regional-scale ice sheet m odels... 196

7.7 Calculation of the load of ice grounded below sea-level... 202

7.8 Sea-level change due to the regional model... 205

7.9 Comparison of regional models with the actual Antarctic Ice Sheet... 206

7.10 Predicted RSL history at a site on the present ice sheet margin...207

7.11 Effect of late Holocene melting on the RSL at a coastal site... 208

7.12 Ice load component of sea-level change at a coastal site predicted by the three regional ice-melting curves... 209

7.13 Predicted relative sea-level curves at a coastal site for constant-h and constant-b ice sconstant-heets and tconstant-hree melting constant-histories...210

7.14 The effect on the predicted relative sea-level history at a coastal site of the volume of the regional ice sheet model...211

7.15 The effect of lithospheric thickness on RSL at a coastal site...212

7.16 The effect of upper mantle viscosity on RSL at a coastal site...214

7.17 The effect of lower mantle viscosity on RSL at a coastal site... 216

7.18 Subdivision of an ice load into annular rings...220

7.19 The distribution of removed ice for several melting styles and two values of total ice volume... 222

7.20 The effect of ice volume and distribution on predicted sea-level change in the region beyond the peripheral bulge... 224

7.21 The effect of upper mantle viscosity on predicted sea-level change in the region beyond the peripheral bulge... 225

7.22 The effect of lithospheric thickness on predicted sea-level change in the region beyond the peripheral bulge...225

(13)

7.24 Comparison between observed and predicted sea-level in the

(14)

List of Tables

2.1 Depths and ages of drowned shorelines in the central GBR... 22

2.2 Radiocarbon ages of samples from sediment cores in Princess Charlotte Bay...23

2.3 Radiocarbon ages of mangrove muds from the transgressive unit, northern GBR... 25

2.4 Proposed transgressive sea-level Stillstands... 25

2.5 Radiocarbon ages of raised Porites microatolls...29

2.6 Radiocarbon ages of coral samples from Orpheus Island...39

2.7 Radiocarbon ages of Tridacna clam shell samples...40

2.8 Study sites in the central and northern GBR... 46

2.9 Radiocarbon dates from central and northern GBR... 47

2.10 Comparison of elevations of modern and ancient oyster shell deposits...50

2.11 Locations of sites for which borehole data have been compiled...54

3.1 Locations of Antarctic sites discussed in this chapter... 57

3.2 Estimates of the marine reservoir correction at some locations in East Antarctica... 63

3.3 Radiocarbon assays from modern algae from the Vestfold Hills...64

3.4 Radiocarbon ages from marine organisms in emerged marine terraces in the Vestfold Hills... 65

3.5 Published former sea-levels derived from isolation of lakes in the Vestfold Hills... 70

3.6 Lakes cored in the Vestfold Hills...73

3.7 Descriptions of cores taken from lakes in the Vestfold Hills... 75

3.8 Conventional radiocarbon ages obtained from cores from lakes in the Vestfold Hills... 78

3.9 Ages of marine-lacustrine transitions in lakes in the Vestfold Hills...80

3.10 Radiocarbon ages from emerged marine sediments in McMurdo Sound... 85

3.11 Radiocarbon ages from Terra Nova Bay...88

3.12 Marine limit observations from the Windmill Islands...93

3.13 Radiocarbon ages from sediments in lakes in the Windmill Islands... 94

3.14 Radiocarbon ages from the Bunger Hills...98

4.1 Age of lowest organic-rich lacustrine sediments in lakes of the Vestfold Hills... 110

4.2 Radiocarbon ages from till in cores from the Vestfold Hills...110

6.1 Rheological models used in this chapter... 160

(15)

6.3 Scatter of observed-predicted sea-level residuals...181

7.1 - Regional contributions to postglacial sea-level rise from the Antarctic Ice Sheet,...189

7.2 The three regional ice sheet ESL models... 197

7.3 ESL models used in the continental-scale ice sheet models... 198

7.4 Removed ice distribution models in the continental-scale ice sheet models... 198

7.5 Initial and final dimensions of the continental-scale ice sheet m odels...199

7.6 Rheological models in this chapter... 203

7.7 Effect on RSL curve of varying the model parameters...218

7.8 Surface elevation changes at the centre of the ice sheet...222

7.9 Estimated former ice thicknesses and margin retreat distances at sites in East Antarctica...228

7.10 Estimated former ice thicknesses and margin retreat distances at sites in East Antarctica, using a standard earth model and only considering models which include significant late-Holocene m elting... 229

7.11 Predicted spatial gradient in the height of the 6 ka sea-level highstand... 238

7.12 Components of RSL at 6 ka at several East Australian sites... 244

(16)

Chapter 1

Introduction: Sea-level change, ice sheets and

c lim a te

1.1

What can sea-level tell us about climate and climate history?

The h isto ry of E arth's clim ate in the Cenozoic is dom inated by gradual cooling and associated accum ulation of ice on land since the Oligocene. A lthough the prim ary control on the onset of glaciation is tectonic, nam ely the iso latio n of the A ntarctic co n tin en t allow ing the fo rm atio n of the circum -polar current, ice sheets them selves play a critical role in the global clim ate system . For exam ple, their high albedo influences the n et solar rad iatio n retained by the earth; the form ation of sea ice at their m argins controls the production of deep cold ocean currents, which are an im portant global h e at reservoir; and th ey store significant a m o u n ts of w ater, controlling global sea-levels. Sea-level in tu rn influences the p a ttern of ocean cu rren ts, the d istrib u tio n of land and w ater on the continental shelves, and hence the p attern s of vegetation, w ind and rainfall w hich constitute the earth's climate. A record of sea-level, then, w hen com pared to other climatic records, provides us w ith an opportunity to establish some of the m echanism s o p eratin g b etw een different aspects of the clim atic system . Also, the form er sea-level is a m easure of the am ount of ice on land, so the interrelationships betw een ice sheets, sea-level and clim ate can be investigated through sea-level studies.

N atu rally , these interrelationships, and m any others w hich com prise the clim ate system , are very com plex and are the subject of a great deal of current research. General Circulation M odels (GCMs) are used to stu d y the physical and therm odynam ic interaction of the atm osphere and oceans, and produce predictions for tem perature, w ind and precipitation patterns. These all d ep en d prim arily on short-term processes, and require that the climate- influencing features such as ice distribution and sea-level, w hich change on longer time scales, be given as p a rt of the initial conditions. The predictions p ro d u ced are then com pared w ith clim ate records such as those derived from vegetation histories, deep sea sedim ents and ice cores. The degree of ag reem en t betw een predictions and observations is u sed to assess the v alid ity of the m odelling and the significance of the processes w hich are m odelled. Thus, if the climate system is to be properly un d ersto o d using this m ethod, a good record of sea-level change and ice distribution and an u n d erstan d in g of the relationship betw een them is essential.

(17)

change due to global w arm ing involve estim ating the effects of sea-level rise due to th erm al expansion of the oceans and increased m eltin g of ice, balanced against the possibility of sea-level fall due to increased snow fall onto existing ice sh eets, in a w arm er, m ore h u m id e n v iro n m e n t (eg W arrick and O erlem ans 1990).

D eterm ining the relative im portance of these effects involves com paring global tem perature and sea-level records, w hich are generally only available w ith sufficient resolution for the past few decades. The sea-level history, as recorded by tide gauges, is noisy and the com ponent due to global w arm ing can only be determ ined if all other effects are accounted for. The continuing sea-level changes due to the deglaciation w hich was largely com plete 6,000 years ago m ay still be significant in this regard.

Some of the m ajor o u tstan d in g problem s of Q uaternary clim ate change w hich m ay be addressed by studies of ice sheets and sea-level change are discussed below.

1.1.1 What forces climatic cyclicity?

The m ost striking feature of the Q uaternary climate record is its cyclicity: glacial an d interglacial p erio d s have alternated w ith a p erio d of aro u n d 100,000 years for m ost of this time. Analysis of oxygen isotope records from deep sea cores show s th at frequencies of around 23,000, 41,000 and 100,000 years are significant. These m atch well w ith the periodic variations in the earth's orbital param eters, w hich control the am ount of the su n 's rad iatio n the p lan et receives and its d istrib u tio n w ith latitude and th ro u g h o u t the year. This suggests a strong link betw een insolation and climate, but by no m eans an exclusive one. O ther possible controls on periodic climate change include the irregular cyclicity inherent in complex system s w ith feedback m echanism s, and singular events such as volcanic eru p tio n s, w hich m ay reduce th e earth 's insolation by injecting d u st into the atm o sp h ere, or induce w arm ing by the addition of greenhouse gases.

(18)

decay of ice sheets. To resolve this problem , the various clim ate records m u st be well constrained in time, so that the leads and lags betw een the associated processes can be determined.

1.1.2 How much ice was on Antarctica at the LGM?

As m entioned above, the eustatic sea-level record is a direct m easure of the am o u n t of terrestrial ice, b u t only the total volum e is indicated. For a record of the spatial distribution of ice, w e need in d ep en d en t evidence from the continents w here the ice was located. Most of this is the geom orphological sig n atu re of ice m asses - m oraines, striated pavem ents, roches m outonees, and o th er glacial and periglacial features. Along coastlines, the record of post-glacial reb o u n d also describes the d istrib u tio n of form er ice sheets. Because of the usually destructive nature of glacial advances, and the fact th a t the earth 's tim e constant for isostatic relaxation is a ro u n d a few th o u sa n d years, the records from geom orphological and reb o u n d data respectively are useful only for the last cycle of deglaciation. Luckily, this time period falls completely w ithin the range of radiocarbon dating.

The existence and approxim ate size of form er ice sheets in N o rth America, Fennoscandia and Britain and their effects on sea-level had been established from geom orphological evidence long before the sea-level record w as discovered. M ore recently, significant ice over the Barents Sea region has been established prim arily using rebound data (Lambeck 1995), bringing the total n o rth ern hem isphere contribution to ~90 m (N akada and Lam beck 1988, T ushingham and Peltier 1991). O bservations of sea-level at the last glacial m axim um range from -114 to -175 m (eg Veeh and Veevers 1970, C arter and Johnson 1986), so it is apparent that large bodies of ice m ust have existed elsewhere as well. Com parison of m odels and observations (N akada and Lambeck 1988, Tushingham and Peltier 1991) indicates th at around 25 - 35 m of equivalent sea-level was contributed from other sites. A ntarctica is the m o st favoured candidate for this ice, b u t other sites such as eastern or w estern Siberia rem ain as unproven possibilities (Grosswald 1980). A better determ ination of the total volume of the Antarctic ice sheet at the last glacial m axim um is therefore a useful addition to reconstructing global clim ate at that time.

1.1.3 How was it distributed?

(19)

D enton and others 1991), as w ell as isotopic studies of ice cores (eg Lorius and others 1985) and some isostatic rebound observations (eg C olhoun and others 1992). These studies indicate that at the last glacial m axim um , the Antarctic ice sheet w as considerably w ider than at present, extending to the edge of the continental shelf, b u t that the central height was about the sam e or perh ap s even lower. M ore observations and detailed isostatic m odelling will add to the understanding of ice sheet dynamics.

1.1.4 When did it melt?

(20)

1.2

Modelling sea-level change

1.2.1 Glacio-hydro-isostasy

Reliable clim ate records are not sufficient to establish the processes w hich link the m any observable aspects of climate, past and present: we need to describe the relationship betw een causes and effects quantitatively. As is the case in GCM studies, the problem is approached by using a m odel and a set of in itial co n d itio n s to m ake a p red ictio n , w hich is co m p ared w ith observations. In this study, the system being m odelled is the isostatic response of the w hole earth to changing loads at the surface. Schematically, w e m ay consider the system as a set of forces F pro d u cin g a response R filtered according to an earth model E.

The forces, F, are the changing loads of ice and w ater on the earth's surface, as m ass is m oved from the oceans onto the h ig h -latitu d e land m asses d uring glacial periods, and then redistributed w hen the ice sheets melt.

The param eters of the m odel E are the rheological properties of the earth - density, elasticity, and viscosity, and their variation w ith depth. The radial variation in the elastic m oduli has been estim ated from seismological data, and one of the m odels thus obtained (PREM, D ziew onski and A nderson 1981) is used for this work. The earth's rheological properties are time-scale- d ep en d an t, and in studying clim ate-related processes, we are interested in tim escales of 1,000 to 100,000 years, com pared w ith seconds to hours in seism ological studies. Similarly, the loads we are interested in are oceans and ice sheets (hundreds of m etres to a few kilom etres thick, and tens to th o u sa n d s of km w ide), and the resp o n ses are observable as vertical m ovem ents of up to h u n d red s of m etres, com pared w ith the small ground displacem ents recorded by seism om eters. Similar studies have been m ade m ade on these long timescales by looking at changes in the earth's rotation rate, also due to changing ice loads, and other loading problem s on the crust and u p p e r m antle.

Because the processes w hich link different aspects of climate are complex, nonlinear system s, I w on't try to solve them directly from the observations as an inverse problem . Instead, I w ill use the m athem atical m odel to p ro d u ce p red ictio n s from a range of plau sib le m odels and use sets of observations to identify the effects of the different variables and select the best-fitting solution.

1.2.2 Previous results from glacio-hydro-isostasy

(21)
(22)

1.3

Content of this thesis

This thesis addresses some of the problem s discussed above, using sea-level an d ice sheet observations from A ntarctica and A ustralia, and num erical m odels of sea-level change. The thesis is divided into tw o parts. Part I presents the observations, w ith little discussion of their w ider im plications. In Part II, glacio-hydro-isostatic m odels of sea-level change are presented, and the observations from Part I are used to constrain several aspects of the ice sh eet/earth /sea-lev el system.

In C h a p te r 2, new sea-level o bservations from n o rth Q u een slan d are p resen ted . These observations are all from the m id-late H olocene, w hen sea-level has been at roughly its present level. In particular, form er sea- levels w ere recorded for several islands in the L izard Island - Princess Charlotte Bay region, and a detailed record is obtained for O rpheus Island, in H alifax Bay near Townsville. In addition to the new data, published and un p u b lish ed sea-level observations are compiled and review ed, including a large data set of dated coral from reef flat cores, w hich constrain sea-level during the period of rapid rise before 6000 years ago.

Records of A ntarctic sea-level change are presented in C hapter 3. A new record from the Vestfold Hills, based on the isolation of lakes from the sea, defines sea-level for the last 6000 - 7000 years w ith m uch greater precision th an can be obtained by dating raised beach features, w hich has been the m ain source of A ntarctic sea-level observations. Published records from several sites d istrib u ted aro u n d East A ntarctica are also com piled and review ed.

The recent history of the A ntarctic ice sheet is discussed in C h ap ter 4, particularly addressing the problem of how m uch additional ice w as stored there d u rin g the last glaciation. First, the evidence from glacial geology in the coastal oases and the Transantarctic M ountains is discussed, including new observations from the Vestfold Hills. Results from m arine geology, ice cores and theoretical m odels are also compiled.

Part II begins w ith a general presentation of the principles of glacio-hydro- isostasy in C hapter 5. U sing the theory and num erical m odels of Johnston (1993), som e implications for sea-level change of m elting a circular ice sheet sim ilar in size to the form er Fennoscandian ice sheet are investigated. Particular attention is paid to coastal sites both close to the ice sheet and a long distance from it, as these are analogous to the Antarctic and A ustralian observation sites, respectively.

(23)

has occurred in the late H olocene. The last of these com ponents is also studied using the compiled sea-level observations from the last 6000 years. The m elting history of the Antarctic ice sheet is modelled in C hapter 7. Two types of m odel ice sheets are investigated, in order to investigate tw o aspects of the ice sheet history on w hich we have observational constraints. Both classes of m odels are circular, w ith parabolic cross-section, and do not realistically sim ulate the detailed geographical distribution of ice. First, regional m odels are constructed, which are intended to sim ulate the history of a part of the ice sheet m argin on the scale of -1000 km. These m odels are used to estim ate the form er ice thickness at those coastal sites from w hich sea-level records have been obtained (Chapter 3). The second type of m odel sim ulates the entire A ntarctic ice sheet on the continental scale. These m odels are used to investigate the predicted elevation change in the centre of the ice sheet, and the isostatic effects of changes in the Antarctic ice sheet on sea-level change in eastern A ustralia. These sea-level changes are m ainly d ep en d an t on the total ice volum e change, and A ustralian sea-level observations are com pared w ith the predictions in an attem p t to b etter constrain this.

1.3.1 Som e d e fin itio n s

Sea-level can appear to vary as a result of m any processes, such as eustasy (ie w ater volum e changes), regional isostasy, tectonism and ch anging tid al range. As these processes are global, regional and local, the term "sea-level change" can cause some confusion. In this work, any change in the position of the sea surface relative to the land is referred to as relative sea-level change (RSL), w hich is the height of m ean sea-level at that tim e above or below p resen t m ean sea-level. Thus, tectonic uplift of a site d u e to an earthquake causes an instantaneous fall in RSL, and an average uplift rate of 5 m m /y r im plies that RSL 1000 years ago was +5 m (assum ing that the sum of all other causes of sea-level change is zero).

P eriods of tim e in w hich RSL is positive (ie sea-level w as h ig h er th an present) are called highstands, and similarly times of negative RSL are called low stands. If sea-level rem ains at one level for an extended period of time, this is term ed a Stillstand.

(24)

Part I

Sea-level and ice sheet

o b serv a tio n s

Chapter 2 : Sea-level Observations from Far North Q ueensland... 11

Chapter 3 : Sea-level Observations from Antarctica... 57

(25)
(26)

Chapter 2

Sea-level Observations from Far North

Queensland

2.1

Introduction

To constrain the param eters of a glacio-hydro-isostatic m odel, it is necessary to have a know ledge of sea-level change both close to and distant from the form er ice sheets. Sea-level change in the form erly glaciated regions is dom inated by the effects of glacio-isostatic rebound, and these observations are necessary to constrain the form er d istrib u tio n of ice and its m elting history. The large am plitude of the rebound, however, m akes the near field record insensitive to som e aspects of the sea-level m odel, such as the am ount of eustatic sea-level rise in the late Holocene. In contrast, sea-level change at d istan t coastlines is dom inated by the eustatic sea-level rise and the hydro-isostatic response of the coastal region to this w ater-load. Sea- level change after 6 ka, w hen significant m elting of the ice sheets ceased, is most sensitive to this response. Also, if the geom etry of the form er ice sheet is n o t w ell k n o w n , th e n a n u m b e r of d ifferen t i c e /e a r th m o d el com binations m ay adequately describe the observed sea-level change in the near field. Sea-level change in the far field, how ever, is insensitive to the geom etry of the ice load, so better constraints on the m antle rheology m ay be obtained. These processes are discussed in detail in C hapter 5. N orth Q ueensland is a suitable far field region, because it is on the m argin of a tectonically stable continent, an d the n a tu ra l en v iro n m en t of the G reat Barrier Reef (GBR) is highly sensitive to sea-level change, and m ay preserve records of it .

The v ast length and b read th of the GBR m ake it a suitable region in which to investigate the differences in the sea-level record along transects both p e rp e n d icu lar to the coastline and, w ith observations from the rest of eastern A ustralia and Tasmania, along a line roughly radial to the Antarctic ice sheet. Also, the proxim ity of the Gulf of C arpentaria, w hich w as entirely above sea-level d uring the last glacial m axim um and th u s form s a large p o stg lacial w a te r load, m akes N o rth Q u een slan d an id eal reg io n to investigate the effects of large m igrations of the coastline on sea-level change.

(27)
(28)

2.2

Methods

This section sum m arises some of the principal m ethods used to reconstruct the sea-level history of north Q ueensland. N ot all these m ethods have been used in this study, but all are found in the large body of literature on the subject. All m ethods relate a geom orphic feature to some p a rt of the tidal range, and thus all are susceptible to error due to possible change in the tidal range. It is assum ed throughout this work that the tidal range has rem ained constant th ro u g h time, and this is a reasonable assum ption for sm all reef islands in the last 6 ka, w hich are surrounded by ocean and have not grow n sufficiently to im pede tidal w ater flow. This m ay not be the case for m ainland coastal sites, w here sedim ent deposition may have changed w ater d ep th enough to alter tidal range. A guide to tidal planes referred to in this chapter is show n in Figure 2.1.

Highest Astronomical Tide (H.A.T.)

Mean High Water Springs (M.H.W.S.)

Mean High Water N eaps (M.H.W.N.)

:y;y;y;y$ _

Australian Height Datum '••.(Ä.H.D.j-.A. Mean S e a Level (M.S.L.)

'• Mean Low Water N eaps (M.L.W.N.)

'•‘^ • S ® Mean Low Water Springs (M.L.W.S.)

Datum of Predictions ■Aa'Ä Low Water Datum (L.W.D.)

Port Datum Lowest Astronomical Tide. (L.A.T.)

■mm

Figure 2.1 : A guide to tidal planes and their relationship to A ustralian H eight Datum. (From 1992 Tidal Notes, Q ueensland M inistry of Transport).

2.2.1 Microatolls

[image:28.545.101.476.275.551.2]
(29)

a)

[image:29.559.12.542.16.806.2]

b)

(30)

In ad d itio n , they described the various form s w hich can resu lt from the complex histories possible during the lifetime of a coral colony. It m ust be rem em bered th at the changes in "relative sea-level" that a coral experiences on a reef flat are not necessarily those experienced by the region as a whole. For instance, the form ation of rubble ram parts can cause w ater to be m oated on the reef flat at low tide, allowing m icroatolls to grow w ith an artificially high surface. M oat-form ing ram parts m ay be created and destroyed du rin g the lifetime of a single coral, producing complex forms on its surface (Scoffin and Stoddart, 1978). Some of these forms are illustrated in Figure 2.3. Since w e are fairly sure th a t relative sea-level does not change this rapidly on a regional scale in tectonically stable areas, complex m icroatoll form s m ay be an indication of form er m oating, and these colonies should be used carefully to infer form er sea-levels. In other circum stances, m oats m ay last long enough to form m icroatolls w hich are indistinguishable from open-w ater examples, and these w ould provide incorrect estim ates of sea-level change.

surface

z i ...

[image:30.559.91.474.295.691.2]
(31)

M oats m ay form on the reef flat w hen coral rubble accum ulates there, throw n u p from the front of the reef during storm s. For this to occur, the reef flat m ust be relatively wide, and there m ust be a significant am ount of coral grow th nearby as source m aterial for the rubble. These conditions were unlikely to be fulfilled w hen sea-level first reached its present level, as the reefs were m ainly grow ing upw ards, to keep up with the rapid sea-level rise. M icroatolls of this age are therefore likely to be reliable sea-level indicators. A fter a few th o u san d years, how ever, the reef flat becom es sufficiently m ature to accum ulate rubble, as show n by the ages obtained from cem ented rubble terraces on reef islands. Sea-level reached its present level around 6 ka, yet the terrace m aterial d ated by M cLean and others (1978) form ed predom inantly betw een 4 ka and 2 ka. M icroatolls of this age and younger are therefore m ore likely to have been m oated w hen they occur on broad reef flats.

Because they form at sea-level, the sea-level record from fossil m icroatolls exposed on reef flats only covers the time since the m id Holocene, w hen sea- level has been relatively stable. M icroatolls w ould be difficult to identify in drill core, and it w ould be im possible to determ ine w hether m oating had occurred on the reef flat w hen they form ed, so the record is unlikely to extend further back than it does at present. In any case, the period of rapid sea-level rise m ay have p rev en ted the form ation of reef flats on w hich m icroatolls could have formed. The last time sea-level was stable w as at the last glacial m axim um , 18-20 ka, and preserved in-situ m icroatolls of this age w ould provide a precise estim ate of sea-level at that time. Such sam ples are probably at d epths of 120-130 m, and w ould alm ost certainly have to be collected using a subm ersible vehicle.

2.2.2 Reef core data

Bore holes drilled th ro u g h reef flats can p rovide inform ation about the n atu re of reef grow th at that point. U nfortunately, the rate of u p w ard reef grow th is alm ost alw ays lower than the rate of sea-level rise through m ost of the postglacial transgression, so the reef grow th curve is n o t necessarily a sea-level curve (H opley 1986b). No species on the GBR can be used as a "zone coral" such as Acropora palmata in the Caribbean, w hich now grows w ith in a precise d ep th range on the reef crest (Lighty and others 1982). N evertheless, know ing th at coral grow th is restricted to less th an ~30 m w a ter d e p th , w ith en o u g h borehole data it is possible to construct an envelope of sea-level change (eg G rindrod and Rhodes, 1984). Problem s associated w ith this type of record include uncertainty that any coral sam ple from a core is in g ro w th p o sitio n , and the difficulty of ex tractin g a stratigraphically correct core w hen the reef flat is composed of large am ounts of coral rubble.

(32)

small (Hopley 1986b). The advantage of this type of record, com pared to the use of microatolls, is that it can be extended m uch further back in time. The d ep th to the Pleistocene surface ranges from sea-level in Torres Strait to in excess of 30 m in the central GBR (H opley 1986b), so it is theoretically possible to obtain a sea-level record over this depth range.

2.2.3 Beachrock

The term "beachrock" has been used to describe a w ide range of lithified m aterials in the littoral zone, b u t have been m ore rigidly defined as "beach sed im en ts found w ith in the in tertid al zone, and cem ented by calcium carbonate" (Hopley 1986a). If beachrock is to be used as a paleo-sea-level indicator, th en we m u st know the elevation range in w hich it p resen tly form s. A conservative estim ate of sea-level change w o u ld th en be the difference in elevation betw een the beachrock and the m axim um elevation at w hich it presently forms. H opley (1986a) supports the argum ent for the form ation of aragonite cem ent by inorganic precipitation from sea-w ater, w hich suggests that the upperm ost level of cem entation w ould lie betw een m ean high w ater spring tides and the highest astronom ical tide. H ow ever, cem entation above this level is also possible if the water table is tem porarily raised by storm conditions, for exam ple. C em entation can occur over periods of time ranging from m onths to decades (Hopley 1986a), so the range of possible form ation elevations is large. M cLean and o th ers (1978) exam ined beachrock deposits throughout the northern GBR, and concluded that there w as no positive evidence of sea-level change.

In ad d itio n to the uncertainty regarding the elevation at w hich it form ed, the precise dating of beachrock is a difficult problem . If unrecrystallised fragm ents of the clast m aterial can be obtained, then a m axim um age for the deposit can be ascertained, although the time betw een grow th of the clast and its incorporation into the beachrock may range from years to thousands of years (H opley 1986a). If it can be isolated, the radiocarbon age of the cem ent m ay provide the age of cem entation, although recrystallisation is possible and undetectable. H opley (1971) reports that at H erald Island, near Tow nsville, a clast consisting of a shell of Trochus sp. gave an age of 4280±100 BP, w hile a w hole-rock sam ple of fin e-g rain ed b each ro ck , containing up to 50% cement, w as dated at only 3540±90 BP.

D espite the relatively large uncertainties in determ ining the age of a sam ple and its relationship to sea-level at that time, beachrock m ay prove to be a useful indicator of sea-level at the Last Glacial M axim um (LGM). Few sam ples from this time have been collected, and the uncertainty in LGM sea- level is greater than the uncertainty in the form ation range of beachrock.

2.2.4 Chenier plains

(33)

tid a l m udflats" (Otvos and Price 1979). This im plies fo rm atio n in an e n v iro n m e n t w h ere m u d fla t p ro g ra d a tio n is in te rru p te d by ch en ier form ation. C heniers are distinct from "true" beach ridges in that they are shallow -rooted and rest on m udflat deposits, while the latter are based on shoreface deposits and represent the top of a laterally extensive beach deposit. This difference makes cheniers m uch m ore useful for interpreting paleo-sea-levels, because the base of a chenier forms in the intertidal zone, w hile other beach ridge deposits m ay have no clearly defined base. The elevation of the crest of a chenier or beach ridge is less ap p ro p riate as an indicator of sea-level change, because it is initially controlled by the wave energy at the time of form ation, w hich is unrelated to sea-level, and m ay be altered later by erosion.

C heniers m ay form by a variety of processes, which involve interruption of the p ro g rad atio n of the m udflat (A ugustinus 1989). The coarse sedim ent m ay develop by w av e-w in n o w in g and su b seq u en t concentration in an offshore b ar, or arrive by longshore tran sp o rt of a shallow -based spit. C om binations of both usually occur (A ugustinus 1989). After form ation, the ridge m ay m igrate landw ards due to continuing or episodic rew orking, until it is bey o n d the range of storm w aves and sp rin g tides. C o n tin u in g progradation of the m udflat seaw ard of the ridge m ay also halt its landw ard m ig ratio n .

C henier p lain s preserve a less reliable record of sea-level change than m icroatolls do, as they are formed w ithin a w ider sea-level range, and are affected by slow sedim ent compaction. Smart (1976) noted th at some chenier ridges ap p ear to have m ost of their volum e above sea-level, w hile others have a considerable volum e below it. This could be caused by variable com paction of the u n d erly in g sedim ent. N evertheless, they are useful because th ey occur on the continental coast w here the m id-H olocene h ig h stan d w as the greatest, and w here corals cannot grow because of the abundance of terrigenous sedim ent. Example of such sites are K arum ba, in the southeast Gulf of C arpentaria (Rhodes 1980), and Princess C harlotte Bay, described by Chappell and G rindrod (1984).

2.2.5 Oyster shells

(34)

E stim ating the form er sea-level represented by an oyster shell d ep o sit is subject to several errors. The nom inal grow th range of the oyster Saccostrea cited above represents a vertical range of 0.2 - 0.3 m above MSL in the GBR region (Q ueensland M inistry of T ransport 1992). H ow ever, Beam an and others (1994) reported the upper limit of the m odern oyster bed on M agnetic Island, near Townsville, to be 0.4 m above AHD (A ustralian H eight D atum , w hich in the Townsville area is equivalent to MSL). O ther m easurem ents, presented later in this chapter, show the limit to be up to +0.8 m MSL on exposed rocks at W est Cliff Island, w hich places it betw een the levels of MHWS and M HW N, although this m easurem ent d epends on the validity of correlating the tidal range from Flinders Island. Furtherm ore, on Lizard Island the u p p er lim it of m odern oyster grow th w as seen to range from +0.46 m MSL on exposed rocks to +0.53 m MSL at a sheltered site in the sam e bay. This suggests that fossil oyster beds, which are usually found in protected cracks and caves, grew above the range of contem porary oysters in exposed sites and w ould thus indicate a larger sea-level change than had in fact occurred. The uncertainty in estim ated sea-level change u sin g fossil oyster beds is thus around ±0.35 m, neglecting any uncertainty in surveying the level of the sample. This is considerably larger than the 0.14 m im plied by Beaman and others (1994).

Problem s m ay also be encountered in dating, as the finely-foliated n atu re of the shells m eans that clean surfaces rapidly absorb CO2 containing m o d ern carbon from the atm osphere during sam ple processing. This can be avoided eith er by p rep arin g sam ples u n d e r a nitrogen atm osphere, or rem oving surface m aterial w ith acid imm ediately before dating.

2.2.6 Other methods

(35)

2.3

Previous work

This section review s previous w ork on late Pleistocene and H olocene sea- level change in no rth Q ueensland. It is presented in three sections, each of w hich have a different significance for glacio-hydro-isostatic m odelling. First, I review observations of sea-level at the last glacial m axim um (LGM). The m inim um sea-level at this tim e constrains the total volum e of ice that m ust have existed on land. Second, the period of rapid sea-level rise until ~6 ka. Evidence of either steady or episodic sea-level rise d u rin g this period has direct im plications for the n atu re of the disintegration of the large ice sheets, w hich is not w ell understood. Third, sea-level change in the m id- late H olocene. Precise sea-level observations are m ost ab u n d an t in this period, and the spatial and tem poral variation supplies inform ation on the rheology of the earth and recent eustatic sea-level change.

2.3.1 LGM sea-level

G lacio-hydro-isostatic m odelling of sea-level change is ham pered by the fact that the total am ount of w ater taken from the oceans to form the ice sheets of the LGM is only approxim ately known. A t present, different w orkers use values w hich are consistent w ith their ow n ice sheet reconstructions and sea-level observations, b u t too few observations are available from LGM tim es to constrain the total ice volum e tightly. The m odel of N akada and Lambeck (1988) contains a total ice volume of 129.3 m equivalent sea-level, w hile th at of Peltier (1994) uses 105.2 m. It is im portant to establish the actual v alu e, because the figure is not only u sed in global isostatic calculations, b u t also in com parisons of sea-level and isotope records (C happell and Shackleton 1986, Shackleton 1987) on w hich m any paleo- climate studies are based.

Sea-level at the LGM varies spatially, because of the effects of isostasy as well as local tectonism . This variation will be exam ined in C hapter 6. Here, we will review published observations of LGM sea-level from Q ueensland and n o rth ern A ustralia.

Veeh and Veevers (1970) dated a sam ple of beach rock from -150 m and a shallow w ater coral, Galaxea clavus, from a terrace at -175 m off One Tree Island in the Capricorn Group. Radiocarbon dates of 13 860 ± 220 and 13 600 ± 220 w ere obtained for the sam ples. A U-Th age of 17 000 ± 1000 w as also obtained for the coral. As the coral can grow to at least 25 m d ep th and occasionally as deep as 75 m (Veeh and Veevers 1970), these sam ples could represent a low stand of 150 m, w ith the coral grow ing 25 m below sea-level. Veeh and Veevers, how ever, consider it m ore likely that the coral grew in shallow w ater, and th at sea-level stood near -175 m som etim e betw een 13 600 and 17 000 years ago.

(36)

BP. Jongsm a (1970) reported beachrock observed from a subm arine in the A rafura Sea, and a sam ple dredged from about the same depth gave an age of 18 700 ± 350 BP.

C arter and Johnson (1986) used PDR (precision depth recorder) profiles to identify several bathym etric steps on the continental slope in the central GBR. A prom inent terrace at -133 m has a sedim ent cover up to 10 m thick, w hich w edges out at a slightly eroded edge around -114 m. No dates are available for the sedim ent, and the LGM shoreline could be either the -114 m notch or the -133 m terrace. C arter and Johnson also consider the possibility th at a higher terrace, at -103 m, could m ark the LGM sea-level m in im u m .

2.3.2 The postglacial transgression

On a general scale, the rate at w hich the LGM ice sheets m elted is know n from oxygen isotope records and far field sea-level observations. The m ost rapid period of sea-level rise w as betw een 14 ka and 8 ka, w ith an average rate of about 15 m m /y r. H ow ever, the question of w hether sea-level rise w as continuous or episodic remains.

Recent evidence from G reenland ice cores (Alley and others 1993, D ansgaard and others 1993, Taylor and others 1993) and N orth Atlantic sedim ent cores (Bond and others 1992, Lehm an and Keigw in 1992) suggest th at the last deglaciation involved su d d en reorganisation of the global clim ate system . Collapse events in the L aurentide and A ntarctic ice sheets are am ong the processes w hich are inferred to have occurred (Blanchon and Shaw 1995). These w ould cause episodic eustatic sea-level rise, the effects of which w ould vary spatially, due to isostatic effects. In the sea-level record from Barbados, Blanchon and Shaw (1995) propose three periods of "catastrophic sea-level rise" w hich they associate w ith ice sh eet collapse events: tw o in the L aurentide and one in the Antarctic. In this section, I review the sea-level record for the sam e period of tim e in N o rth Q ueensland. C om parison betw een the predicted and observed effects of episodic sea-level rise will be m ade in C hapter 6.

Maxwell (1968, p61) interpreted w idespread bathym etric features of constant d e p th as form er stra n d lin es, and su g g e ste d a sequence of sea-level transgressions and Stillstands w hich could have form ed them. He proposed Stillstands in the postglacial sea-level rise at 102 m, 88 m, 66 m and 29 m, follow ed by a fall to 59 m (depths ro u n d ed to the nearest m etre). Further rise was punctuated by Stillstands at 37 m and 18 m.

(37)

m aterial in the associated sedim ent bodies, and the rem ainder were assigned ages by com parison w ith the sea-level curve from N ew Z ealand of Gibb (1985). Since sea-level varies spatially, the N ew Zealand record cannot be considered to be a eustatic curve, and dates obtained in this w ay are dubious unless corrected for glacio-hydro-isostatic effects. W here terrace-covering sedim ents can be dated, they provide a lower limit for sea-level at that time. C arter and Johnson (1986) interpreted the sequence as the result of a history of alternating Stillstands and rapid sea-level rise. In a m ore localised study of C leveland Bay, offshore from Townsville, Carter and others (1993) found evidence "consistent w ith pauses or slow dow ns in the post-glacial sea-level rise" c o rre sp o n d in g w ith sh o relin es S2 an d S3, a lth o u g h the featu re corresponding w ith S2 was found at -10 m.

Table 2.1: Depths and ages of drowned shorelines in the central GBR. From Carter and Johnson (1986).

sh o relin e age (ka) depth (m)

S8 18 114

S7 17 88

S6 15 75

S5 12 56

S4 11 45

S4a 10 39

S3 9.5 28

S3a 9 23

S2 7.5 9

SI 6.5 0

Of these shorelines, S7 and S3 correlate w ith M axwell's observations, and S4a is w ith in 2 m of another, w hich is the stated erro r for C arter and Johnson's depths. Also, the 103 m shoreline m entioned b y C arter and Johnson (1986), b u t not included in their list of transgressive shorelines, m atches closely w ith the 102 level of Maxwell (1968). A part from these, the terrace levels do not correspond.

W hile these shorelines m ay have been form ed during the post-glacial sea- level tran sg ressio n , the conclusion that they represent Stillstands is not necessarily w arran ted , because processes other than sea-level influence the distribution of coastal sedim ent. Lateral m igration of sedim ent distribution channels, for exam ple leaves a sedim ent record at any given point w hich could be in terp reted as indicative of episodic sedim entation. One of the m ost pow erful arg u m en ts in favour of Stillstands is th at the shorelines ap p aren tly p ersist laterally at roughly constant d e p th over w ide areas, although this is not alw ays the case and could also be explained by clim ate changes w hich affect the supply of sedim ent throughout the region.

(38)

sw am p environm ents. Salam a (1990) recovered m angrove "peats" and unspecified shells from several sedim ent cores taken in Princess C harlotte Bay. Radiocarbon ages of these sam ples are sum m arised in Table 2.2 and Figure 2.4. A part from two m angrove sam ples, the data form a reasonable trend for sea-level rise betw een -40 and -10 m. There is not sufficient sam ple resolution to determ ine w hether the rise w as continuous or episodic, b u t it is w orth noting th at the five peat sam ples betw een 23.3 m and 24.5 m form a p lateau, in agreem ent w ith the 23 m shoreline (S3a) of C arter an d Johnson (1986).

Table 2.2: Radiocarbon ages of samples from sedim ent cores in Princess Charlotte Bay, from Salama (1990). The m arine reservoir correction applied to the dates from shells is -450±35 yr (Gillespie and Polach, 1979). The tw o samples older than 20 000 years are considered to be rew orked, and are not included in Figure 2.4.

field code S lat. E long. lab. num ber d e p th

below MSL (m)

conventional age (yr BP)

corrected age (yr BP)

sam ple

8/1/220 14.10 144.34 SAU 2373 33.9 31100 ± 800 30650 ± 800 sh e lls

15/1/125 14.20 144.30 SAU 2374 22.7 29900 ± 600 29450 ± 601 sh e lls

18/2/240 14.12 144.08 SAU 2375 31.6 9190 ± 140 8740 ± 144 sh e lls

28/3/287 14.11 144.13 BETA 5814 36.8 8810 ± 100 8810 ± 100 p e a t

29/2/180 14.18 144.15 BETA 9283 23.9 8570 ± 140 8570 ± 140 p e a t

29/2/205 14.18 144.15 BETA 9284 24.2 8980 ± 200 8980 ± 200 p e a t

29/2/240 14.18 144.15 SAU 2375 24.5 8640 ± 120 8640 ± 120 p e a t

36/3/224 14.21 144.02 BETA 5814 16.7 8280 ± 70 8280 ± 70 p e a t

41/2/170 14.42 144.06 BETA 5816 10.1 7220 ± 120 7220 ± 120 p e a t

49/1/355 14.42 143.98 SAU 2377 9.0 6730 ± 180 6280 ± 183 crab

55/2/75 14.16 144.06 BETA 9285 24.1 8570 ± 140 8570 ± 140 p e a t

57/1/320 13.91 143.80 BETA 21804 23.3 8800 ± 90 8800 ± 90 p e a t

125/1/60 14.14 143.90 SAU 2378 18.4 8100 ± 140 8100 ± 140 p e a t

134/1/113 14.29 143.89 BETA 22248 12.8 8760 ± 140 8760 ± 140 p e a t

220/3/325 14.09 144.09 BETA 6648 27.8 9020 ±90 9020 ± 90 p e a t

224/1/110A 14.07 144.18 SAU 2379 33.2 8920 ± 70 8470 ± 78 sh e lls

224/1/HOB 14.07 144.18 SAU 2380 33.2 9320 ± 120 9320 ± 120 p e a t

228/1/97 14.04 144.18 SAU 2381 42.0 9510 ± 120 9060 ± 125 sh e lls

241/1/140 13.99 144.98 SAU 2382 31.0 9360 ± 150 8910 ± 154 sh e lls

[image:38.559.79.508.265.540.2]

Figure

Figure 2.1 : A guide to tidal planes and their relationship to Australian Height Datum
Figure 2.2 : a) Living, and b) fossil microatolls at Iris Point, Orpheus Island, on the Great Barrier Reef
Figure 2.3 : The response of microatolls to different patterns of sea-level change. From Woodroffe and McLean (1990).
Table 2.2: Radiocarbon ages of samples from sediment cores in Princess Charlotte Bay, from Salama (1990)
+7

References

Related documents

• Follow up with your employer each reporting period to ensure your hours are reported on a regular basis?. • Discuss your progress with

Nebraska’s efforts towards integrated management have the po- tential to support more adaptive approaches to water resources man- agement and could serve as a guidepost for

Only regular and honorary members of the Corporation in good standing who have demonstrated active participation in the Corporation and who meet the established qualifications

The corona radiata consists of one or more layers of follicular cells that surround the zona pellucida, the polar body, and the secondary oocyte.. The corona radiata is dispersed

Proposed Framework Form a project team Communicat e objectives to all concerned Confirm the value drivers Define Objective s. Based on the results of data analysis, the

Using text mining of first-opinion electronic medical records from seven veterinary practices around the UK, Kaplan-Meier and Cox proportional hazard modelling, we were able to

This conclusion is further supported by the following observations: (i) constitutive expression of stdE and stdF in a Dam + background represses SPI-1 expression (Figure 5); (ii)

4.1 The Select Committee is asked to consider the proposed development of the Customer Service Function, the recommended service delivery option and the investment required8. It