<p>The mid- to late Pliocene (3.3-2.6 Ma) spans one of the most significant climatic transitions of the Cenozoic. It is characterised by global cooling from a climate with an atmospheric CO2 concentration of ~400 ppm and temperatures of 2-3°C warmer-than-present, to one marked by the progressive expansion of ice sheets on northern hemisphere. Consequently, the mid-Pliocene warm period (MPWP; 3.3-3.0 Ma) provides the most accessible and recent geological analogue for global sea-level variability relevant to future warming. Global mean sea level has been estimated at 22 ± 10 m above present-day for MPWP. However, recent re-evaluations of this estimate suggest that spatially-varying visco-elastic responses of the crust, local gravitational changes and dynamic topography from mantle processes may preclude ever being able to reconstruct peak Pliocene mean sea level. The Whanganui Basin, New Zealand, contains a ~5 km thick stratigraphic succession of Pliocene-Pleistocene (last 5 Ma), shallow-marine, cyclical sedimentary sequences demonstrated to record orbitally-paced, glacial-interglacial global sea-level fluctuations. A limitation of the Whanganui sea level record, to date, has been an inability to resolve the full amplitude of glacial-interglacial water depth change due to the occurrence of cycle bounding unconformities representing sub-aerial erosion during glacial lowstands. This thesis analyses a new ~900 m-thick, mid- (3.3-3.0 Ma) to late Pliocene (3.0-2.6 Ma), shallow-marine, cyclical sedimentary succession from a remote and relatively understudied part of Whanganui Basin. Unlike previous studies, these shelf sediments were continuously deposited, and were not eroded during sea-level lowstands, and thus provide the potential to reconstruct the full amplitude of glacial-interglacial sea-level change. On orbital timescales the influence of mantle dynamic processes is minimal. The approach taken applies lithofacies, sequence stratigraphy, and benthic foraminiferal analyses and a novel depth-dependent sediment grain size method to reconstruct the paleowater depths for, two continuously-cored drill holes, which are integrated with studies of outcropping sections. The thesis presents a new record of the amplitude and frequency of orbitally-paced, global sea-level changes from a wave-graded continental shelf, that is independent of the benthic δ¹⁸O proxy record of global ice-volume change. Paleobathymetric interpretations are underpinned by analysis of extant benthic foraminiferal census data and a statistical correlation with the distribution of modern taxa. In general, water depths derived from foraminiferal modern analogue technique are consistent with variability recorded by lithofacies. The inferred sea-level cycles co-vary with a qualitative climate record reconstructed from a census of extant pollen and spores, and a modern temperature relationship. A high-resolution age model is established using magnetostratigraphy constrained by biostratigraphy, and the dating and correlation of tephra. This integrated chronostratigraphy allows the recognition of 23 individual sedimentary cycles, that are correlated “one-to-one” across the paleo-shelf and are compared to the deep-ocean benthic oxygen isotope (δ ¹⁸O) record. A grain size-water depth technique was developed to quantify the paleobathymetry with more precision than the relatively insensitive benthic foraminifera approach. The method utilises a water depth threshold relationship between wave-induced near bed velocity and the velocity required to transport sand. The resulting paleobathymetric records of the most sensitive sites, the mid-Pliocene Siberia-1 drill core and the late Pliocene Rangitikei River section, were selected to compile a composite paleobathymetry. A one-dimensional backstripping method was then applied to remove the effects of tectonic subsidence, sediment and water loading on the record, to derive a relative sea level (RSL) curve. The contribution of glacio-hydro-isostatic (GIA) processes to the RSL record was evaluated using a process-based forward numerical solid Earth model for a range of plausible meltwater scenarios. The Whanganui Basin RSL record approximates eustatic sea level (ESL) in all scenarios when variability is dominated by Antarctic Ice Sheet meltwater source during the mid-Pliocene, but overestimates ESL once Northern Hemisphere ice sheet variability dominates in the late Pliocene. The RSL record displays 20 kyr precession-paced sea level variability during the MPWP with an average amplitude of ~15 ± 8 m, in-phase with southern high-latitude summer insolation. These are interpreted as ~20 m Antarctic Ice Sheet contributions, offset by ~ 5 m anti-phased Greenland Ice Sheet contribution, in the absence of a significant Northern Hemisphere ice sheets. This interpretation is supported by a previously published ice-proximal precession-paced, ice-berg-rafted debris record recovered off the coast of Wilkes Land. The Whanganui RSL record is not consistent with a dominant 40 kyr pacing observed the benthic oxygen isotope stack at this time. While the deep ocean benthic δ¹⁸O stack is of varying temporal and spatial resolution, during this time interval, the Whanganui RSL record implies a more complex relationship between ice-volume and oxygen isotope composition of sea water (δ¹⁸Oseawater). The relative influences of varying composition of the polar ice sheets, marine versus land based ice, the out-of-phase behaviour of polar ice sheet growth and retreat, and a potential decoupling of ocean bottom water temperature and δ¹⁸Oseawater are explored. The late Pliocene relative sea level record exhibits increasing ~40 kyr obliquity-paced amplitudes of ~20 ± 8 m. This is interpreted as a response to the expansion of Northern Hemisphere ice sheets after ~2.9 Ma. During this time the Antarctic proximal ice-berg rafted debris records display continuing precession-paced ice-volume fluctuations, but with decreasing amplitude suggesting cooling and stabilisation of the East Antarctic Ice Sheet. With the bipolar glaciation, the ocean δ¹⁸O signal became increasingly dominated by northern hemisphere ice-volume. However, the RSL record implies relatively limited ice-volume contributions (up to ~25 m sea level equivalent) prior to ~2.6 Ma. The large amplitude contribution of Antarctic Ice Sheets to global sea level during the MPWP has significant implications for the sensitivity of the Antarctica Ice Sheet to global temperatures 2-3°C above preindustrial levels, and atmospheric CO₂ forecast for the coming decades.</p>
Simulating the Antarctic ice sheet in the late-Pliocene warm period: PLISMIP-ANT, an ice-sheet model intercomparison project
