PP22A-01 INVITED
PRISM Sea Surface Temperatures: Overview and new PRISM3 Low Latitude Pacific SSTs.
The Pliocene Research, Interpretation and Synoptic Mapping (PRISM) data set is a comprehensive paleoenvironmental reconstruction for the period ~3.3-3.0 Ma, now used in paleoclimate- model comparisons of periods of global warmth. PRISM data have been used by GISS and UKMO modeling groups interested in the mid-Pliocene as a possible analog for future climate change. Initially developed in the 1990's, the PRISM reconstruction has undergone major revision in the past two years. New data are available for almost every component of the climate system (sea-surface temperature [SST], sea-ice distribution, land-ice configuration, vegetation and land cover, sea-level and topography) which are being integrated into the PRISM3 reconstruction. Deep-sea bottom temperature data are still being developed. This presentation will summarize the development of the PRISM SST data set and focus on the new low latitude Pacific Ocean SST configuration based upon mutiproxy data and modeling efforts.
PP22A-02
Modelling high latitude climates and ice sheets during the mid-Pliocene warm period
Reduction in the polar ice caps and associated climate feedbacks are implicated in the warming of pre- Quaternary palaeoclimates. General Circulation Model (GCM) simulations of the last such warm period, the mid- Pliocene (3.29-2.97 Ma), suggests global surface temperatures were between 1.4°C and 3.6°C warmer than today. However, the changes are amplified in the high latitudes, where the PRISM (Pliocene Research, Interpretation and Synoptic Mapping) palaeoenvironmental reconstruction specifies a 50% reduction in the Greenland Ice Sheet (GrIS) and a 33% reduction in the Antarctic Ice Sheet. These ice sheets configurations are largely based on poorly constrained sea level estimates and are one of the least well-known boundary conditions for the mid-Pliocene. Utilizing a suite of mid-Pliocene GCM experiments, evaluated against available palaeoenvironmental information, and a 3-D thermomechanically coupled ice sheet model, the state of the GrIS and East Antarctic Ice Sheet (EAIS) during this interval has been reconstructed. Ensemble models of Greenland, which compare favourably to evidence of ice-rafted debris and mid-Pliocene vegetation, suggest a reduction in the ice sheet to 30 - 40% of the modern ice volume. In East Antarctica increased surface temperatures during the mid-Pliocene lead to significant melt over the Wilkes and Aurora Subglacial Basins and a reduction in the extent of the ice sheet. These ice losses are partially offset by an increase in snowfall over the Antarctic plateau. Marine diatoms in the Transantarctic Mountains have been used as evidence of major East Antarctic deglaciations during the Pliocene. However, our EAIS predictions show that the modelled mid-Pliocene climate is insufficient to cause the hypothesized magnitude of ice sheet retreat. Finally, the mid-Pliocene has been suggested as a possible palaeoclimate analogue for the climate of the late 21st century. Here we compare predictions of mid-Pliocene ice sheets with observations of modern high latitude cryospheric change and discuss the implications for future ice sheet and climate stability of the high latitudes.
PP22A-03 INVITED
CO2 Climate-Glaciation linkages During the Late Paleozoic Ice Age and the Earth's Penultimate Deglaciation
The Late Paleozoic Ice Age (LPIA) was the longest-lived (330 to 260 Myr) and most intense glaciation of the past half-billion years. Emerging high-latitude Southern Hemisphere records document a much more dynamic ice age – one defined by multiple short-lived (1 to 7 myr duration) icehouse periods punctuated by warmer periods of glacial minima. These major climate shifts throughout the LPIA and its demise at the close of the Early Permian provide the only ‘vegetated-Earth' analogues of major climate change in an icehouse. As our climate system departs from the well-studied Pleistocene glacial-interglacial cycles, a ‘deep-time' perspective of pCO2-climate- glaciation linkages during past icehouse-to-greenhouse transitions provides a unique perspective into what may be the Earth's most epic deglaciation. Here we apply the carbon isotopic compositions of soil-formed carbonates and fossil plant material (cuticle, coals, charcoals) from several terrestrial basins in North America to a soil CO2-diffusion model and Monte Carlo modeling to estimate atmospheric pCO2 for the LPIA and its transition to the ensuing Mesozoic greenhouse state. Best estimates of Late Paleozoic pCO2 indicate repeated shifts from present-day levels to values of up to 2500 to 3000 ppmv during periods of glacial minima and possibly fully deglaciated greenhouse states. To evaluate the nature of the CO2-climate relationship during these major climate transitions, we developed a time-equivalent record of paleotropical sea-surface temperatures (SSTs) using δ18O values from a global compilation of well-preserved latest Permo-Carboniferous tropical shallow-water brachiopods. The observed covariance between shifts in inferred paleotropical SSTs, pCO2 and high- latitude Gondwanan glaciation implies a strong CO2-climate-glaciation linkage that is consistent with the range predicted by Permian climate simulations for a change in radiative CO2-forcing from 1 to 8 fold present-day levels. This apparent CO2-climate-glaciation link suggests that atmospheric CO2 levels may have been the primary driver for the repeated buildup and retreat of continental ice sheets during the Late Paleozoic. Integration of these climate proxy records with newly developed tropical paleobotanical records for paleotropical Euramerica reveals repeated major restructuring of flora in-step with climate and pCO2 shifts illustrating the impact on tropical floral ecosystems associated with past CO2-forced climate transitions.
PP22A-04
Quantifying Ice Volume and Temperature Change for the Greenhouse to Icehouse Transition: A Coupled Palaeoceanographic and Palaeoclimate Modelling Approach
The abrupt and widespread glaciation of Antarctica in the earliest Oligocene marked a fundamental change in global climate leading to the Earth's current glaciated state. An increase in benthic δ18O of up to 1.5 ‰ occurred over a 300-400 kyr interval and is widely assumed to document both cooling and ice sheet growth marking the inception of the icehouse world. Resolving the relative contribution of ice volume and temperature changes to this shift is essential to understanding, and accurately modelling, this climate transition. In an attempt to quantify relative ice volume and temperature changes at the Eocene-Oligocene boundary we present the initial results from a coupled paleoceanographic and paleoclimate modelling approach. Coupled δ18O and Mg/Ca records of surface dwelling Turborotalia ampliapertura and thermocline dwelling Subbotina angiporoides have the potential to document upper ocean temperature and δ18O seawater at ODP site 1263, Walvis Ridge in the South Atlantic. %CaCO3 measurements from the suite of sites drilled on ODP Leg 208 place the lyscoline at ~3.8 km in the latest Eocene, prior to deepening in the earliest Oligocene. Collected at a present day water depth of 2717 m, ODP Site 1263 was positioned above the lysocline throughout the Eocene-Oligocene transition. Carbonate concentrations vary between 84 and 96 % (within the studied interval) and planktonic foraminifera appear well preserved. Spanning 33.8 to 32.8 Ma, initial records have a temporal resolution averaging <20 kyrs and compliment an existing benthic δ18O record from the same site with data points every ~6 kyrs. Preliminary records suggest a temperature shift of less than 0.5° C in the Mg/Ca records of both the surface and subsurface-dwelling species. This finding matches that of the simulated response of sea surface temperatures (SST) at the paleolatitude associated with ODP Site 1263 to the growth of an Antarctic ice sheet during the earliest Oligocene using the HadCM3L General Circulation Model. Two Early Oligocene experiments are being performed which are identical in all respects expect in their prescribed Antarctic ice volumes (either zero ice volume or an ice volume close to modern given the Early Oligocene land-sea mask employed). The response of SST to the growth of an Antarctic ice sheet is predicted to be spatially heterogeneous with some areas subject to changes >3° C, yet a temperature change <0.5° C at ODP Site 1263 is suggested in both the modelling and proxy data presented here. Furthermore, modelling results suggest that intermediate and deep waters cooled by not more than 0.5° C in open ocean areas, indicating that the benthic δ18O shift at the Eocene-Oligocene boundary cannot be accounted for by glaciation of Antarctica alone. Further work will include the incorporation of a Northern Hemisphere ice sheet in an additional model run to assess the response of SST and BWT to bipolar glaciation and the production of geographically diverse SST and BWT records to compare with model simulations.
PP22A-05 INVITED
Pole-to-pole sea surface temperatures from the Paleocene-Eocene thermal maximum using organic and inorganic paleothermometers: the ultimate test case for climate models
The Paleocene-Eocene Thermal Maximum was a short-lived (about 170 kyr) episode of globally elevated temperatures, superimposed on already warm late Paleocene – early Eocene greenhouse climates. Recent application of the relatively new organic paleothermometer TEX86 as well as oxygen isotope analyses on well-preserved foraminifera on marginal marine PETM sections has allowed direct comparison of absolute sea surface temperature (SST) reconstructions by independent proxies. These records show relatively good correspondence, suggesting they are both suitable to reconstruct late Paleocene and early Eocene absolute SSTs. Both low- and high-latitude temperatures were (much) warmer than at present, consistent with biogeographical patterns, but the difference between high- and low latitude temperatures was extremely low. Current generation fully coupled climate models fed with early Paleogene boundary conditions indeed predict warming on a global scale with higher CO2 concentrations. However, meridional temperature gradients remain structurally overestimated in the models, implying that important feedbacks during greenhouse climates are not (correctly) implemented in the climate models. Moreover, temperatures during the PETM, rose by 5 to 8 ° C on a global scale (in the absence of ice-albedo feedbacks), thus not showing significant polar amplification. This suggests that the mechanism that caused the reduced meridional temperature gradient was not amplified during the PETM.
PP22A-06 INVITED
Modeling the Climates of the Mid and Latest Permian
Global climate system models are the most comprehensive tools for understanding Earth's climate and changes to the system through time. These models are becoming more complex as chemistry, biogeochemistry and vegetation dynamics are coupled to the physical climate system. Model simulations of the Mid and Latest Permian are compared to paleoclimate data to show how models and data can be used in conjunction to better understand Earth's deep climates. We also consider the implications of rapid change in the oceans at the Latest Permian, and the implications of this rapid change for Earth's future climate. A high-resolution (70km) atmospheric simulation of the Latest Permian will also be presented to show the importance of increased resolution on modeling paleoclimates. We also describe a new opportunity to carry out deep climate simulations for the paleoclimate community.
PP22A-07
Reconstructing Neoproterozoic Paleoclimates Using a Combined Data/Modeling Approach
Climate reconstructions of the Neoproterozoic Era (1,000-542 Ma) face special challenges since many proxies used to constrain younger paleoclimates are not available or applicable in Precambrian time. Given the few available proxies, such deep time climate simulations are best viewed as a means to address more fundamental questions about the nature of climate change, and to address disparities in data interpretation by examining phenomena from a process-related perspective. The Global Climate Model (GCM) simulations presented here were aimed at determining what combination of climate forcings might have permitted the initiation of low- to mid-latitude continental ice sheets during a 'snowball Earth' glacial interval. In addition to natural forcings such as greenhouse gases and solar luminosity, models often require certain input parameters for which no paleoclimate proxies exist. For example, in our experiments the average thickness of the ocean mixed layer is specified. The degrees of freedom inherent in these simulations leave open the possibility that model output might reflect a physically plausible but geologically improbable climatic state, so particular care is needed in validating model results with available data. Analysis of our GCM output suggests that the addition of topographic relief and dynamic ice flow to our simulations would more readily permit the growth and persistence of ice sheets in subtropical continental regions. However, these simulations also suggest that `hard' snowball Earth solutions are only likely for much earlier intervals in Earth history.
PP22A-08
Model-Data Comparisons for Past Warm Climates: How Do We Know When We've Got It Right?
In order to use paleoclimate reconstructions to evaluate the quality of climate system models, we have to assume at some point that our interpretations of climate proxy data are correct. Without that, there is no argument to be made that a model's sensitivity to some system change is right or wrong. (We also must assume that we have identified and reconstructed properly every relevant system boundary condition and forcing, but this is far from the current state of the art.) We generally can believe that the analytical precision with which we measure a sample property is adequate (isotope, metal or compound ratios, magnetism, component weight percentage, etc.). But we also have to believe that a primary environmental signal is preserved in a sample (i.e. that there has been no diagenetic alteration of the quantity or quality being measured). And, most difficult by far, we have to believe that we understand properly how to translate the measurement that is made into the value of the original environmental variable (ocean temperature, atmospheric composition, percent deciduous forest, continental position and elevation, etc.) Addressing the vital issue of climate model sensivity to atmospheric carbon dioxide concentration, we examine the results from coupled and uncoupled model sensitivity experiments in comparison to multiple proxy reconstructions of carbon dioxide and surface temperatures for past warm climate intervals during the Cretaceous and Eocene. The comparison, like all paleoclimate model-data comparisons, forces the question, "How do we know when we've got it right?"