PP51E-01 INVITED
Modelling Pliocene Warmth: Contribution of Atmosphere, Oceans and Cryosphere Revisited
In 2004, Haywood and Valdes published the first fully coupled ocean-atmosphere GCM (OAGCM) study of the mid Pliocene. The work was performed to determine the relative role of the atmosphere, oceans and cryosphere in driving/maintaining elevated global temperatures during the mid Pliocene warm period. The model (the HadCM3 GCM) was initialised with boundary conditions derived from the U.S. Geological Survey PRISM data set and an atmospheric carbon dioxide concentrations of 400 ppmv. The results from the 2004 study were based on a single model run in which all the boundary conditions were changed simultaneously to a mid Pliocene state. This made it difficult to ascertain the exact contribution of each boundary condition change to mid Pliocene warmth. Furthermore, the length of the model integration was insufficient to allow a full equilibrium condition to be reached. Here we present the results from an ensemble of fully coupled OAGCM simulations using the HadCM3 GCM. Our mid Pliocene control run, incorporating a full suite of palaeo-boundary conditions, has now been integrated for more than 850 simulated years. An additional four simulations have been performed which track the principal environmental changes that occurred between the mid Pliocene and pre-industrial (simulation 1. mid Pliocene but with pre-industrial CO2; simulation 2. mid Pliocene with pre-industrial CO2 and modern ice sheets; simulation 3. mid Pliocene with pre-industrial CO2, modern ice sheets and soils but retaining mid Pliocene vegetation in non-glaciated regions; simulation 4. pre-industrial). Results from this ensemble of experiments broadly support the conclusions of the 2004 study indicating that elevated CO2 and the smaller Greenland and Antarctic ice sheet provide the largest contribution to mid Pliocene warmth. Changes in orography, soils and vegetation provide additional but minor contributions. Our control mid Pliocene simulation continues to indicate no increase relative to the pre-industrial in ocean heat transport from the tropics to the high latitudes and no enhancement in thermohaline circulation.
PP51E-02 INVITED
Zonal and Meridional Sea Surface Temperature Gradients and Orbital Variability During the Plio-Pleistocene Transition
As the most recent interval of sustained warmth in Earth's history the Pliocene represents a potential analog for future climate scenarios. Yet, the ultimate causes of both early Pliocene warmth and the subsequent major climatic transition, which resulted in the development of large ice sheets in the Northern Hemisphere, remain unclear. Here, we compare and contrast three orbital-resolution, alkenone-derived sea surface temperature (SST) records for the interval from 4 to 1.4 Ma, shedding new light on the evolution of the ocean surface temperature field during the Plio-Pleistocene Transition. We examine the trends and orbital scale variability as well as the zonal and meridional temperature gradients elucidated by SST data from Ocean Drilling Program Site 846 (3°S, 91°W) in the eastern equatorial Pacific (EEP), Site 982 (58°N, 16°W) in the North Atlantic, and Site 662 (1°S, 12°W) in the Eastern Equatorial Atlantic (EE ATL). Our data indicate that although the rate of cooling in the EEP (1°C/Myr) was nearly twice that in the EE ATL (0.6°C/Myr), the overall structure of the data from these two sites is remarkably similar. This marked similarity in structure suggests the operation of a large-scale forcing mechanism, such as a change in atmospheric CO2 concentrations. The zonal temperature gradient between our two tropical sites grew steadily from <1°C at 4 Ma to ~2°C at 1.4 Ma, perhaps as a result of strengthening Walker circulation in the equatorial Pacific. The North Atlantic meridional temperature gradient was ~8°C between 4 and 3.5 Ma and grew steadily to ~12°C at 2.5 Ma when it plateaued until the end of the record at 1.4 Ma. Not surprisingly, the variance at our high latitude site is much greater than that at our tropical sites. However, the presence of significant variance at Site 982 prior to the intensification of Northern Hemisphere Glaciation (~2.75 Ma) is anomalous given the very small amplitude of variations in all other climatic time series from this time period. Obliquity variations are notable in all three records. Significant obliquity variability exists throughout the Site 982 time series while the obliquity response at both tropical sites increases markedly at ~2.7 Ma. In contrast to the minor role of precessional variations at Site 846 and Site 982, precession played very prominent role at Site 662, likely related to variations in the strength of the African monsoon system.
PP51E-03
Climate variability in a Pliocene boreal forest: evidence from tree-rings of sub-fossil wood.
In recent decades, the response of high latitudes to changes in climate has led to renewed interest in the study of ancient ecosystems under greenhouse conditions. Making the study of past Arctic climates particularly informative when considering the response of high latitudes to future warming. Tree rings have enormous potential to provide both climatic and ecological in the modern climatic system models there exist several well known oceanic cycles, such as, ENSO, PDO AO/NAO and the AMO that have been observed to exert a strong influence on regional and global climate. Recently Global Circulation Models (GCM) have begun to incorporate these climate cycles in an effort to study their effects on past and future climate. Huber and Caballero, (2003) demonstrated evidence for ENSO during the Eocene and Wara et al. (2005) and Fedorov et al., (2006) have demonstrated that these familiar cycles were in operation during the Pliocene. However, empirical evidence for climatic oscillations operating during these times remains sparse. Tree rings have enormous potential to provide both climatic and ecological information at an annual resolution. Our inter-annual study of Pliocene sub-fossil wood provides direct observations to demonstrate whether these inter-annual and inter-decadal climatic oscillations were indeed persistent through the Pliocene.
PP51E-04 INVITED
Permanent El Nino Conditions and Meridional Expansion of the Tropical Ocean Warm-Water Pool in the Early Pliocene: Evidence, Mechanisms and Global implications.
Proxy temperature records obtained during the last several years show that in the early Pliocene, approximately 3 to 5 million years ago, the tropics were characterized by a permanently warm El Nino-like state. During this time interval, and possibly before, the zonal SST gradient along the equator was significantly reduced or virtually absent. Thus far, there has been no satisfactory explanation for the climate state of the Pliocene, especially for the climate conditions in the tropics and subtropics. State-of-the-art climate models fail to reproduce a permanent El Nino, even when forced by CO2 concentrations many times larger than those estimated for the early Pliocene. Here, we summarize available evidence for a permanent El Nino state and then use available proxy temperature data from Mg/Ca and Alkenone techniques to reconstruct the latitudinal distribution of temperature in the Pacific ocean for the period of around 4Ma, exactly when the east-west temperature gradient along the equator fully collapses. Several assumptions and temperature adjustments are made to compensate for the coarseness of data. Our results indicate that a strong reduction of the meridional temperature gradient (especially from the equator to the subtropics) occurred in the Pacific ocean during the early Pliocene (in comparison with today's climate). This implies a large poleward expansion of the warm-water pool in the tropical Pacific, which may have contributed to maintaining permanent El Nino conditions. Other potential implications of the reduced meridional temperature gradient may have included a significant reduction in the strength of the Hadley cell, and changes in global precipitation and wind stress patterns.
PP51E-05
Seasonal Variations Preserved in an Extinct Neogene Scallop, Chesapecten, from Florida to Delaware, USA and its Implications for Paleobiogeography
High-resolution records of climate variability on deep-time scales are needed to advance our understanding of the impact of a warming climate on seasonality and ecological change along a latitudinal gradient. The Middle Pliocene Warm Interval (MPWI: 3.2-2.8 Ma) provides insight into a globally warmer world, in which, relative to today, continental and oceanic configurations and atmospheric CO2 levels were similar; sea and continental ice were reduced; and interiors of continents were arid. Accretionary hard parts of marine organisms serve as physical (growth lines and increments) and chemical (87Sr/86Sr, δ18 O and δ13C) archives of life history, ecology, and environmental conditions during the life of the animal. Our goal was to examine variations in seasonality across latitudinal (~27° to37°N) and biogeographic (tropical to cold-temperate) gradients of the Middle Atlantic Coastal Plain (MACP) during two intervals of warming: the Middle Miocene Climate Optimum (MMCO: 17-15 Ma) and the Middle Pliocene Warm Interval (MPWI: 3.2-2.8 Ma). We analyzed variations in annual shell growth and isotope ratios (87Sr/86Sr, δ18O and δ13C) of 12 Chesapecten shells from the extremes of their biogeographic range (tropical to warm- temperate) (MMCO: Florida and Delaware; and MPWI: Florida and Virginia). Today, Florida is warm-temperate, and Delaware and Virginia are cold-temperate. Chesapecten are an extinct genus of scallop commonly preserved in MAPC deposits. They inhabited subtidal marine environments during the Miocene and Pliocene. We used 87Sr/86Sr ratios to tightly refine the timing and modeled age of the MMCO and MPWI. Modeled ages across Chesapecten's biogeographic extremes include: Florida (MMCO) 15.5 to 14.1±0.6 Ma; Delaware (MMCO) 18.0 to 17.5±0.4 Ma; Florida (MPWI) 3.75 to 2.05±0.9 Ma; and Virginia (MPWI) 2.45 to 1.65±0.4 Ma. We estimated seasonal temperature from the δ18O time series assuming interglacial δ18OSEAWATER values of -0.05‰ and -0.35‰ for the MMCO (Florida and Delaware, respectively) and 0.96‰ and -0.50‰ for the MPWI (Florida and Virginia, respectively), accounting for latitudinal effects. Temperatures during the MMCO from Florida shells ranged from 19.7 to 31.4°C, and from Delaware shells ranged from 13.3 to 26.1°C. These data reflect diminished seasonality from Florida to Delaware at that time. Prominant annual growth lines were not apparent in these shells, a common characteristic in tropical shells. Like the MMCO temperature estimates, temperature recorded in MPWI shells from Florida lacked seasonal variability, as expected from low latitudes, with temperatures ranging from 15.6 to 21.3°C. Furthermore, like the MMCO shells, prominent annual growth lines were not evident. MPWI shells from Virginia displayed seasonal temperatures from 2.7 to 20.1°C indicating warm- temperate conditions. We predicted annual growth line formation during summer months based on studies that show modern warm-temperate bivalves inhabiting mid to high latitudes form annual growth lines during summer months. However, isotope sclerochronology revealed that these scallops produced winter growth lines. These findings have important implications for paleoceanographic and atmospheric circulation during the MPWI.
PP51E-06
Mounting Mid-Pliocene Paradoxes, Mounting Concern About Climate Sensitivity
The warm interval of the middle Pliocene continues to vex proxy data and climate modelers with its unusual combination of much warmer than modern sea surface temperatures, reduced ice sheet mass and consequent sea level rise, together with atmospheric carbon dioxide amounts that are within the error bars of present day values. Coupled ocean atmosphere GCMs simulating the mid-Pliocene have characteristically been incapable of producing sea surface temperatures at high latitudes that are as warm as those portrayed by the PRISM2 global data set, while simultaneously maintaining relatively unchanged tropical SSTs. Alkenone data for the east equatorial Pacific, and reanalysis of Mg/Ca proxies may show that warmer tropical SSTs were indeed present at this time, but it remains to be seen whether or not proxy paleo-CO2 data can be reinterpreted to lend modelers a forcing mechanism that would generate such warm ocean temperatures. Further complicating the Pliocene scenario is the fact that deep ocean temperature data that are now becoming available may not present a smoking gun that would be suggestive of ocean circulation intensification. This may not prove to be paradoxical to prior interpretations of carbon isotope ratios, which seemed indicative of increased NADW production, but once again a mechanism (increased ocean heat transport) that may have been consistent with much warmer North Atlantic SSTs, might be at odds with proxy data. Regardless, coupled models seem locked into solutions that show warming climates are generally accompanied by a weakening of the North Atlantic circulation. Thus, one of the few negative feedbacks that we consistently find in global warming scenarios is triggered in a location where the mid-Pliocene warm SST anomaly peaks. These paradoxes are explored using one of the latest versions of the GISS GCM (Model III). With the mid-Pliocene the most recent period in which global warmth approached something like late 21st century projections, we should perhaps be concerned that our closest global warming analog is telling us that our best models are not sensitive enough.
PP51E-07
The Late Pliocene Tropical Pacific
To test the permanent El Niño hypothesis and to investigate the character of tropical climate variability during the Pliocene, we used Mg-Paleothermometry in the planktonic foraminifer Globigerinoides ruber to reconstruct detailed thermal and d18O-seawater histories from ODP site 806B, in the western equatorial Pacific (WEP). Our records span the Late Pliocene between 2.3 and 3.1 Ma with an average resolution of 1.7 ky. Hole 806B G. ruber Mg/Ca-derived SST data indicate an average Late Pliocene SST of 28°C, a modest long-term cooling of 0.3°C and a glacial to interglacial SST range of 2°C. Our western equatorial Pacific SST record shows spectral characteristics previously unknown or unreported. In contrast with foraminiferal d18O records which show a dominant 41 ky-period component during the Late Pliocene, our WEP SST record shows a dominant ~100 ky-period component and a weaker contribution at 41 ky. Cross-spectral analysis reveals that SST leads ODP Hole 806B benthic foraminiferal d18O by 7.4 and 1.9 ky at the ~100 ky and 41 ky-dominant periods during the Late Pliocene. Comparison between our Hole 806B SST record with an ODP Hole 846 SST record from the eastern equatorial Pacific (EEP) cold tongue suggests that both records are coherent and dominated by 100 ky cycles and a weaker contribution at 41 ky. Furthermore, comparisons between these records indicate a ~3° C east-west zonal equatorial Pacific SST gradient and the strengthening of this gradient from 3° C to 4° C between 3.1 and 2.3 Ma. Our results are not compatible with the notion of a permanent El Niño state in the Late Pliocene. The character of Tropical Pacific thermal evolution, with a dominant 100 ky period and a weaker contribution at 41 ky observed in both eastern and western equatorial Pacific SSTs, suggests that the dominant driver of this variability was radiative forcing by atmospheric greenhouse gases. The larger glacial-interglacial amplitudes observed in eastern equatorial Pacific SSTs suggests an additional influence of EEP thermocline changes on EEP SSTs, as previously suggested (Lawrence et al., 2006).
PP51E-08
Comparison of different mechanisms proposed to explain the onset of Northern Hemisphere Glaciation
Different mechanisms have been proposed to explain the transition from the warm climate of the middle Pliocene to the following cold climate and the onset of glaciation in the Northern Hemisphere approximately 2.75 million years ago (Ma). The proposed mechanism include increased obliquity and eccentricity forcing (Maslim et al. 1998), reduced atmospheric CO2 concentrations (Mudelsee and Raymo, 2005), closure of the Central America seaway causing the strengthnening of the Atlantic Meridional Overturning Circulation and associated increased moisture supply to the areas of ice sheet formation (Haug and Tiedemann, 1998), increased seasonality in the North Pacific (Haug et al. 2005), and/or permanent El Ninno conditions prior to the transition at ~2.75 Ma (Wara et al. 2005). In this study these hypothesis are compared via simulations with an Atmospheric General Circulation Model coupled to a slab ocean model. The amplitude and spatial patterns of climate change associated to each hypothesis are analysed, and a Equilibrium Line Altitude calculation is used in order to evaluate the relevance of these signals for the initiation of ice sheets.