Search

The radiocarbon analysis of uranium‐thorium‐dated cold‐water corals (CWCs) provides an excellent opportunity for qualitative reconstruction of past ocean circulation and water mass aging. While mid‐depth water mass aging has been studied in the Atlantic Ocean, the evolution of the thermocline is still largely unknown. Here we present a combined 14C and 230Th/U age record obtained from thermocline dwelling CWCs at various sites in the eastern Atlantic Ocean, with intermittently centennial resolution over the last 32 ka. Shallow dwelling CWCs off Angola, located in the South Atlantic, infer a link between the mid‐depth equatorial Atlantic and Southern Ocean. They confirm a 14C drawdown during the Last Glacial Maximum (LGM) and advocate for a consistent Southern Hemisphere radiocarbon aging of upper thermocline waters, as well as strong depth gradients and high variability. Direct comparison with 14C simulations carried out with an Ocean General Circulation Model yield good agreement for Angola. In contrast, the North Atlantic thermocline shows well‐ventilated water with strong variations near the position of today's Azores Front (AF), neither of which are captured by the model. During the Bølling‐Allerød, we confirm the important role of the AF in separating North and South Atlantic thermocline waters and provide further evidence of a 500 year long deep convection interruption within the Younger Dryas (YD). We conclude that the North and South Atlantic thermocline waters were separately acting carbon reservoirs during the LGM and subsequent deglaciation until the modern circulation was established during the YD. Key Points: North Atlantic cold‐water corals trace well‐ventilated thermocline waters near major oceanic fronts since the Last Glacial MaximumAcross the South Atlantic into the Southern Ocean, aged waters with large variability and connectivity are evident during the last glacialThe modern state of radiocarbon ventilation of the thermocline Atlantic is initiated during the Younger Dryas cold reversal [ABSTRACT FROM AUTHOR]

Since it is currently not understood how changes in 14C production rate (Q), and in the carbon cycle, can be combined to explain the reconstructed atmospheric Δ14C record, we discuss possible reasons for this knowledge gap. Reviewing the literature, we exclude that changes in the content of atoms in the atmosphere, which produce cosmogenic 14C after being hit by galactic cosmic rays, might be responsible for parts of the observed differences. When combining Q with carbon cycle changes, one needs to understand the changes in the atmospheric 14C inventory, which are partially counterintuitive. For example, during the Last Glacial Maximum, Δ14C was ∼400‰ higher compared with preindustrial times, but the 14C inventory was 10% smaller. Some pronounced changes in atmospheric Δ14C do not correspond to any significant changes in the atmospheric 14C inventory, since CO2 was changing simultaneously. Using two conceptually different models (BICYCLE‐SE and LSG‐OGCM), we derive hypothetical Qs by forcing the models with identical atmospheric CO2 and Δ14C data. Results are compared with the most recent data‐based estimates of Q derived from cosmogenic isotopes. Millennial‐scale climate change connected to the bipolar seesaw is missing in the applied models, which might explain some, but probably not all, of the apparent model‐data disagreement in Q. Furthermore, Q based on either data from marine sediments or ice cores contains offsets, suggesting an interpretation deficit in the current data‐based approaches. Key Points: No important change found in the level of precursor material, which produces cosmogenic radiocarbon after being hit by galactic cosmic raysTransient changes in the bipolar seesaw are needed in applied carbon cycle models to improve interpretation of the radiocarbon cycleSediment core‐ and ice core‐based radiocarbon production rates differ systematically, suggesting missing processes [ABSTRACT FROM AUTHOR]

The past provides evidence of abrupt climate shifts and changes in the frequency of climate and weather extremes. We explore the nonlinear response to orbital forcing and then consider climate millennial variability down to daily weather events. Orbital changes are translated into regional responses in temperature, where the precessional response is related to nonlinearities and seasonal biases in the system. We question regularities found in climate events by analyzing the distribution of interevent waiting times. Periodicities of about 900 and 1,150 yr are found in ice cores besides the prominent 1,500 yr cycle. However, the variability remains indistinguishable from a random process, suggesting that centennial‐to‐millennial variability is stochastic in nature. New numerical techniques are developed allowing for a high resolution in the dynamically relevant regions like coasts, major upwelling regions, and high latitudes. Using this model, we find a strong sensitivity of the Atlantic meridional overturning circulation depending on where the deglacial meltwater is injected into. Meltwater into the Mississippi and near Labrador hardly affect the large‐scale ocean circulation, whereas subpolar hosing mimicking icebergs yields a quasi shutdown. The same multiscale approach is applied to radiocarbon simulations enabling a dynamical interpretation of marine sediment cores. Finally, abrupt climate events also have counterparts in the recent climate records, revealing a close link between climate variability, the statistics of North Atlantic weather patterns, and extreme events. Plain Language Summary: Predicting the future spread of possible climates, the risk of climate extremes and the risk of rapid transitions is of high socioeconomic relevance. The past provides evidence of abrupt climate change and the frequency of extremes. This allows to separate anthropogenic signals from natural climate variability. Earth system models applied both to past and future scenarios will enhance our ability to detect regime shifts which are necessary to potentially predict climate extremes and transitions. We consider the response of the system to regular orbital forcing and then focus on shorter time scales down to weather. The appearance of precession is linked to nonlinear responses of the climate system to external orbital forcing. Furthermore, we find that centennial‐to‐millennial variability is stochastic in nature. We also discuss recent developments of climate models with superior resolution in typical retrieval regions of paleoclimate records, such as continental margins and coasts. Using this model, we find a strong sensitivity of the Atlantic meridional overturning circulation depending on where the deglacial meltwater is injected into. Meltwater into the Mississippi and near Labrador hardly affect the large‐scale ocean circulation, whereas subpolar hosing related to icebergs yields a quasi shutdown. Our multiscale approach is applied to radiocarbon simulations enabling a dynamical interpretation of marine sediment cores. Finally, we explore a close link between climate variability, the statistics of weather patterns, and extreme events. Key Points: Nonlinear response of the climate system to external orbital forcing yields responses in the precessional bandCentennial and millennial variability are stochastic in natureOcean and radiocarbon modeling with high resolution at the coasts and high latitudes provides a suitable framework for multi‐scale dynamicsThe past provides evidences of abrupt climate changes and weather extremes [ABSTRACT FROM AUTHOR]

Ocean warming may lead to smaller body sizes of marine ectotherms, because metabolic rates increase exponentially with temperature while the capacity of the cardiorespiratory system to match enhanced oxygen demands is limited. Here, we explore the impact of rising sea water temperatures on Atlantic cod ( Gadus morhua), an economically important fish species. We focus on changes in the temperature-dependent growth potential by a transfer function model combining growth observations with climate model ensemble temperatures. Growth potential is expressed in terms of asymptotic body weight and depends on water temperature. We consider changes between the periods 1985-2004 and 2081-2100, assuming that future sea water temperatures will evolve according to climate projections for IPCC AR5 scenario RCP8.5. Our model projects a response of Atlantic cod to future warming, differentiated according to ocean regions, leading to increases of asymptotic weight in the Barents Sea, while weights are projected to decline at the southern margin of the biogeographic range. Southern spawning areas will disappear due to thermal limitation of spawning stages. These projections match the currently observed biogeographic shifts and the temperature- and oxygen-dependent decline in routine aerobic scope at southern distribution limits. [ABSTRACT FROM AUTHOR]