Feng He

The Earth’s energy imbalance (EEI) that develops at the top of the atmosphere is accommodated by gains or losses of energy in Earth’s heat reservoirs, leading to temperature and sea level change. The global ocean has stored about 90% of the EEI from anthropogenic forcing, but the attribution of past changes of EEI remains largely unknown, thus obscuring our understanding of past climate change. Here, we reconstruct changes in mean ocean temperature over the past 150,000 years that, with a reconstruction of ice sheet–volume changes, allow us to isolate the contributions of the dominant ocean and ice sheet heat reservoirs to the global energy inventory as well as to derive their associated contributions to EEI. We attribute orbital-scale EEI variability to joint precessional and CO ₂ forcing that caused changes in rates of ice sheet energy storage. In contrast, millennial-scale EEI variability can be attributed to radiative responses to decreases in the Atlantic meridional overturning circulation and increasing CO ₂ during Heinrich stadials that caused changes in rates of ocean energy storage.

The connection between abrupt high-latitude warming during the last glacial period and rapid temperature/hydrologic changes at lower latitudes has revealed strong inter-hemispheric teleconnections in the ocean–atmosphere system. How this connection influenced climate in the midcontinent of North America is unclear because climate archives with sufficient time resolution are scarce. This study investigates paleoclimate changes ~70–45 thousand years ago (ka) across multiple timescales (sub-annual to millennial) from speleothems in southern Wisconsin and Minnesota (USA). To do this we use fluorescent imaging (confocal laser microscopy) and a secondary ion mass spectrometer (SIMS) to collect high-resolution δ18O measurements (5-10µm) across individual stalagmite growth bands. Annual growth bands are 20-80 µm in width, allowing for subannual sampling resolution. Abrupt climate changes that were previously identified in a stalagmite from this region (Batchelor et al. 2023) are recorded as several large (2.0 to 3.0‰), negative δ18O excursions that occurred between 61–55 ka, coinciding with Dansgaard–Oeschger (DO) events 17–14. Here we extend the record from our prior work by 10 ka from a higher accumulation rate speleothem. Our high-resolution δ18O SIMS data show similar large (1 to 1.5‰) δ18O shifts during DO events 13 and 12 that occur in less than 30 years. When comparing the intra-annual δ18O gradient measured in different fluorescent bands, we observe a weaker gradient during DO interstadials, and a stronger gradient during DO stadials suggesting that seasonal changes in the hydrological system are also occurring during these events. Our results demonstrate large, abrupt temperature and hydrological responses to DO events during the last glacial period in midcontinent North America and indicate a strong seasonal climate response to DO warming and cooling. At the meeting we will present our published and unpublished measurements and related modeling simulations.

Freshwater forcing (FWF) is recognized as a primary driver of abrupt Atlantic Meridional Overturning Circulation (AMOC) changes during significant paleoclimate variations. Numerous hosing experiments have been conducted to examine the sensitivity of AMOC to FWF; however, most of research effort was on the AMOC shutdown with few examining its recovery. Therefore, the processes underlying AMOC recovery remain uncertain.This study employs version II of the TraCE-21K simulation (TraCE-21K-II) using the Community Climate System Model version 3 (CCSM3) to investigate key factors and mechanisms behind the two-phase AMOC recovery, shown in the TraCE-21K-II simulation, following the sudden cessation of FWF input at the end of the Younger Dryas (YD). Our findings reveal a strong correlation between AMOC strength and sea surface salinity (SSS) patterns in the North Atlantic, influenced differently by FWF and sea ice cover during various paleoclimate periods. Before the initial AMOC recovery, SSS is dominated by FWF input, with sea ice mostly covering high-latitude regions throughout the YD. As the FWF ceases, the interaction of warm, salty subtropical water with cold, fresh northern water induces AMOC strength oscillations. Eventually, persistent northward flow of warm, salty water leads to sea ice collapse, resulting in a sharp SSS increase and triggering the second recovery phase. These results highlight the critical role of SSS-related processes in AMOC variation and argue that the AMOC recovery may not be solely influenced by the cessation of FWF. Investigating these mechanisms further could prompt a reassessment of our current knowledge of AMOC dynamics and their implications for future climate projections.

The last deglaciation (∼20–11 ka BP) is a period of a major, long-term climate transition from a glacial to interglacial state that features multiple centennial- to decadal-scale abrupt climate variations whose root cause is still not fully understood. To better understand this time period, the Paleoclimate Modelling Intercomparison Project (PMIP) has provided a framework for an internationally coordinated endeavour in simulating the last deglaciation whilst encompassing a broad range of models. Here, we present a multi-model intercomparison of 17 transient simulations of the early part of the last deglaciation (∼20–15 ka BP) from nine different climate models spanning a range of model complexities and uncertain boundary conditions and forcings. The numerous simulations available provide the opportunity to better understand the chain of events and mechanisms of climate changes between 20 and 15 ka BP and our collective ability to simulate them. We conclude that the amount of freshwater forcing and whether it follows the ice sheet reconstruction or induces an inferred Atlantic meridional overturning circulation (AMOC) history, heavily impacts the deglacial climate evolution for each simulation rather than differences in the model physics. The course of the deglaciation is consistent between simulations except when the freshwater forcing is above 0.1 Sv – at least 70 % of the simulations agree that there is warming by 15 ka BP in most places excluding the location of meltwater input. For simulations with freshwater forcings that exceed 0.1 Sv from 18 ka BP, warming is delayed in the North Atlantic and surface air temperature correlations with AMOC strength are much higher. However, we find that the state of the AMOC coming out of the Last Glacial Maximum (LGM) also plays a key role in the AMOC sensitivity to model forcings. In addition, we show that the response of each model to the chosen meltwater scenario depends largely on the sensitivity of the model to the freshwater forcing and other aspects of the experimental design (e.g. CO₂ forcing or ice sheet reconstruction). The results provide insight into the ability of our models to simulate the first part of the deglaciation and how choices between uncertain boundary conditions and forcings, with a focus on freshwater fluxes, can impact model outputs. We can use these findings as helpful insight in the design of future simulations of this time period.

Transient simulations of the last deglaciation have been increasingly performed to better understand the processes leading to both the overall deglacial climate trajectory as well as the centennial- to decadal- scale climate variations prevalent during deglaciations. The Paleoclimate Modelling Intercomparison Project (PMIP) has provided a framework for an internationally coordinated effort in simulating the last deglaciation (~20 – 11 ka BP) whilst encompassing a broad range of models. Here, we present a multi-model intercomparison of 17 simulations of the early part of the last deglaciation (~20 – 15 ka BP) from nine different climate models spanning a range of model complexities and uncertain boundary conditions/forcings. A main contrasting element between the simulations is the method by which groups implement freshwater fluxes from the melting ice sheets and how this forcing then impacts ocean circulation and surface climate. We find that the choice of meltwater scenario heavily impacts the deglacial climate evolution, but the response of each model depends largely on the sensitivity of the model to the freshwater forcing as well as to other aspects of the experimental design (e.g., CO₂ forcing or ice sheet reconstruction). There is agreement throughout the ensemble that warming begins in the high latitudes associated with increasing insolation and delayed warming in the tropics aligned with the later increases in atmospheric CO₂ concentration. The delay in this warming in the tropics is dependent on the timescale of the CO₂ reconstruction used by the modelling group. Simulations with freshwater forcings greater than 0.1 Sverdrup (Sv) after 18 ka BP experience delayed warming in the North Atlantic, whereas simulations with smaller freshwater forcings begin deglaciating sooner. All simulations show a strong correlation between North Atlantic temperatures, atmospheric CO₂ concentrations, and the AMOC. In simulations with a freshwater forcing greater than 0.1 Sv, North Atlantic temperatures correlate strongly with changes in the AMOC. Simulations with a smaller freshwater forcing show stronger correlations with atmospheric CO₂. This indicates that the amount of meltwater strongly controls the climate trajectory of the deglaciation. Comparing multiple simulations run by the same model demonstrate model biases by showing similar surface climate spatial patterns despite the use of different ice sheet reconstructions and/or meltwater flux scenarios. Simulations run with different models, but similar boundary conditions, have provided insight into the sensitivity of individual models to particular forcings, such as the amount freshwater forcing, which has been highly debated in previous studies. This debate has stemmed from the so-called ‘meltwater paradox’ that exists in choosing how much meltwater to input into simulations of the last deglaciation (i.e., large and geologically inconsistent meltwater forcings that successfully produce abrupt climate events versus glaciologically realistic meltwater fluxes that do not). The results of this research highlight how important this decision is.

Transient simulations of the last deglaciation have been increasingly performed to better understand the processes leading to both the overall deglacial climate trajectory as well as the centennial- to decadal- scale climate variations prevalent during deglaciations. The Paleoclimate Modelling Intercomparison Project (PMIP) has provided a framework for an internationally coordinated effort in simulating the last deglaciation (~20 – 11 ka BP) whilst encompassing a broad range of models. Here, we present a multi-model intercomparison of 17 simulations of the early part of the last deglaciation (~20 – 15 ka BP) from nine different climate models spanning a range of model complexities and uncertain boundary conditions/forcings. A main contrasting element between the simulations is the method by which groups implement freshwater fluxes from the melting ice sheets and how this forcing then impacts ocean circulation and surface climate. We find that the choice of meltwater scenario heavily impacts the deglacial climate evolution, but the response of each model depends largely on the sensitivity of the model to the freshwater forcing as well as to other aspects of the experimental design (e.g., CO₂ forcing or ice sheet reconstruction). There is agreement throughout the ensemble that warming begins in the high latitudes associated with increasing insolation and delayed warming in the tropics aligned with the later increases in atmospheric CO₂ concentration. The delay in this warming in the tropics is dependent on the timescale of the CO₂ reconstruction used by the modelling group. Simulations with freshwater forcings greater than 0.1 Sverdrup (Sv) after 18 ka BP experience delayed warming in the North Atlantic, whereas simulations with smaller freshwater forcings begin deglaciating sooner. All simulations show a strong correlation between North Atlantic temperatures, atmospheric CO₂ concentrations, and the AMOC. In simulations with a freshwater forcing greater than 0.1 Sv, North Atlantic temperatures correlate strongly with changes in the AMOC. Simulations with a smaller freshwater forcing show stronger correlations with atmospheric CO₂. This indicates that the amount of meltwater strongly controls the climate trajectory of the deglaciation. Comparing multiple simulations run by the same model demonstrate model biases by showing similar surface climate spatial patterns despite the use of different ice sheet reconstructions and/or meltwater flux scenarios. Simulations run with different models, but similar boundary conditions, have provided insight into the sensitivity of individual models to particular forcings, such as the amount freshwater forcing, which has been highly debated in previous studies. This debate has stemmed from the so-called ‘meltwater paradox’ that exists in choosing how much meltwater to input into simulations of the last deglaciation (i.e., large and geologically inconsistent meltwater forcings that successfully produce abrupt climate events versus glaciologically realistic meltwater fluxes that do not). The results of this research highlight how important this decision is.