André Paul
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
André Paul has a total of 152 co-authors. Here are the top co-authors by number of shared publications:
- Michael Schulz
University of Bremen
25 shared publications - Stefan Mulitza
MARUM‐Center for Marine Environmental Sciences University of Bremen Bremen Germany
12 shared publications - Lukas Jonkers
MARUM - Center for Marine Environmental Sciences, University of Bremen, Bremen, Germany
11 shared publications - Martin Werner
Alfred Wegener Institute, Helmholtz Center for Polar and Marine Research, Bremerhaven, Germany
11 shared publications - Kira Rehfeld
Department of Geosciences, University of Tübingen, Tübingen, Germany
10 shared publications - Marie-Luise Kapsch
Max Planck Institute for Meteorology, Hamburg, Germany
10 shared publications
Latest Publications
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
Constraining glacial ocean carbon cycle – A multi-model study
The ocean contained a larger carbon content at the Last Glacial Maximum (LGM, ~21kyr before present) compared to the late Holocene, making a considerable contribution to the deglacial atmospheric CO2 rise of about 90 ppm. Yet, there’s no consensus on the mechanisms controlling the glacial-interglacial changes in oceanic carbon storage due to uncertainties and sparseness of proxy data. Numerical simulations have been widely used to quantify the impact of key factors, such as changes in sea surface temperatures, ocean circulation and biological production, on glacial ocean carbon sequestration. However, the robustness of these findings is subject to further testing due to the differences in process representation, parameterization, model architecture, or external forcing employed by models.Towards further constraining the LGM ocean carbon cycle, we conducted a multi-model comparison with three comprehensive Earth System Models (Alfred Wegener Institute Earth System Model, AWI-ESM; Community Earth System Model, CESM; Max Planck Institute Earth System Model, MPI-ESM) and one Earth system Model of Intermediate Complexity (CLIMBER-X). We carried out three coordinated experiments with each model: 1) PI (the pre-industrial control simulation), 2) LGM-PMIP (following PMIP4 LGM protocol) and 3) LGM-LowCO2 (as LGM-PMIP, but with boosted alkalinity inventory to lower atmospheric CO2 to about 190 ppm. All experiments were conducted with the prognostic CO2 for the carbon cycle, considering only the atmosphere and ocean reservoirs, and prescribed CO2 for radiative forcing.All models consistently show that applying the PMIP4 LGM boundary conditions alone leads to only a 5-40 ppm decrease in atmospheric CO2. Globally, the glacial CO2 drawdown in LGM-PMIP is mainly controlled by the enhanced solubility pump. The spatial distribution of the increased glacial DIC depends on the ocean circulation state in each model. In MPI-ESM and CLIMBER-X, the shallower and weaker AMOC facilitates carbon storage in the deep Atlantic. An LGM atmospheric CO2 of 190 ppm can be achieved by boosting alkalinity by 5-8% in scenario LGM-LowCO2. In all models, boosting LGM alkalinity inventory increases DIC in the bottom water. However, comparison to proxy data reveals that the models lack respired carbon, particularly in the deep Pacific. This suggests a need to enhance the glacial biological carbon pump in the models.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessConstraining glacial ocean carbon cycle – A multi-model study
The ocean contained a larger carbon content at the Last Glacial Maximum (LGM, ~21kyr before present) compared to the late Holocene, making a considerable contribution to the deglacial atmospheric CO2 rise of about 90 ppm. Yet, there’s no consensus on the mechanisms controlling the glacial-interglacial changes in oceanic carbon storage due to uncertainties and sparseness of proxy data. Numerical simulations have been widely used to quantify the impact of key factors, such as changes in sea surface temperatures, ocean circulation and biological production, on glacial ocean carbon sequestration. However, the robustness of these findings is subject to further testing due to the differences in process representation, parameterization, model architecture, or external forcing employed by models.Towards further constraining the LGM ocean carbon cycle, we conducted a multi-model comparison with three comprehensive Earth System Models (Alfred Wegener Institute Earth System Model, AWI-ESM; Community Earth System Model, CESM; Max Planck Institute Earth System Model, MPI-ESM) and one Earth system Model of Intermediate Complexity (CLIMBER-X). We carried out three coordinated experiments with each model: 1) PI (the pre-industrial control simulation), 2) LGM-PMIP (following PMIP4 LGM protocol) and 3) LGM-LowCO2 (as LGM-PMIP, but with boosted alkalinity inventory to lower atmospheric CO2 to about 190 ppm. All experiments were conducted with the prognostic CO2 for the carbon cycle, considering only the atmosphere and ocean reservoirs, and prescribed CO2 for radiative forcing.All models consistently show that applying the PMIP4 LGM boundary conditions alone leads to only a 5-40 ppm decrease in atmospheric CO2. Globally, the glacial CO2 drawdown in LGM-PMIP is mainly controlled by the enhanced solubility pump. The spatial distribution of the increased glacial DIC depends on the ocean circulation state in each model. In MPI-ESM and CLIMBER-X, the shallower and weaker AMOC facilitates carbon storage in the deep Atlantic. An LGM atmospheric CO2 of 190 ppm can be achieved by boosting alkalinity by 5-8% in scenario LGM-LowCO2. In all models, boosting LGM alkalinity inventory increases DIC in the bottom water. However, comparison to proxy data reveals that the models lack respired carbon, particularly in the deep Pacific. This suggests a need to enhance the glacial biological carbon pump in the models.
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
Climate model evaluation based on ecological principles: using fossil plankton biodiversity patterns to probe simulated thermal gradients in the ice age North Atlantic
The Last Glacial Maximum (23,000 – 19,000 years ago; LGM) is the most recent time when Earth’s climate was fundamentally different from today. The LGM hence remains an important target to evaluate climate models under boundary conditions different from today. Evaluation of paleoclimate simulations is mostly done using proxy-based reconstructions. However, such reconstructions are indirect and associated with marked uncertainty, which often renders model-data comparison equivocal. Here we take a different approach and use macro-ecological patterns preserved in fossil marine zooplankton to evaluate simulations of LGM near-surface ocean temperature.We utilise the distance-decay pattern in planktonic foraminifera to evaluate modelled temperature gradients. Distance decay emerges because of differences in habitat preferences among species that cause the compositional similarity between assemblages to decrease the further apart they are from each other in environmental space. Distance decay is a fundamental concept in ecology and is observed in many different taxa and ecosystems, including planktonic foraminifera that show a monotonous decrease in similarity with increasing difference in temperature. Because the ecological niches of planktonic foraminifera are unlikely to have changed since the LGM, the distance-decay relationship based on simulated LGM temperatures and LGM assemblages should in principle be identical to the modern distance decay pattern. Thus we can use fossil planktonic foraminifera species assemblages to evaluate climate model simulations based on ecological principles.Our analysis is based on an extended new LGM planktonic foraminifera database (2,085 assemblages from 647 unique sites) and a suite of 10 simulations from state-of-the-art climate models (PMIP3 and 4). We find markedly different planktonic foraminifera distributions during the LGM, primarily due to the equatorward expansion of polar assemblages at the expense of transitional assemblages. The distance-decay pattern that emerges when the LGM assemblages are combined with simulated ocean temperatures is different from the modern pattern. All simulations suggest large thermal gradients between regions where the planktonic foraminifera indicate no, or only weak, gradients. This pattern arises from the pronounced shift to polar species assemblages in the North Atlantic where the simulations predict only moderate cooling. In general, the models predict spatially rather uniform cooling, whereas the microfossil evidence suggests more pronounced regional differences in the temperature change. The difference between reconstructions and the simulations reaches up to 10 K in the North Atlantic.Importantly, simulations with a reduced AMOC and hence lower North Atlantic near sea surface temperatures, yield a distance-decay pattern that is much more similar to the modern pattern. The planktonic foraminifera assemblages thus question the view of the LGM ocean as an equilibrium response to external forcing. Our method opens up new avenues to evaluate climate model simulations in other settings using fundamental ecological patterns.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessClimate model evaluation based on ecological principles: using fossil plankton biodiversity patterns to probe simulated thermal gradients in the ice age North Atlantic
The Last Glacial Maximum (23,000 – 19,000 years ago; LGM) is the most recent time when Earth’s climate was fundamentally different from today. The LGM hence remains an important target to evaluate climate models under boundary conditions different from today. Evaluation of paleoclimate simulations is mostly done using proxy-based reconstructions. However, such reconstructions are indirect and associated with marked uncertainty, which often renders model-data comparison equivocal. Here we take a different approach and use macro-ecological patterns preserved in fossil marine zooplankton to evaluate simulations of LGM near-surface ocean temperature.We utilise the distance-decay pattern in planktonic foraminifera to evaluate modelled temperature gradients. Distance decay emerges because of differences in habitat preferences among species that cause the compositional similarity between assemblages to decrease the further apart they are from each other in environmental space. Distance decay is a fundamental concept in ecology and is observed in many different taxa and ecosystems, including planktonic foraminifera that show a monotonous decrease in similarity with increasing difference in temperature. Because the ecological niches of planktonic foraminifera are unlikely to have changed since the LGM, the distance-decay relationship based on simulated LGM temperatures and LGM assemblages should in principle be identical to the modern distance decay pattern. Thus we can use fossil planktonic foraminifera species assemblages to evaluate climate model simulations based on ecological principles.Our analysis is based on an extended new LGM planktonic foraminifera database (2,085 assemblages from 647 unique sites) and a suite of 10 simulations from state-of-the-art climate models (PMIP3 and 4). We find markedly different planktonic foraminifera distributions during the LGM, primarily due to the equatorward expansion of polar assemblages at the expense of transitional assemblages. The distance-decay pattern that emerges when the LGM assemblages are combined with simulated ocean temperatures is different from the modern pattern. All simulations suggest large thermal gradients between regions where the planktonic foraminifera indicate no, or only weak, gradients. This pattern arises from the pronounced shift to polar species assemblages in the North Atlantic where the simulations predict only moderate cooling. In general, the models predict spatially rather uniform cooling, whereas the microfossil evidence suggests more pronounced regional differences in the temperature change. The difference between reconstructions and the simulations reaches up to 10 K in the North Atlantic.Importantly, simulations with a reduced AMOC and hence lower North Atlantic near sea surface temperatures, yield a distance-decay pattern that is much more similar to the modern pattern. The planktonic foraminifera assemblages thus question the view of the LGM ocean as an equilibrium response to external forcing. Our method opens up new avenues to evaluate climate model simulations in other settings using fundamental ecological patterns.
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
Causal drivers of North Atlantic glacial climate variability
Terrestrial and marine proxy records suggest that climate variability was higher during the last glacial cycle than today. This is particularly true in the North Atlantic region where this enhanced variability is associated with changes in the strength of the Atlantic Meridional Overturning Circulation (AMOC) on different time scales. Despite prolonged efforts to explain an oscillatory behaviour of glacial AMOC strength, it is still not fully understood; moreover, studies often describe its mechanism qualitatively or attempt to identify its drivers using methods that do not imply causation.In this study we use the isotope-enabled Earth system model iCESM1.2 to simulate the glacial climate under conditions representing the Marine Isotope Stage 3 (about 38 thousand years before present). Our results show the AMOC oscillating with a period of roughly 500 years and an amplitude of 4 Sv (1 Sv = 106 m3s-1). Surface air temperature varies by about 1-2°C in Northern Europe, 4°C in Greenland, and up to 15°C over the North Atlantic Ocean where sea ice cover varies the most. Based on a supervised machine learning method (causal discovery), we find causal links between AMOC and North Atlantic Ocean salinity and meridional salt transport. We show how changes in the salinity of water advected into key deep-water forming areas feed back to the AMOC, thus driving the oscillation.Our findings from applying causal discovery outline the mechanism of a salt-oscillator in a fully coupled model, and indicate the potential of this method to identify causal drivers that trigger variations in AMOC strength on different time scales too.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessCausal drivers of North Atlantic glacial climate variability
Terrestrial and marine proxy records suggest that climate variability was higher during the last glacial cycle than today. This is particularly true in the North Atlantic region where this enhanced variability is associated with changes in the strength of the Atlantic Meridional Overturning Circulation (AMOC) on different time scales. Despite prolonged efforts to explain an oscillatory behaviour of glacial AMOC strength, it is still not fully understood; moreover, studies often describe its mechanism qualitatively or attempt to identify its drivers using methods that do not imply causation.In this study we use the isotope-enabled Earth system model iCESM1.2 to simulate the glacial climate under conditions representing the Marine Isotope Stage 3 (about 38 thousand years before present). Our results show the AMOC oscillating with a period of roughly 500 years and an amplitude of 4 Sv (1 Sv = 106 m3s-1). Surface air temperature varies by about 1-2°C in Northern Europe, 4°C in Greenland, and up to 15°C over the North Atlantic Ocean where sea ice cover varies the most. Based on a supervised machine learning method (causal discovery), we find causal links between AMOC and North Atlantic Ocean salinity and meridional salt transport. We show how changes in the salinity of water advected into key deep-water forming areas feed back to the AMOC, thus driving the oscillation.Our findings from applying causal discovery outline the mechanism of a salt-oscillator in a fully coupled model, and indicate the potential of this method to identify causal drivers that trigger variations in AMOC strength on different time scales too.
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
Modelling the Ocean Circulation during the mid-Cretaceous using an Energetically Consistent Internal Wave Model in the Community Earth System Model
Deep-time climate simulations typically disregard tidal dynamics while trying to reconstruct the paleoclimate. Tidally induced mixing is a dominant part of vertical mixing in the deep ocean, which is key for maintaining ocean stratification and influences the strength of the Meridional Ocean Circulation (MOC). We add this missing tidal mixing component to the Community Earth System Model (CESM) and try to reconstruct the mid-Cretaceous (~90Ma) climate, which is known for its warm, equable climate and low meridional temperature gradient. In the next step, as an improvement over the default tidal mixing scheme, an energetically consistent internal wave model IDEMIX is used in CESM to get the vertical diffusivity coefficients in simulating the mid-Cretaceous climate. Initially, 90Ma simulations were performed in the conventional method with enhanced constant background diffusivity coefficients for the default vertical mixing scheme and then with the tidal mixing component enabled. Preliminary results from the simulations with tidal mixing show that there is a considerable reduction in the global ocean mean temperature and a change in the strength of MOC in the deep ocean when compared to the simulations without the tidal mixing component. We will be also presenting results from additional experiments that are being performed with the internal wave model IDEMIX as the tidal mixing parameterization in the model. IDEMIX is forced by the dissipated barotropic tidal energy, which is modelled from the mid-Cretaceous bathymetry. With the IDEMIX parameterization, we expect more realistic results for ocean circulation hoping to reduce the disagreements between proxy data and model simulations.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessModelling the Ocean Circulation during the mid-Cretaceous using an Energetically Consistent Internal Wave Model in the Community Earth System Model
Deep-time climate simulations typically disregard tidal dynamics while trying to reconstruct the paleoclimate. Tidally induced mixing is a dominant part of vertical mixing in the deep ocean, which is key for maintaining ocean stratification and influences the strength of the Meridional Ocean Circulation (MOC). We add this missing tidal mixing component to the Community Earth System Model (CESM) and try to reconstruct the mid-Cretaceous (~90Ma) climate, which is known for its warm, equable climate and low meridional temperature gradient. In the next step, as an improvement over the default tidal mixing scheme, an energetically consistent internal wave model IDEMIX is used in CESM to get the vertical diffusivity coefficients in simulating the mid-Cretaceous climate. Initially, 90Ma simulations were performed in the conventional method with enhanced constant background diffusivity coefficients for the default vertical mixing scheme and then with the tidal mixing component enabled. Preliminary results from the simulations with tidal mixing show that there is a considerable reduction in the global ocean mean temperature and a change in the strength of MOC in the deep ocean when compared to the simulations without the tidal mixing component. We will be also presenting results from additional experiments that are being performed with the internal wave model IDEMIX as the tidal mixing parameterization in the model. IDEMIX is forced by the dissipated barotropic tidal energy, which is modelled from the mid-Cretaceous bathymetry. With the IDEMIX parameterization, we expect more realistic results for ocean circulation hoping to reduce the disagreements between proxy data and model simulations.
- Open Access
- 8 April 2024
- Published by Copernicus GmbH in Climate of the Past
- Vol. 20 (4) , 865-890
Towards spatio-temporal comparison of simulated and reconstructed sea surface temperatures for the last deglaciation
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations' reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital and millennial-scale temperature variations. The use of proxy forward modeling allows for accounting for non-climatic processes that affect the temperature reconstructions. It further avoids the need to reconstruct gridded fields or regional mean temperature time series from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model–data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales, which suggests that optimizing for agreement with the temporal patterns of a small set of proxies might be insufficient for capturing the spatial structure of the deglacial temperature variability. We attribute the diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model–data agreement. Additionally, the algorithm can be applied to variables like oxygen isotopes and climate transitions such as the penultimate deglaciation and the last glacial inception.
- 8 April 2024
- Published by Copernicus GmbH in Climate of the Past
- Vol. 20 (4) , 865-890
Open AccessTowards spatio-temporal comparison of simulated and reconstructed sea surface temperatures for the last deglaciation
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations' reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital and millennial-scale temperature variations. The use of proxy forward modeling allows for accounting for non-climatic processes that affect the temperature reconstructions. It further avoids the need to reconstruct gridded fields or regional mean temperature time series from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model–data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales, which suggests that optimizing for agreement with the temporal patterns of a small set of proxies might be insufficient for capturing the spatial structure of the deglacial temperature variability. We attribute the diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model–data agreement. Additionally, the algorithm can be applied to variables like oxygen isotopes and climate transitions such as the penultimate deglaciation and the last glacial inception.
- Open Access
- 15 December 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-42
Reply on RC1
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations’ reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital- as well as millennial-scale temperature variations. The use of proxy forward modeling avoids the need to reconstruct gridded or regional mean temperatures from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model-data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales. We attribute this diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model-data agreement.
- 15 December 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-42
Open AccessReply on RC1
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations’ reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital- as well as millennial-scale temperature variations. The use of proxy forward modeling avoids the need to reconstruct gridded or regional mean temperatures from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model-data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales. We attribute this diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model-data agreement.
- Open Access
- 15 December 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-42
Reply on RC2
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations’ reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital- as well as millennial-scale temperature variations. The use of proxy forward modeling avoids the need to reconstruct gridded or regional mean temperatures from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model-data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales. We attribute this diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model-data agreement.
- 15 December 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-42
Open AccessReply on RC2
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations’ reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital- as well as millennial-scale temperature variations. The use of proxy forward modeling avoids the need to reconstruct gridded or regional mean temperatures from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model-data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales. We attribute this diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model-data agreement.
- Open Access
- 18 October 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-42
Comment on egusphere-2023-986
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations’ reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital- as well as millennial-scale temperature variations. The use of proxy forward modeling avoids the need to reconstruct gridded or regional mean temperatures from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model-data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales. We attribute this diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model-data agreement.
- 18 October 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-42
Open AccessComment on egusphere-2023-986
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations’ reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital- as well as millennial-scale temperature variations. The use of proxy forward modeling avoids the need to reconstruct gridded or regional mean temperatures from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model-data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales. We attribute this diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model-data agreement.
- Open Access
- 26 June 2023
- Published by Copernicus GmbH in Climate of the Past
- Vol. 19 (6) , 1275-1294
Effects of Last Glacial Maximum (LGM) sea surface temperature and sea ice extent on the isotope–temperature slope at polar ice core sites
Stable water isotopes in polar ice cores are widely used to reconstruct past temperature variations over several orbital climatic cycles. One way to calibrate the isotope–temperature relationship is to apply the present-day spatial relationship as a surrogate for the temporal one. However, this method leads to large uncertainties because several factors like the sea surface conditions or the origin and transport of water vapor influence the isotope–temperature temporal slope. In this study, we investigate how the sea surface temperature (SST), the sea ice extent, and the strength of the Atlantic Meridional Overturning Circulation (AMOC) affect these temporal slopes in Greenland and Antarctica for Last Glacial Maximum (LGM, ∼ 21 000 years ago) to preindustrial climate change. For that, we use the isotope-enabled atmosphere climate model ECHAM6-wiso, forced with a set of sea surface boundary condition datasets based on reconstructions (e.g., GLOMAP) or MIROC 4m simulation outputs. We found that the isotope–temperature temporal slopes in East Antarctic coastal areas are mainly controlled by the sea ice extent, while the sea surface temperature cooling affects the temporal slope values inland more. On the other hand, ECHAM6-wiso simulates the impact of sea ice extent on the EPICA Dome C (EDC) and Vostok sites through the contribution of water vapor from lower latitudes. Effects of sea surface boundary condition changes on modeled isotope–temperature temporal slopes are variable in West Antarctica. This is partly due to the transport of water vapor from the Southern Ocean to this area that can dampen the influence of local temperature on the changes in the isotopic composition of precipitation and snow. In the Greenland area, the isotope–temperature temporal slopes are influenced by the sea surface temperatures near the coasts of the continent. The greater the LGM cooling off the coast of southeastern Greenland, the greater the transport of water vapor from the North Atlantic, and the larger the temporal slopes. The presence or absence of sea ice very near the coast has a large influence in Baffin Bay and the Greenland Sea and influences the slopes at some inland ice core stations. The extent of the sea ice far south slightly influences the temporal slopes in Greenland through the transport of more depleted water vapor from lower latitudes to this area. The seasonal variations of sea ice distribution, especially its retreat in summer, influence the isotopic composition of the water vapor in this region and the modeled isotope–temperature temporal slopes in the eastern part of Greenland. A stronger LGM AMOC decreases LGM-to-preindustrial isotopic anomalies in precipitation in Greenland, degrading the isotopic model–data agreement. The AMOC strength modifies the temporal slopes over inner Greenland slightly and by a little on the coasts along the Greenland Sea where the changes in surface temperature and sea ice distribution due to the AMOC strength mainly occur.
- 26 June 2023
- Published by Copernicus GmbH in Climate of the Past
- Vol. 19 (6) , 1275-1294
Open AccessEffects of Last Glacial Maximum (LGM) sea surface temperature and sea ice extent on the isotope–temperature slope at polar ice core sites
Stable water isotopes in polar ice cores are widely used to reconstruct past temperature variations over several orbital climatic cycles. One way to calibrate the isotope–temperature relationship is to apply the present-day spatial relationship as a surrogate for the temporal one. However, this method leads to large uncertainties because several factors like the sea surface conditions or the origin and transport of water vapor influence the isotope–temperature temporal slope. In this study, we investigate how the sea surface temperature (SST), the sea ice extent, and the strength of the Atlantic Meridional Overturning Circulation (AMOC) affect these temporal slopes in Greenland and Antarctica for Last Glacial Maximum (LGM, ∼ 21 000 years ago) to preindustrial climate change. For that, we use the isotope-enabled atmosphere climate model ECHAM6-wiso, forced with a set of sea surface boundary condition datasets based on reconstructions (e.g., GLOMAP) or MIROC 4m simulation outputs. We found that the isotope–temperature temporal slopes in East Antarctic coastal areas are mainly controlled by the sea ice extent, while the sea surface temperature cooling affects the temporal slope values inland more. On the other hand, ECHAM6-wiso simulates the impact of sea ice extent on the EPICA Dome C (EDC) and Vostok sites through the contribution of water vapor from lower latitudes. Effects of sea surface boundary condition changes on modeled isotope–temperature temporal slopes are variable in West Antarctica. This is partly due to the transport of water vapor from the Southern Ocean to this area that can dampen the influence of local temperature on the changes in the isotopic composition of precipitation and snow. In the Greenland area, the isotope–temperature temporal slopes are influenced by the sea surface temperatures near the coasts of the continent. The greater the LGM cooling off the coast of southeastern Greenland, the greater the transport of water vapor from the North Atlantic, and the larger the temporal slopes. The presence or absence of sea ice very near the coast has a large influence in Baffin Bay and the Greenland Sea and influences the slopes at some inland ice core stations. The extent of the sea ice far south slightly influences the temporal slopes in Greenland through the transport of more depleted water vapor from lower latitudes to this area. The seasonal variations of sea ice distribution, especially its retreat in summer, influence the isotopic composition of the water vapor in this region and the modeled isotope–temperature temporal slopes in the eastern part of Greenland. A stronger LGM AMOC decreases LGM-to-preindustrial isotopic anomalies in precipitation in Greenland, degrading the isotopic model–data agreement. The AMOC strength modifies the temporal slopes over inner Greenland slightly and by a little on the coasts along the Greenland Sea where the changes in surface temperature and sea ice distribution due to the AMOC strength mainly occur.
- Open Access
- 15 June 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-42
Comment on egusphere-2023-986
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations’ reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital- as well as millennial-scale temperature variations. The use of proxy forward modeling avoids the need to reconstruct gridded or regional mean temperatures from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model-data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales. We attribute this diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model-data agreement.
- 15 June 2023
- Published by Copernicus GmbH
- Vol. 2023, 1-42
Open AccessComment on egusphere-2023-986
An increasing number of climate model simulations is becoming available for the transition from the Last Glacial Maximum to the Holocene. Assessing the simulations’ reliability requires benchmarking against environmental proxy records. To date, no established method exists to compare these two data sources in space and time over a period with changing background conditions. Here, we develop a new algorithm to rank simulations according to their deviation from reconstructed magnitudes and temporal patterns of orbital- as well as millennial-scale temperature variations. The use of proxy forward modeling avoids the need to reconstruct gridded or regional mean temperatures from sparse and uncertain proxy data. First, we test the reliability and robustness of our algorithm in idealized experiments with prescribed deglacial temperature histories. We quantify the influence of limited temporal resolution, chronological uncertainties, and non-climatic processes by constructing noisy pseudo-proxies. While model-data comparison results become less reliable with increasing uncertainties, we find that the algorithm discriminates well between simulations under realistic non-climatic noise levels. To obtain reliable and robust rankings, we advise spatial averaging of the results for individual proxy records. Second, we demonstrate our method by quantifying the deviations between an ensemble of transient deglacial simulations and a global compilation of sea surface temperature reconstructions. The ranking of the simulations differs substantially between the considered regions and timescales. We attribute this diversity in the rankings to more regionally confined temperature variations in reconstructions than in simulations, which could be the result of uncertainties in boundary conditions, shortcomings in models, or regionally varying characteristics of reconstructions such as recording seasons and depths. Future work towards disentangling these potential reasons can leverage the flexible design of our algorithm and its demonstrated ability to identify varying levels of model-data agreement.