Search

Oral presentation at the CLIMERI-France workshop in Bordeaux, France, to present modeling results based on the newly developed French CMIP6 models

The APPLICATE Task 1.2.3 is dedicated to the development of diagnostics/metrics relevant to investigate climate linkages between the Arctic and the mid-latitudes. Here we report the progress made within this task and provide several lists of diagnostics/metrics that can be used by APPLICATE partners and the wider community for weather and climate model evaluation and/or climate change analyses. There are still important gaps in our knowledge regarding the linkages between Arctic changes and mid-latitude variability and changes, both in the atmosphere and in the ocean. Substantial work has been made in Task 1.2.3 to review the literature and select the most appropriate diagnostics for documenting Arctic - mid-latitude linkages. Novel diagnostics have been developed, and several others are still under development. Partners involved in this task already use these diagnostics as metrics for the evaluation of their institutional weather and climate models. The software codes used to calculate the new diagnostics/metrics are now available for APPLICATE partners and very soon available for the wider community. Key diagnostics/metrics will become available in ESMValTool v2.0. The upcoming release of CMIP6 and PAMIP multi-model data constitutes an ideal opportunity (Task 1.3) to sorely test these software codes and to implement well-tested diagnostics/metrics into ESMValTool. The ultimate objectives are (i) to document potential improvements in the representation of mid-latitude dynamics since CMIP5 (including linkages with the Arctic), (ii) to quantify CMIP6 projected changes in the atmospheric circulation, their uncertainties, and the contribution of the Arctic region, and (iii) to narrow model uncertainties in future projections by using the developed metrics as emergent constraints (Task 1.5).

Different practices are currently being used to couple the different components of global models for weather forecasts (several days to weeks ahead) and climate applications (several decades to centuries). This report provides a summary of the coupling methods currently used for Numerical Weather Prediction (NWP) and climate applications. We have reviewed the different approaches employed in the community and used them as a motivation to design novel experiments aimed at improving the coupling between the atmosphere, ocean and sea ice. The results of these experiments, together with prior results from the literature, have guided the recommendations provided in this report. To improve coupling methodology in NWP and climate models, we recommend using fluxes that are consistent at the interface between the atmosphere, the ocean and the sea ice. More specifically, we show that accounting for the differences in horizontal resolution between the different components of the Earth System when computing the surface fluxes, or accounting for the different sea ice thickness categories can lead to small but detectable differences in the representation of key variables like Arctic sea ice extent and volume. Even if the differences are small, computing the fluxes in a consistent way is more physical and could lead to a better representation of the atmospheric boundary layer and consequently of near-surface weather. We further show that the representation of key physical processes such as snow over sea ice is essential for a realistic representation of near surface temperature in coupled models and it is particularly important for NWP. Representing processes such as snow over sea ice is key to reduce model biases in the polar regions and hence to improve short term predictions and the representation of the model climate. We expect that the results presented in this report will help weather and climate model developments in the future.

In such a remote and harsh environment as the Arctic, the monitoring of essential climate variables is expensive and, therefore, sporadic. In turn, satellites provide observations constrained to the surface. Also, because of technical restrictions, satellites cannot sample, at least not year-round, a set of essential variables, such as the sea ice thickness. To overcome this difficulty, numerical models are key tools for studying and predicting the Arctic weather and climate. Numerical models aim at reproducing the interactions between different climate components such as the land, atmosphere, ocean, and sea ice. Such interactions are complex and described with non-linear functions. This is one reason numerical models are in constant improvement. The primary goals of Work Package 2 are to promote improvements in numerical models and establish the impact of these improvements on model results, both for the study of the Arctic climate and numerical weather prediction. This deliverable presents a set of model enhancements that are implemented and tested in fully coupled models (AWI-CM1, HadGEM3-GC3.1, EC-Earth3 PRIMAVERA, and ECMWF IFS CY45R1) and forced-mode (NEMO3.6-LIM3 and GELATO). All model developments aim to improve the representation of physical processes that take place in the sea ice or snow. These consist of a better representation of the turbulent exchanges of heat and momentum between the atmosphere and sea ice (form drag), a description of the melt ponds that appear on the sea ice surface during the melt season (melt ponds), a parameterization which takes into account the effect of the sea ice attached to the shore and ocean floor (landfast ice), a scheme which accounts for the size of the floes that form the sea ice cover (floe size distribution), and an improved representation of the snow both over land and sea ice (multilayer snow scheme). We assess the benefit of the model developments based on the following five aspects: (i) changes in various components of the Arctic surface energy budget, (ii) changes in the transfer of momentum from the atmosphere to the ocean, (iii) the overall realism of the simulated climate system, (iv) effects on the Arctic Ocean circulation, and (v) changes in the Arctic climate sensitivity. Common results emerged from the form drag experiments: increased sea ice drift speed in the marginal ice zone, a general decrease in ice thickness, and a marginal decrease of ice concentration at the ice edge in summer. The form drag parameterization improved the large-scale atmospheric and ocean-driven ocean circulation. The melt pond parameterization shows a clear impact on the albedo and sea ice variability and reinforces that a reduced sea ice regime in the Arctic impacts the large-scale, density-driven ocean circulation. The multilayer snow scheme makes the models more sensitive to the surface thermodynamic forcing than the control run with a single layer of snow and shows a more realistic albedo. In numerical weather prediction, the multilayer scheme leads to an improved prediction of the 2-m temperature diurnal cycle over land. Over sea ice, the new scheme reduces large positive biases of outgoing longwave radiation. Fast ice developments lead to more realistic sea ice conditions supported by evidence from observational data. Parameterization of floe size distribution reveals sea ice growth caused by large ice floes in the marginal ice zone during summer.

Oral presentation at the CLIMERI-France workshop in Bordeaux, France, to present modeling results based on the newly developed French CMIP6 models

In such a remote and harsh environment as the Arctic, the monitoring of essential climate variables is expensive and, therefore, sporadic. In turn, satellites provide observations constrained to the surface. Also, because of technical restrictions, satellites cannot sample, at least not year-round, a set of essential variables, such as the sea ice thickness. To overcome this difficulty, numerical models are key tools for studying and predicting the Arctic weather and climate. Numerical models aim at reproducing the interactions between different climate components such as the land, atmosphere, ocean, and sea ice. Such interactions are complex and described with non-linear functions. This is one reason numerical models are in constant improvement. The primary goals of Work Package 2 are to promote improvements in numerical models and establish the impact of these improvements on model results, both for the study of the Arctic climate and numerical weather prediction. This deliverable presents a set of model enhancements that are implemented and tested in fully coupled models (AWI-CM1, HadGEM3-GC3.1, EC-Earth3 PRIMAVERA, and ECMWF IFS CY45R1) and forced-mode (NEMO3.6-LIM3 and GELATO). All model developments aim to improve the representation of physical processes that take place in the sea ice or snow. These consist of a better representation of the turbulent exchanges of heat and momentum between the atmosphere and sea ice (form drag), a description of the melt ponds that appear on the sea ice surface during the melt season (melt ponds), a parameterization which takes into account the effect of the sea ice attached to the shore and ocean floor (landfast ice), a scheme which accounts for the size of the floes that form the sea ice cover (floe size distribution), and an improved representation of the snow both over land and sea ice (multilayer snow scheme). We assess the benefit of the model developments based on the following five aspects: (i) changes in various components of the Arctic surface energy budget, (ii) changes in the transfer of momentum from the atmosphere to the ocean, (iii) the overall realism of the simulated climate system, (iv) effects on the Arctic Ocean circulation, and (v) changes in the Arctic climate sensitivity. Common results emerged from the form drag experiments: increased sea ice drift speed in the marginal ice zone, a general decrease in ice thickness, and a marginal decrease of ice concentration at the ice edge in summer. The form drag parameterization improved the large-scale atmospheric and ocean-driven ocean circulation. The melt pond parameterization shows a clear impact on the albedo and sea ice variability and reinforces that a reduced sea ice regime in the Arctic impacts the large-scale, density-driven ocean circulation. The multilayer snow scheme makes the models more sensitive to the surface thermodynamic forcing than the control run with a single layer of snow and shows a more realistic albedo. In numerical weather prediction, the multilayer scheme leads to an improved prediction of the 2-m temperature diurnal cycle over land. Over sea ice, the new scheme reduces large positive biases of outgoing longwave radiation. Fast ice developments lead to more realistic sea ice conditions supported by evidence from observational data. Parameterization of floe size distribution reveals sea ice growth caused by large ice floes in the marginal ice zone during summer.

Public presentation at the College of Engineering,Mathematics and Physical Sciences, University of Exeter, April 2019

Polar amplification – the phenomenon where external radiative forcing produces a larger change in surface temperature at high latitudes than the global average – is a key aspect of anthropogenic climate change, but its causes and consequences are not fully understood. The Polar Amplification Model Intercomparison Project (PAMIP) contribution to the sixth Coupled Model Intercomparison Project (CMIP6; Eyring et al., 2016) seeks to improve our understanding of this phenomenon through a coordinated set of numerical model experiments documented here. In particular, PAMIP will address the following primary questions: (1)what are the relative roles of local sea ice and remote sea surface temperature changes in driving polar amplification? (2)How does the global climate system respond to changes in Arctic and Antarctic sea ice? These issues will be addressed with multi-model simulations that are forced with different combinations of sea ice and/or sea surface temperatures representing present-day, pre-industrial and future conditions. The use of three time periods allows the signals of interest to be diagnosed in multiple ways. Lower-priority tier experiments are proposed to investigate additional aspects and provide further understanding of the physical processes. These experiments will address the following specific questions: what role does ocean–atmosphere coupling play in the response to sea ice? How and why does the atmospheric response to Arctic sea ice depend on the pattern of sea ice forcing? How and why does the atmospheric response to Arctic sea ice depend on the model background state? What have been the roles of local sea ice and remote sea surface temperature in polar amplification, and the response to sea ice, over the recent period since 1979? How does the response to sea ice evolve on decadal and longer timescales? A key goal of PAMIP is to determine the real-world situation using imperfect climate models. Although the experiments proposed here form a coordinated set, we anticipate a large spread across models. However, this spread will be exploited by seeking “emergent constraints” in which model uncertainty may be reduced by using an observable quantity that physically explains the intermodel spread. In summary, PAMIP will improve our understanding of the physical processes that drive polar amplification and its global climate impacts, thereby reducing the uncertainties in future projections and predictions of climate change and variability.

Oral presentation at the 2019 PAMIP workshop at Devon, UK. This international workshop was aimed at sharing initial results from the Polar Amplification Model Intercomparison Project (PAMIP) to improve our understanding of the physical processes that drive polar amplification and its global climate impacts.

We analyze the Northern Hemisphere atmospheric response to Arctic sea ice decline in the CNRM-CM6 model (130km resolution, 91 vertical levels) using atmosphere-only experiments that follow the PAMIP protocoldescribed in Smith et al. (2019). The high-top atmospheric component of CNRM-CM6 is forced by the same pattern of SST but different idealized patterns of sea ice that correspond to i) pre-industrial conditions (PI), ii) present day conditions (PD), and iii) future conditions (FUT) associated with a 2C warming with respect to i). Each experiment is run for 14 months starting in April, using at least 100 members. We focus on the cold season response (October to March) and describe the atmospheric response for different sea ice forcings applied either in to whole Arctic or to specificregions of the Arctic.

Work package 3 within APPLICATE examines atmospheric and oceanic linkages with the objectives of (1) advancing our understanding of the mechanisms by which the mid-latitude weather and climate could respond to the substantial Arctic climate change that is expected in the coming decades and (2) coordinate a suite of novel multi-model experiments designed to identify the oceanic and atmospheric linkages between the Arctic region and the northern mid- latitudes. Deliverable 3.2 presents results from two of these experiments in which the impact of Arctic sea-ice loss is investigated using state-of-the-art climate models forced by sea-ice concentrations from the present-day climate and from a climate with reduced sea-ice concentrations in the Arctic. Investigating the difference between these model experiments allows us to assess the effect of Arctic sea-ice decline, focusing on lower latitudes. This is relevant for the project objectives of APPLICATE, which is to develop enhanced predictive capacity for weather and climate in the Arctic and beyond, and to determine the influence of Arctic climate change on Northern Hemisphere mid-latitudes, for the benefit of policy makers, businesses and society. Results show that Arctic sea-ice loss is associated with significant low-level warming over the Arctic region. The warming is strongest during boreal autumn and winter. In these seasons, the near-surface temperature response is largest in the regions where the sea ice is changing the most, over the Arctic Basin during autumn and more over the marginal seas during winter; but temperature changes are also occurring in regions in Eastern Europe, Siberia and North America. Most models show an equatorward shift of the Northern Hemisphere tropospheric jet stream during winter, but there is large variation in the magnitude of the response. Furthermore, even the sign of the response of the stratospheric jet is uncertain. We find a significant low- pressure anomaly residing over the central Arctic during fall. During winter, the response varies more from model to model, but there is generally an increase in pressure in the Icelandic region and decrease further south (though the significance varies between models), in line with a negative phase of the NAO and an equatorward shift of the tropospheric jet. A shift in the tropospheric jet has potential ramifications for extreme events. We find a significant decrease in surface winds and precipitation during winter over Northern Europe in HadGEM3-GA7.1, which corroborates results from the recent APPLICATE case study 1 which highlights the potential role of low autumn Arctic sea ice leading to extreme climate events at mid-latitudes the following winter. Extreme weather and Arctic amplification has also been linked to changes in the waviness of the atmospheric flow. We find a significant influence of Arctic sea-ice loss on planetary-scale waves during summer in CNRM-CM6. However, no significant differences are found during winter, nor for synoptic-scale waves for either season. These results are corroborated by results from ECHAM6.3, using a sinuosity index. The spread in model responses reported here is expected and desirable since it potentially allows the real-world situation to be diagnosed using the “emergent constraint” framework. A key part of this will be to understand the physical processes in detail and hence develop an observable metric to explain the differences between models and reduce the uncertainty. Other experiments within APPLICATE are aimed at understanding the physical processes, specifically using coupled models and testing the sensitivity to regional sea-ice changes and model biases. These will be reported in future deliverables D3.1 and D3.3. An important achievement of APPLICATE has been establishing the Polar Amplification Model Intercomparison Project (PAMIP), which includes the experiments described above and expands the set of models beyond those run by the APPLICATE partners. An overall synthesis of results from APPLICATE and PAMIP will be presented in D3.4.

Work package 3 within APPLICATE examines atmospheric and oceanic linkages with the objectives of (1) advancing our understanding of the mechanisms by which the mid-latitude weather and climate could respond to the substantial Arctic climate change that is expected in the coming decades and (2) coordinate a suite of novel multi-model experiments designed to identify the oceanic and atmospheric linkages between the Arctic region and the northern mid- latitudes. Deliverable 3.2 presents results from two of these experiments in which the impact of Arctic sea-ice loss is investigated using state-of-the-art climate models forced by sea-ice concentrations from the present-day climate and from a climate with reduced sea-ice concentrations in the Arctic. Investigating the difference between these model experiments allows us to assess the effect of Arctic sea-ice decline, focusing on lower latitudes. This is relevant for the project objectives of APPLICATE, which is todevelop enhanced predictive capacity for weather and climate in the Arctic and beyond, and to determine the influence of Arctic climate change on Northern Hemisphere mid-latitudes, for the benefit of policy makers, businesses and society. Results show that Arctic sea-ice loss is associated with significant low-level warming over the Arctic region. The warming is strongest during boreal autumn and winter. In these seasons, the near-surface temperature response is largest in the regions where the sea ice is changing the most, over the Arctic Basin during autumn and more over the marginal seas during winter; but temperature changes are also occurring in regions in Eastern Europe, Siberia and North America.Most models show an equatorward shift of the Northern Hemisphere tropospheric jet stream during winter, but there is large variation in the magnitude of the response. Furthermore, even the sign of the response of the stratospheric jet is uncertain.We find a significant low- pressure anomaly residing over the central Arctic during fall. During winter, the response varies more from model to model, but there is generally an increase in pressure in the Icelandic region and decrease further south (though the significance varies between models), in line with anegative phase of the NAO and an equatorward shift of the tropospheric jet. A shift in the tropospheric jet has potential ramifications for extreme events. We find a significant decrease in surface winds and precipitation during winter over Northern Europe in HadGEM3-GA7.1, which corroborates results from the recent APPLICATE case study 1 which highlights the potential role of low autumn Arctic sea ice leading to extreme climate events at mid-latitudes the following winter. Extreme weather and Arctic amplification has also been linked to changes in the waviness of the atmospheric flow. We find a significant influence of Arctic sea-ice loss on planetary-scale waves during summer in CNRM-CM6. However, no significant differences are found during winter, nor for synoptic-scale waves for either season. These results are corroborated by results from ECHAM6.3, using a sinuosity index. The spread in model responses reported here is expected and desirable since it potentially allows the real-world situation to be diagnosed using the “emergent constraint” framework. A key partof this will be to understand the physical processes in detail and hence develop an observable metric to explain the differences between models and reduce the uncertainty. Other experiments within APPLICATE are aimed at understanding the physical processes, specifically using coupled models and testing the sensitivity to regional sea-ice changes and model biases. These will be reported in future deliverables D3.1 and D3.3. An important achievement of APPLICATE has been establishing the Polar Amplification Model Intercomparison Project (PAMIP), which includes the experiments described above and expands the set of models beyond those run by the APPLICATE partners. An overall synthesis of results from APPLICATE and PAMIP will be presented in D3.4.