Periodic Reporting for period 1 - OptimESM (Optimal High Resolution Earth System Models for Exploring Future Climate Changes)
Okres sprawozdawczy: 2023-01-01 do 2024-03-31
OptimESM will develop a novel generation of Earth system models (ESMs), combining high resolution with an unprecedented representation of key physical and biogeochemical processes. These models will be used to deliver cutting-edge and policy-relevant knowledge around the consequences of reaching or exceeding different levels of global warming, including the risk of rapid change in key Earth system phenomena and the regional impacts arising both from the level of global warming and the occurrence of abrupt changes. OptimESM will realise these goals by bringing together four ESM groups with Integrated Assessment Modelling teams, as well as experts in model evaluation, Earth system processes, machine learning, climate impacts and science communication.
OptimESM will further develop new policy-relevant emission and land use scenarios, including ones that realise the Paris Agreement, and others that temporarily or permanently overshoot the Paris targets. Using these scenarios, OptimESM will deliver long-term projections that will increase our understanding of the risk for triggering potential tipping points in phenomena such as; ice sheets, sea ice, ocean circulation, marine ecosystems, permafrost, and terrestrial ecosystems. OptimESM will further our understanding of the processes controlling such tipping points, attribute the risk of exceeding various tipping points to the level of global warming, and develop a range of techniques to forewarn the occurrence of tipping points in the real world.
New knowledge and data from OptimESM will be actively communicated to other disciplines, such as the impacts and policy research communities, as well as the general public. This knowledge will provide a solid foundation for actionable science-based policies.
OptimESM will further develop new policy-relevant emission and land use scenarios, including ones that realise the Paris Agreement, and others that temporarily or permanently overshoot the Paris targets. Using these scenarios, OptimESM will deliver long-term projections that will increase our understanding of the risk for triggering potential tipping points in phenomena such as; ice sheets, sea ice, ocean circulation, marine ecosystems, permafrost, and terrestrial ecosystems. OptimESM will further our understanding of the processes controlling such tipping points, attribute the risk of exceeding various tipping points to the level of global warming, and develop a range of techniques to forewarn the occurrence of tipping points in the real world.
New knowledge and data from OptimESM will be actively communicated to other disciplines, such as the impacts and policy research communities, as well as the general public. This knowledge will provide a solid foundation for actionable science-based policies.
OptimESM has started developments in three different but interlinked areas. First, Earth system models (ESM); second, new land use and emission scenarios and third, machine learning (ML) based methods for regional downscaling of climate data.
The ESM development included both integrating recent developments into project ESMs, and actions to bring together improved process realism and high resolution in the new generation of ESMs. These actions focus on the representation of key ESM processes (e.g. ice sheets, permafrost, carbon and methane cycles, fires, soil and vegetation), but also on the development of new methods to accelerate the spin-up and calibration of coupled ESMs, and to develop tools to run ocean biogeochemical and atmospheric chemical model components on coarser grids than the respective physical models in ocean and atmosphere, thus increasing model computing speed.
The development of new land and emission scenarios has focused on improving the modeling of land-use changes to represent the forestry sector and the associated carbon stock and flow dynamics in the Integrated Assessement Framwork REMIND-MAgPIE. Moreover, along with afforestation, bioenergy with carbon capture and sequestration, and Direct Air Capture option to remove carbon via Ocean Alkalinity Enhancement has been added to the REMIND-MAgPIE model.
The development of ML-based tools for regional downscaling made good progress and they have successfully been applied for downscaling reanalysis data from different coarser to different higher spatial resolution (up to 5 km). The developed methods have been assessed against reanalysis and against high resolution regional climate models.
In addition to the developments, first emission (CO2-emissions) driven ESM simulations for the pre-industrial climate, historical climate and first idealised warming scenarios have been performed.
Tools for evaluation and analysis of the ongoing and upcoming ESM simulations have been developed and have been tested on existing model simulations that have been carried our as part of the 6th phase of the Coupled Model Intercomparison Project (CMIP6).
In particular, algorithms have been developed to detect events, possibly associated with tipping points, with an initial focus on ocean and sea-ice. Further, a start has been made in developing new techniques for early warning signals under rapid.
OptimESM has closely collaborated with other ongoing EU-projects and international activities to reach synergies and maximize the impact of the work.
The ESM development included both integrating recent developments into project ESMs, and actions to bring together improved process realism and high resolution in the new generation of ESMs. These actions focus on the representation of key ESM processes (e.g. ice sheets, permafrost, carbon and methane cycles, fires, soil and vegetation), but also on the development of new methods to accelerate the spin-up and calibration of coupled ESMs, and to develop tools to run ocean biogeochemical and atmospheric chemical model components on coarser grids than the respective physical models in ocean and atmosphere, thus increasing model computing speed.
The development of new land and emission scenarios has focused on improving the modeling of land-use changes to represent the forestry sector and the associated carbon stock and flow dynamics in the Integrated Assessement Framwork REMIND-MAgPIE. Moreover, along with afforestation, bioenergy with carbon capture and sequestration, and Direct Air Capture option to remove carbon via Ocean Alkalinity Enhancement has been added to the REMIND-MAgPIE model.
The development of ML-based tools for regional downscaling made good progress and they have successfully been applied for downscaling reanalysis data from different coarser to different higher spatial resolution (up to 5 km). The developed methods have been assessed against reanalysis and against high resolution regional climate models.
In addition to the developments, first emission (CO2-emissions) driven ESM simulations for the pre-industrial climate, historical climate and first idealised warming scenarios have been performed.
Tools for evaluation and analysis of the ongoing and upcoming ESM simulations have been developed and have been tested on existing model simulations that have been carried our as part of the 6th phase of the Coupled Model Intercomparison Project (CMIP6).
In particular, algorithms have been developed to detect events, possibly associated with tipping points, with an initial focus on ocean and sea-ice. Further, a start has been made in developing new techniques for early warning signals under rapid.
OptimESM has closely collaborated with other ongoing EU-projects and international activities to reach synergies and maximize the impact of the work.
OptimESM has started to progress beyond the state of the art in five key areas:
a) Delivering the next generation of high-resolution ESMs
OptimESM work has focused on:
• Increasing model resolution: OptimESM has started to develop tools to run ocean biogeochemistry and atmospheric chemistry at lower resolution than the physical components and to explore methods for more efficient tuning and spin-up of the ESMs. These are first important steps beyond the state of the art towards high-resolution ESMs
• Increasing process realism: OptimESM work so far progressed on the representation of continental ice sheets and their interactions with ice shelves and ocean, permafrost, wildfires and land surface processes and their link to regional climate and CO2 uptake and the risk of dryland expansion in a warming climate.
b) Developing new, policy-relevant emission and land use scenarios
The modeling of land-use changes to represent the forestry sector and the associated carbon stock has been improved. Moreover, along with afforestation, bioenergy with carbon capture and sequestration, and Direct Air Capture option to remove carbon via Ocean Alkalinity Enhancement has been added to the REMIND-MAgPIE model.
A preliminary set of new land use and emission scenarios until 2100 has been derived with REMIND-MAgPIE. Important steps towards the extensions beyond 2100 until 2300 have been undertaken using the small climate model ACC2.
c) Evaluating ESMs and uncertainty in climate projections
OptimESM has completed a compendium of existing observational datasets relevant for identifying the current climate trends and evaluating ESMs. The document outlines the latest observational datasets for Earth system variables relevant to tipping points, and will be kept up to date throughout the project. OptimESM has prepared evaluation scripts and developed new tools and indices to evaluate the models.
d) Delivering new knowledge on potential abrupt climate changes
OptimESM work focused on making an inventory of occurrence of strong nonlinear surprises including possible tipping points in the CMIP6 database with an initial focus on ocean and sea ice. The second focus was on developing new techniques for early warning signals under rapid climate forcing. These early warming techniques have then been tested for sub-polar gyre collapses in CMIP6 simulations.
e) Delivering new knowledge on regional consequences
OptimESM aims to provide new understanding of the regional climate response to long-term global change and to the occurrence of different abrupt changes.
As first steps beyond state of the art, OptimESM developed new methods for statistical downscaling based on artificial neural networks of varying topology and tested these methods against reanalysis data of different resolution.
a) Delivering the next generation of high-resolution ESMs
OptimESM work has focused on:
• Increasing model resolution: OptimESM has started to develop tools to run ocean biogeochemistry and atmospheric chemistry at lower resolution than the physical components and to explore methods for more efficient tuning and spin-up of the ESMs. These are first important steps beyond the state of the art towards high-resolution ESMs
• Increasing process realism: OptimESM work so far progressed on the representation of continental ice sheets and their interactions with ice shelves and ocean, permafrost, wildfires and land surface processes and their link to regional climate and CO2 uptake and the risk of dryland expansion in a warming climate.
b) Developing new, policy-relevant emission and land use scenarios
The modeling of land-use changes to represent the forestry sector and the associated carbon stock has been improved. Moreover, along with afforestation, bioenergy with carbon capture and sequestration, and Direct Air Capture option to remove carbon via Ocean Alkalinity Enhancement has been added to the REMIND-MAgPIE model.
A preliminary set of new land use and emission scenarios until 2100 has been derived with REMIND-MAgPIE. Important steps towards the extensions beyond 2100 until 2300 have been undertaken using the small climate model ACC2.
c) Evaluating ESMs and uncertainty in climate projections
OptimESM has completed a compendium of existing observational datasets relevant for identifying the current climate trends and evaluating ESMs. The document outlines the latest observational datasets for Earth system variables relevant to tipping points, and will be kept up to date throughout the project. OptimESM has prepared evaluation scripts and developed new tools and indices to evaluate the models.
d) Delivering new knowledge on potential abrupt climate changes
OptimESM work focused on making an inventory of occurrence of strong nonlinear surprises including possible tipping points in the CMIP6 database with an initial focus on ocean and sea ice. The second focus was on developing new techniques for early warning signals under rapid climate forcing. These early warming techniques have then been tested for sub-polar gyre collapses in CMIP6 simulations.
e) Delivering new knowledge on regional consequences
OptimESM aims to provide new understanding of the regional climate response to long-term global change and to the occurrence of different abrupt changes.
As first steps beyond state of the art, OptimESM developed new methods for statistical downscaling based on artificial neural networks of varying topology and tested these methods against reanalysis data of different resolution.