Periodic Reporting for period 3 - ECCLES (Emergent Constraints on Climate-Land feedbacks in the Earth System)
Période du rapport: 2020-10-01 au 2022-03-31
The carbon cycle is currently carrying-out an important service for humankind by absorbing about a half of our CO2 emissions. However, it is clear from the long-term measurements of CO2 in the atmosphere that the carbon cycle is sensitive to climate. The Earth System Models (ESMs) which appear in the reports of the Intergovernmental Panel on Climate Change now routinely include climate-carbon cycle feedbacks.
Unfortunately, the feedbacks from the land carbon cycle still vary by a factor of more than five, which impedes action to mitigate and adapt to climate change.
There are three major reasons for this uncertainty: (a) uncertainty in the sensitivity of climate to carbon dioxide; (b) uncertainty in the extent to which climate change will lead to a release of carbon from soil and vegetation; and (c) uncertainty in the extent to which CO2 increase will enhance photosynthesis and therefore increase the land carbon sink.
The ECCLES project addresses these uncertainties by searching for Emergent Constraints, which are essentially relationships between observable aspects of the Earth System and projected future changes, which are common across the full ensemble of models.
Emergent constraints are very attractive because they make use of the range of projections of future climate to reduce the uncertainty in the future climate of the real world. However, there is also a danger that misleading relationships could arise from blind data-mining of the multidimensional outputs of state-of-the-art models. ECCLES therefore focuses on emergent constraints that have a firm basis in mathematical theory and known physical mechanisms.
In short, ECCLES sets-out to reduce uncertainties in climate-carbon projections using theory-based emergent constraints.
Unfortunately, the feedbacks from the land carbon cycle still vary by a factor of more than five, which impedes action to mitigate and adapt to climate change.
There are three major reasons for this uncertainty: (a) uncertainty in the sensitivity of climate to carbon dioxide; (b) uncertainty in the extent to which climate change will lead to a release of carbon from soil and vegetation; and (c) uncertainty in the extent to which CO2 increase will enhance photosynthesis and therefore increase the land carbon sink.
The ECCLES project addresses these uncertainties by searching for Emergent Constraints, which are essentially relationships between observable aspects of the Earth System and projected future changes, which are common across the full ensemble of models.
Emergent constraints are very attractive because they make use of the range of projections of future climate to reduce the uncertainty in the future climate of the real world. However, there is also a danger that misleading relationships could arise from blind data-mining of the multidimensional outputs of state-of-the-art models. ECCLES therefore focuses on emergent constraints that have a firm basis in mathematical theory and known physical mechanisms.
In short, ECCLES sets-out to reduce uncertainties in climate-carbon projections using theory-based emergent constraints.
The ECCLES project team have made excellent progress on these aims, as outlined below in terms of the 4 Tasks listed in the ‘Description of Action’ for the project.
Task 1: Theoretical Basis for Emergent Constraints
The emergent constraint technique involves using the full ensemble of models to find an across-ensemble relationship between an observable feature of the Earth System (such as a trend, interannual variation or change in seasonality) and an uncertain aspect of the future. The power of emergent constraints is that they use the enduring range in model projections to reduce uncertainty in the future of the real Earth System, but there are also risks that indiscriminate data-mining and systematic model errors could yield misleading constraints. This Task set-out to mitigate this risk by developing a theory-based approach to emergent constraints.
A first high-profile example of this approach was published early in this project by PI Peter Cox, Mark Williamson (PDRA1) and long-term collaborator Chris Huntingford (Cox et al., Nature, 2018). This study proposed a theory-motivated emergent constraint on ‘Equilibrium Climate Senstivity (ECS)’, using a simple 1-box ‘Hasselmann Model’ to identify a relationship between ECS and a metric of global temperature variability involving both temperature variance and autocorrelation (. It led to an energetic debate amongst climate scientists and four separate ‘Brief Communications Arising’ which were answered by the original three co-authors plus Femke Nijsse (PhD1) who had now joined the ECCLES research group (Cox et al., 2019).
At about the same time, Femke Nijsse (PhD1) published a study showing how decadal temperature variability varies significantly with ECS across the CMIP5 ensemble of Earth System Models (Nijsse et al., Nature Climate Change, 2019). As the study focussed on the unforced control simulations of the CMIP5 models, it was the cleanest demonstration yet of the sensitivity-variability relationship in complex ESMs.
Task 2: Identifying the Key Observed Changes
Work under this task has focussed on Data Analysis and interpretation of forest inventory records which provide data on individual tree sizes (in terms of trunk-diameter or tree mass). The ECCLES Data Analyst Dr Jon Moore (0.5 FTE) has been aided here by PhD student Arthur Argles who is funded directly by the University of Exeter to develop a new model of vegetation demography under the supervision of Prof Cox. Together, Moore, Argles and Cox have developed analytical mathematical solutions for ‘Demographic Equilibrium Theory (DET)’ and compared these to forest inventory data for North America (Moore et al., 2018, see figure 4) and Amazonia (Moore et al., 2020).
Task 3: Emergent Constraints from Temporal Variations
The transient climate response (TCR) is the metric of temperature sensitivity that is most relevant to warming in the next few decades, and contributes the biggest uncertainty to estimates of the carbon budgets consistent with the Paris target. In a recent paper, Femke Nijsse (PhD1) has demostrated an emergent constraint on both TCR and ECS from the trend in global warming since the 1970s (Nijsse et al., submitted).
In the IPCC 5th Assessment Report (AR5), the stated ‘likely’ range of TCR was given as 1 to 2.5K with a central estimate which was broadly consistent with the ensemble mean of the CMIP5 Earth System Models (ESMs) available at the time. Many of the latest CMIP6 ESM have larger climate sensitivities, with 5 of 34 models having TCR values above 2.5K and an ensemble mean TCR of 2.0+/-0.4K. On the face of it, these latest ESM results suggest that the IPCC likely range of TCR may need revising upwards, which would cast further doubt on the feasibility of the Paris targets.
Nijsse et al. (submitted) show that rather than increasing the uncertainty in climate sensitivity, the CMIP6 models help to further constrain the likely range of TCR to 1.3-2.1 K, with a central estimate of 1.68K. This conclusion is reasched through an emergent constraint approach which relates the value of TCR to the global warming from 1975 onwards. A consistent emergent constraint on TCR is derived when the same method is applied to CMIP5 models. These emergent constraints on TCR benefits from both the large range of TCR values across the CMIP6 models, and also from the extension of the historical simulations into a period when the uncertain changes in aerosol forcing have had a far less significant impact on the trend in global warming.
Task 4: Emergent Constraints from Spatial Variations
Rebecca Varney (PhD3) has developed a spatial emergent constraint for the loss of soil carbon due to the acceleration of heterotrophic respiration with global warming (Varney et al., submitted). This builds on the work of Sarah Chadburn (PDRA3) who used a similar method to constrain the loss of permafrost area with warming (Chadburn et al., Nature Climate Change, 2017).
Future changes in soil carbon depend on changes in litter and root inputs from plants and especially on reductions in the turnover time of soil carbon with warming. The latter-term can be diagnosed from projections made with the CMIP6 and CMIP5 Earth System Models (ESMs), and is found to span a large range even at 2K of global warming (-200+/-120 PgC). Varney et al. present a constraint which makes use of the spatial variability of the turnover time inferred from observations. This spatial emergent constraint allows the uncertainty to be more then halved to -230 +/-50 PgC.
Task 1: Theoretical Basis for Emergent Constraints
The emergent constraint technique involves using the full ensemble of models to find an across-ensemble relationship between an observable feature of the Earth System (such as a trend, interannual variation or change in seasonality) and an uncertain aspect of the future. The power of emergent constraints is that they use the enduring range in model projections to reduce uncertainty in the future of the real Earth System, but there are also risks that indiscriminate data-mining and systematic model errors could yield misleading constraints. This Task set-out to mitigate this risk by developing a theory-based approach to emergent constraints.
A first high-profile example of this approach was published early in this project by PI Peter Cox, Mark Williamson (PDRA1) and long-term collaborator Chris Huntingford (Cox et al., Nature, 2018). This study proposed a theory-motivated emergent constraint on ‘Equilibrium Climate Senstivity (ECS)’, using a simple 1-box ‘Hasselmann Model’ to identify a relationship between ECS and a metric of global temperature variability involving both temperature variance and autocorrelation (. It led to an energetic debate amongst climate scientists and four separate ‘Brief Communications Arising’ which were answered by the original three co-authors plus Femke Nijsse (PhD1) who had now joined the ECCLES research group (Cox et al., 2019).
At about the same time, Femke Nijsse (PhD1) published a study showing how decadal temperature variability varies significantly with ECS across the CMIP5 ensemble of Earth System Models (Nijsse et al., Nature Climate Change, 2019). As the study focussed on the unforced control simulations of the CMIP5 models, it was the cleanest demonstration yet of the sensitivity-variability relationship in complex ESMs.
Task 2: Identifying the Key Observed Changes
Work under this task has focussed on Data Analysis and interpretation of forest inventory records which provide data on individual tree sizes (in terms of trunk-diameter or tree mass). The ECCLES Data Analyst Dr Jon Moore (0.5 FTE) has been aided here by PhD student Arthur Argles who is funded directly by the University of Exeter to develop a new model of vegetation demography under the supervision of Prof Cox. Together, Moore, Argles and Cox have developed analytical mathematical solutions for ‘Demographic Equilibrium Theory (DET)’ and compared these to forest inventory data for North America (Moore et al., 2018, see figure 4) and Amazonia (Moore et al., 2020).
Task 3: Emergent Constraints from Temporal Variations
The transient climate response (TCR) is the metric of temperature sensitivity that is most relevant to warming in the next few decades, and contributes the biggest uncertainty to estimates of the carbon budgets consistent with the Paris target. In a recent paper, Femke Nijsse (PhD1) has demostrated an emergent constraint on both TCR and ECS from the trend in global warming since the 1970s (Nijsse et al., submitted).
In the IPCC 5th Assessment Report (AR5), the stated ‘likely’ range of TCR was given as 1 to 2.5K with a central estimate which was broadly consistent with the ensemble mean of the CMIP5 Earth System Models (ESMs) available at the time. Many of the latest CMIP6 ESM have larger climate sensitivities, with 5 of 34 models having TCR values above 2.5K and an ensemble mean TCR of 2.0+/-0.4K. On the face of it, these latest ESM results suggest that the IPCC likely range of TCR may need revising upwards, which would cast further doubt on the feasibility of the Paris targets.
Nijsse et al. (submitted) show that rather than increasing the uncertainty in climate sensitivity, the CMIP6 models help to further constrain the likely range of TCR to 1.3-2.1 K, with a central estimate of 1.68K. This conclusion is reasched through an emergent constraint approach which relates the value of TCR to the global warming from 1975 onwards. A consistent emergent constraint on TCR is derived when the same method is applied to CMIP5 models. These emergent constraints on TCR benefits from both the large range of TCR values across the CMIP6 models, and also from the extension of the historical simulations into a period when the uncertain changes in aerosol forcing have had a far less significant impact on the trend in global warming.
Task 4: Emergent Constraints from Spatial Variations
Rebecca Varney (PhD3) has developed a spatial emergent constraint for the loss of soil carbon due to the acceleration of heterotrophic respiration with global warming (Varney et al., submitted). This builds on the work of Sarah Chadburn (PDRA3) who used a similar method to constrain the loss of permafrost area with warming (Chadburn et al., Nature Climate Change, 2017).
Future changes in soil carbon depend on changes in litter and root inputs from plants and especially on reductions in the turnover time of soil carbon with warming. The latter-term can be diagnosed from projections made with the CMIP6 and CMIP5 Earth System Models (ESMs), and is found to span a large range even at 2K of global warming (-200+/-120 PgC). Varney et al. present a constraint which makes use of the spatial variability of the turnover time inferred from observations. This spatial emergent constraint allows the uncertainty to be more then halved to -230 +/-50 PgC.
We hope to find further new emergent constraints in the remaining 2.5 years of the ECCLES project. A particular target will be to look for an emergent constraint on the carbon budgets consistent with the Paris climate targets of 1.5K and 2K of global warming above the pre-industrial climate.
We also plan to build on our attempts to build a generic underlying theory for emergent constraints.
Finally, we are aiming to build a bridge between our work on emergent constraints and attempts to forewarn of forthcoming climate tipping points.
We also plan to build on our attempts to build a generic underlying theory for emergent constraints.
Finally, we are aiming to build a bridge between our work on emergent constraints and attempts to forewarn of forthcoming climate tipping points.