Earth System Modelling of Abrupt Climate Change

Lead Research Organisation: University of Bristol
Department Name: Geographical Sciences


Recent studies of abrupt climate change have tended to focus on rapid cooling events related to an abrupt cessation or reduction in the Atlantic meridional overturning circulation. However, ice core records from Greenland which span the last glacial to interglacial cycle (approximately 100,000 years) indicate that abrupt warmings of the order of 10C, which took place in a few decades, are more representative of the past climatic record. Furthermore, although a clear understanding is still lacking, recent modelling efforts suggest that atmospheric dynamics could be more important in shaping these abrupt events than previously thought. We propose therefore to use an Earth System climate model which can quantitatively simulate multiple proxies recorded in Greenland ice (including temperature, methane and dust) to assess which patterns of forcings can be reconciled with these multiple ice core data constraints. We will perform sensitivity analyses to assess the potential for thresholds of abrupt atmospheric transitions and investigate their reliance on glacial topography and background climatic state. This project will therefore be the first to synthesise a number of ice-core records of abrupt climate change in a single, quantitative general circulation modelling framework and the results will place significant new constraints on our understanding of abrupt climate change in the past, and potentially in the future.


10 25 50
Description This project focussed on three key areas, (i) the methane and (ii) the dust cycles under ice-age climatic conditions and abrupt warming events of the last ice-age and (iii) the use of palaeoclimate states for the evaluation of Earth System models. The majority of the project involved the use of the HadGEM2-ES Earth System model, which was widely used in CMIP5 (Coupled Model Intercomparison Project 5) and hence in the IPCC (Intergovernmental Panel on Climate Change) 2013 report. This meant that the findings of this project are relevant to the broader future climate projection and Earth Systems scientific communities.
Our findings have highlighted that palaeoclimate states, in this case the Last Glacial Maximum (LGM) can serve as extremely useful benchmarks for Earth System models. We found that HadGEM2-ES provided a particularly poor representation of the LGM climate and vegetation state, and recommended changes in the model to resolve this (Hopcroft & Valdes, 2015, Climate Dynamics). A unique result of our work is that these changes have been incorporated into both JULES (the Joint UK Land Environment Simulator) and the developmental version of UK Earth System Model (UKESM). Additionally, our work has led to the UKESM team beginning a palaeoclimate evaluation exercise prior to the official release of this new flagship model, something that was not planned originally part of the UKESM project.
We have also demonstrated in a comprehensive climate-composition modelling framework that the observed change in the atmospheric methane cycle between the LGM and the pre-industrial era is not properly understood (Hopcroft et al, in revision). This finding strongly implies that current models of the natural sources of methane do not show the correct sensitivity, and we have highlighted several missing processes that need to be investigated to resolve this.
As a first step in this direction we estimated the methane isotopic budget for the LGM and pre-industrial (manuscript in preparation). We find that both the methane concentration and methane isotope change between the two time-periods is underestimated by process-based models. We used a probabilistic framework to identify solutions that can explain both the concentration and methane isotope change in a manner consistent with the observations and to some extent the models. The results support the conclusions that current source models of methane may be under-sensitive.
Our work on mineral dust aerosols (Hopcroft et al 2015, Journal of Geophysical Research) has shown that the nature of the size distribution of dust particles during emission from the land surface is critical for determining the resultant change in dust radiative forcing in altered climates (such as the ice-age, or potentially the future). This may explain a substantial fraction of the range in dust aerosol radiative forcing estimates found in previous studies of the glacial-interglacial change. This also provides a new emphasis on evaluating the models in terms of the size distribution of dust particles at emission.
Exploitation Route The main findings of this project will feed into three separate areas of research in future. Firstly, work is already underway to evaluate the new Earth System model (UKESM) against paleoclimate reconstructions for the LGM. More generally the integration of the Paleoclimate Modelling Inter-comparison Project into CMIP means that climate models used for projections are also evaluated for past conditions. Future work in this area could make use of our findings concerning Earth System model components and their likely feedback effects and associated uncertainties (e.g. Hopcroft & Valdes, 2015, Geophysical Research Letters).
Future work should build on our evaluation of the LGM methane cycle to better understand what is responsible for the natural fluctuations in atmospheric methane more generally. This will involve building and testing new process-based representations of wetlands and other natural methane sources. Our work suggests that the LGM climate state is useful for evaluating the methane cycle because of a large change in atmospheric methane, and the absence of complications arising from anthropogenic activities.
Natural extensions to our work on dust will be possible with the new UKESM model which will include the dust-cloud interactions and a fully prognostic representation of sea-salt aerosols, both of which are missing from our work. Inclusion of the latter may lead to improved estimates of aerosol radiative forcing in past climates like the LGM, and provide clues about the patterns of change, including the issue of polar amplification during LGM.
Sectors Environment

Description UK Met Office 
Organisation Meteorological Office UK
Country United Kingdom 
Sector Academic/University 
PI Contribution This project has helped foster partnerships with the UK Met Office, with one collaborative paper published and another in preparation.
Collaborator Contribution This project has helped foster partnerships with the UK Met Office, with one collaborative paper published and another in preparation.
Impact Publication and one in preparation
Start Year 2015
Description UKESM 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Study participants or study members
Results and Impact Results from this project have focussed attention within the UK community on the utility of palaeo-climates as a method of evaluation of Earth System models. This is because of the extremely poor performance of HadGEM2-ES when simulating the climate of the last glacial maximum (LGM) demonstrated in this project. These results for the LGM have now encouraged the inclusion of paleoclimate evaluation of the UKESM (UK Earth System Model) at the UK Met Office in collaboration with the BRIDGE research group. For this, UKESM will be evaluated against the climate and Earth System responses reconstructed for the mid-Holocene and LGM. This project also led to changes in the land surface scheme JULES (Joint UK Land Environment Simulator).
Year(s) Of Engagement Activity 2013,2014,2015,2016