The Aerosol-Cloud Uncertainty REduction project (A-CURE)

Lead Research Organisation: University of Leeds
Department Name: School of Earth and Environment

Abstract

A-CURE tackles one of the most challenging and persistent problems in atmospheric science - to understand and quantify how changes in aerosol particles caused by human activities affect climate. Emissions of aerosol particles to the atmosphere through industrial activity, transport and combustion of waste have increased the amount of solar radiation reflected by the Earth, which has caused a cooling effect that partly counteracts the warming effect of greenhouse gases. The magnitude of the so-called aerosol radiative forcing is highly uncertain over the industrial period. According to the latest intergovernmental panel (IPCC) assessment, the global mean radiative forcing of climate caused by aerosol emissions over the industrial period lies between 0 and -2 W m-2 compared to a much better understood and tighter constrained forcing of 1.4 W m-2 to 2.2 W m-2 due to CO2 emissions. This large uncertainty has persisted through all IPCC assessments since 1996 and significantly limits our confidence in global climate change projections. The aerosol uncertainty therefore limits our ability to define strategies for reaching a 1.5 or 2oC target for global mean temperature increase.

A-CURE aims to reduce the uncertainty in aerosol radiative forcing through the most comprehensive ever synthesis of aerosol, cloud and atmospheric radiation measurements combined with innovative ways to analyse global model uncertainty. The overall approach will be to produce a large set of model simulations that spans the uncertainty range of the model input parameters. Advanced statistical methods will then be used to generate essentially millions of model simulations that enable the full uncertainty of the model to be explored. The spread of these simulations will then be narrowed by comparing the simulated aerosols and clouds against extensive atmospheric measurements.

Following A-CURE, improved estimates of aerosol forcing on regional and global scales will enable substantial improvements in our understanding of historical climate, climate sensitivity and climate projections. We will use the improved climate model with narrowed uncertainty to determine the implications for reaching either a 1.5 or 2oC target for global mean temperature increase.

Planned Impact

A-CURE will have societal and economic impact primarily through improved climate models and in particular the accuracy and reliability of the projections they make.
We have identified four potential pathways to impact. The Pathways to Impact document describes how we will carry these out.

1. Assessing the impact of reduced aerosol forcing uncertainty on near-term climate projections and attainment of Paris Agreement of 1.5oC target for global mean temperature change. This is A-CURE objective 2. We have built this objective into A-CURE because it is clear that aerosol forcing uncertainty will directly impact our confidence in the climate projections that underpin any assessment of the target. The Met Office letter of support states that "A reduction in the uncertainties associated with aerosol-cloud-climate interactions will make a major contribution to an overall reduction in the uncertainty of future Earth system projections and offer more reliable information to support UK and international policy decision-making directed towards both climate change mitigation and adaptation."

2. Informing the UK Climate Projections in 2018 (UK18). The Met Office will produce climate data that will inform the UK's 3rd national Climate Change Risk Assessment (CCRA). Although A-CURE will not have a direct impact on the next CCRA model simulations (these are already underway), the results of A-CURE will provide important updated and refined information on the plausibility of aerosol forcing effects in the UK18 projections. One outcome is that the narrowed uncertainty range of aerosol radiative forcing from A-CURE will enable users of the UK18 projections to put the data in better context. For example, it would be feasible to say whether projected strong changes in near-term temperature response are more or less plausible based on updated knowledge about aerosol forcings.

3. Impact of new model-observation metrics on climate projection systems, both in the UK and more widely. A longer-term benefit this work beyond 2018 is that model-observation comparisons are vital for assessing many aspects of the physical climate and its responses (e.g. clouds, convection, surface processes). Observational metrics related to aerosol processes are comparatively under-developed. We would aim to assess the value of new aerosol and cloud metrics that could be incorporated into the development of future climate projections beyond 2018. This would place the uncertainty estimates in these future projections on a more robust footing.

4. Long-term improvements in climate models. We aim for A-CURE to leave a long-term legacy of improved climate models able to make more reliable simulations of aerosol effects. To assist the Met Office in this task, we will provide easy-to-use model emulators that will enable very efficient exploration of model tuning options in UKESM and evaluation against measurements. We will dedicate time from the CEMAC researcher at Leeds to generate a web tool.

Publications

10 25 50
 
Description We have worked out how much the uncertainty in aerosol radiative forcing can be reduced by using an extensive set of measurements to constrain the behaviour of the UK global climate model. The study used 1 million "variants" of the HadGEM3-UKCA climate model that sample the combined effects of 26 sources of uncertainty. We then used over 9000 measurements of aerosol optical depth, PM2.5, particle number concentrations, sulphate and organic mass concentrations to find model variants that were consistent with the measurements. This enabled us to reduce the number of "plausible" variants to a few thousand.

We are able to constrain the likely range of uncertain model parameters related to secondary organic aerosol, anthropogenic SO2 emissions, residential emissions, sea spray emissions, dry deposition rates of SO2 and aerosols, new particle formation, cloud droplet pH and the diameter of primary combustion particles.

The standard deviation range of global annual mean direct radiative forcing, RFari, is reduced by 33 % to -0.14 to -0.26 W m-2 and the global annual mean aerosol-cloud radiative forcing, RFaci is reduced by 7 % to -1.66 to -2.48 W m-2.

We have been able to establish which observations provide the greatest constraint on the model uncertainty.

We have quantified the effect of model structural errors and demonstrated that focusing on this source of model error has the potential to enable substantial reductions in radiative forcing uncertainty. We developed a new approach to detect such structural errors, which we will apply in an upcoming NERC project.
Exploitation Route The results can be used by others to decide how to use different aerosol measurements to constrain other climate models. Modelling centres that do not quantify the uncertainty in their model can use our results to estimate what the likely uncertainty will be.
Sectors Environment

 
Description Towards Maximum Feasible Reduction in Aerosol Forcing Uncertainty (Aerosol-MFR)
Amount £704,077 (GBP)
Funding ID NE/X013901/1 
Organisation Natural Environment Research Council 
Sector Public
Country United Kingdom
Start 08/2023 
End 08/2026
 
Title Data for submitted publication: 'Contrasting responses of idealised and realistic simulations of shallow cumuli to aerosol perturbations' 
Description Data for producing figures from the paper 'Contrasting responses of idealised and realistic simulations of shallow cumuli to aerosol perturbations', submitted to Geophyiscal Research Letters in February 2021. 
Type Of Material Database/Collection of data 
Year Produced 2021 
Provided To Others? Yes  
URL https://zenodo.org/record/4513730
 
Title Software for "Building high accuracy emulators for scientific simulations with deep neural architecture search" 
Description This is the code and datasets for "Building high accuracy emulators for scientific simulations with deep neural architecture search". 
Type Of Technology Software 
Year Produced 2020 
Open Source License? Yes  
URL https://zenodo.org/record/3782843
 
Title Software for "Building high accuracy emulators for scientific simulations with deep neural architecture search" 
Description This is the code and datasets for "Building high accuracy emulators for scientific simulations with deep neural architecture search". 
Type Of Technology Software 
Year Produced 2020 
Open Source License? Yes  
URL https://zenodo.org/record/3782842