The stratospheric impact on extreme weather events

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


The winter of 2013/2014 has seen some of the coldest temperatures in the USA, and some of the heaviest rainfall in the Southern UK ever reported. Both are examples of extreme weather events, the impact of which is enormous on national infrastructure and economy. Current research suggests that the frequency of extreme events is increasing, with strong indications this may be due to humans pumping greenhouse gases into the atmosphere. While the weather on Earth occurs within the troposphere (the layer of the atmosphere between ~0-10km), the stratosphere (~10-50km) is known to play an important role in driving such weather events, especially during winter. The purpose of this project is to understand exactly how the stratosphere influences extreme weather events, and how this follows through to impacts on society, such as infrastructure damage, which can run into billions of pounds, and human deaths, which can be in the tens of thousands.

As climate scientists, one tool available to us in understanding atmospheric dynamics and extremes, are computer models of our current and future climate. These models are similar to those that produce the weather forecasts seen on TV, but are adapted to be relevant for longer timescales. In this project the 'weather@home' framework will be utilised, which allows for the model simulations to be run on thousands of volunteered home computers, spreading the load and allowing for many more simulations to be run than otherwise possible. As the climate system is inherently chaotic, it is unlikely that the models will reproduce exactly the observed climate. Therefore typically on supercomputers, models are run ~5-10 times over the same time period, with slightly different starting conditions each time, to build up an ensemble of 'possible' climate scenarios. The advantage of the weather@home set up is that the models can be run many more times than this (of the order of 10,000). This is essential when considering extreme weather events, because they are by definition rare, and the events will only be captured in ~5% of the model simulations.

Currently, the weather@home set up does not include a stratosphere-resolving model. This is a big issue because without the stratosphere providing a source of variability for extreme events at the surface, attribution statements, predictability, and ultimately estimated societal impacts will be biased. In this project the model will be further developed to include higher vertical resolution, allowing for a well-resolved stratosphere. Initial assessment will be to analyse the stratospheric dynamics that lead to these rare events at the surface. While it has been known that the stratosphere can influence the troposphere for over a decade, the driving conditions behind this are largely unknown. How the stratosphere impacts surface extreme climate is largely unknown, with most previous research focussing on the mean-state. With the shear number of model simulations available to capture internal climate variability, this project is ideally aimed at addressing questions relating to extremes.

An advantage of using climate models (in conjunction with observations) is that they allow for experiments that cannot be tested in the real world. Such as what the climate would be like if humans never existed. In this project such experiments will also be performed to assess how the climate can be forced from different sources of variability. Experiments will be set up to look at how stratosphere-troposphere coupling, and subsequently the extreme events are influences by human induced climate change. Given the recent increase in extreme event attribution globally (for instance, the recent National Academy of Sciences review panel on this issue), an increased understanding of the processes driving extreme events will be invaluable for the scientific community.

Planned Impact

Other than the obvious advancements to the academic community, this project will directly benefit operational centres, policy makers and the public community.

Operation centres: The NERC funded Centre for Ecology and Hydrology (CEH) has expressed a keen interest in the research output from this project (see letter of support). Clearly with an increased understanding of how the stratospheric state 'changes' extremes events, the skill of prediction statements of flooding or drought impacts can be significantly enhanced during winter. CEH's work on the initial Climate Change Risk Assessment and the UKCP09 project sets them up well to incorporate this information, and to go beyond what they have done in the past. It is foreseen that CEH would start to feel the benefits of this research during the timescales of the project, with clear on-going benefits as the techniques are developed.

CEH and the UK Met Office already work closely, and as both are project partners I will have the advantage of their already strong cross-collaboration. This is especially true with regards to extremes (for instance the widely cited paper on the Southern UK floods of 2013/14; Huntingford et al, 2014). Utilising the engagement programmes of both partners is clearly a strong way to influence policy. For instance, on UK public policy (e.g. the Met Office's annual 'get ready for winter' programme), the insurance industry (extreme climate event attribution), or the quality of life of the UK public (e.g. 'snowy' winters). Such incorporations into forecasts would be an expected benefit on longer timescales (i.e. after the initial project finishes).

Policy makers: One of the most frightening impacts of climate change is the possibility of enhanced extreme weather events. Many aspects of political decision making involve interpreting the scientific evidence presented. As such, reliable estimates of probabilistic event attribution are required in order to accurately inform these decisions. Research from this project will directly contribute to the reliability of probabilistic event attribution and therefore aid in political outcomes. This would be a foreseeable benefit over longer timescales (i.e. after the initial project finishes), as the reliability of event attributions improves over time, from the results of this project.

The public: Extreme events are currently at the forefront of public concern; mainly as their coverage and linkage to climate change has increased in the recent years. For instance during the 2013/14 winter, the flooding event over the southern UK, and extremely cold temperatures over North America, both prompted huge media coverage. By increase our understanding or both the underlying mechanisms, and the impacts of such extreme events (in the troposphere and stratosphere), this project will clearly be attractive to media outlets and more generally the public community. It is expected that the benefits on these communities will be felt within the span of the project.


10 25 50