New analysis provides “proof of concept” for real-time extreme event attribution

A new analysis published in the journal Environmental Research Letters establishes that seasonal forecast sea surface temperature (SSTs) can be used to perform probabilistic extreme-event attribution, thereby accelerating the time it takes climate scientists to understand and quantify the role of global warming in certain classes of extreme weather events.

To test the theory that seasonal forecasts SSTs would perform as well as observed SSTs and result in very similar attribution statements, scientists from the CPDN team compared weather@home in forecast mode (forced with seasonal forecast SSTs) relative to hindcast mode (forced with observed SSTs) as part of a model validation effort. Specifically, seasonal forecast SSTs (GloSea5) from the UK Met Office were used to simulate recent UK heavy rainfall events. These simulations were compared with the observed record.

“This analysis proves that the modelling approach behind one of the main pillars of real-time event attribution undertaken is sound and applicable to a range of extreme events in regions where we have confidence in the model performance”, said Karsten Haustein, lead author of the new paper.

“With this approach we answer the attribution question in a manner that is more relevant to potential stakeholders. Specifically, we can assess the influence of human-induced warming, given the predictable climate in the current season rather than exactly the observed situation”, said Friederike Otto  – a co-author of the new study and science lead of the ‘World Weather Attribution’ (WWA) initiative.

Watch Karsten and Fredi explain the study!

The analysis used both a case study on the UK floods that occurred during the winter of 2013/2014 as well as Oct 2015 to account for effects from El Nino. The study revealed that there is no difference in the quantitative assessment of the attribution statements of temperature in the two approaches. For precipitation, using forecast SSTs leads to an attribution statement of the same order of magnitude but quantitatively on the more conservative side. Using seasonal forecast SSTs instead of observations to drive the regional climate model allows scientists to disentangle the influence of low-frequency climate modes like El Nino from human-induced warming. This is particularly relevant in tropical climates but allows for an estimate of the component of natural variability in a given extreme event also in extratropical regions such as the UK.

This powerful capability will be applied to a new initiative using state-of-the-art science to help Asian and African societies to understand the role of climate change in extreme weather events and prepare for future ones.

The ‘Raising Risk Awareness’ initiative brings together scientists from the WWA initiative – an effort led by Climate Central with the Red Cross Red Crescent Climate Centre, University of Oxford, University of Melbourne and Royal Netherlands Meteorological Institute – with the Climate and Development Knowledge Network (CDKN). It will assess whether climate change has contributed to extreme weather events such as droughts, floods and heatwaves in several countries across East Africa and South Asia.

The team will use peer-reviewed methods, including this paper, to discover the links – and will distribute the information widely to the press, policy-makers and the public in these countries.

Heidi Cullen, Chief Scientist at Climate Central, said: “It makes a real difference to policy-makers and planners if they can find out quickly after a disaster whether such extreme events are becoming more or less frequent. Some decisions about recovery and reconstruction need to be made within days or weeks.

The ‘Raising Risk Awareness’ project launched in May 2016 and will run until March 2017

Receive our future bulletins

To register your contact details with WWA so that you can receive timely information about the role of climate change in future extreme weather events, please write to with your request.


Reference: Haustein et al 2016 Environ. Res. Lett. 11, 064006, doi: 10.1088/1748-9326/11/6/064006

« Back to News