Project concepts

This product is a first attempt to address in real-time the question:
"Are our greenhouse gas emissions to blame for this weather event?"

What is the attribution forecast?
The Weather Risk Attribution Forecast is a product run in parallel with the seasonal forecast produced by the Climate Systems Analysis Group (CSAG) at the University of Cape Town. Along with the standard "real world" seasonal forecast, a forecast is produced for a hypothetical "non-GHG world" in which human activities had never emitted greenhouse gases. These forecasts are then compared in order to estimate the contribution of greenhouse gas emissions to the risk of unusual weather events.

The approach is inspired by work conducted by, the Met Office Hadley Centre, and Risk Management Solutions, in particular a pilot study described here and here.

The rationale
"Are our greenhouse gas emissions to blame for this weather event?" As everyone becomes increasingly aware and concerned about climate change this question keeps popping up. There are three reasons for this.
  • Understanding: How does the present climate relate to what has occurred in the past and to what might occur in the future? Even though climate change is generally considered an issue of the past and particularly of the future, people live in the present. In order to better understand what is likely to happen in the future, we feel the need to understand how the present is changing and how it relates to the past and future climates.
  • Risk assessment: Do we really know how likely those damaging weather events are? In climate change simulations of climate models it takes a couple of decades or so for the climate to noticeably change. But this does not mean that we do not need to worry about adapting to climate change for a couple of decades. Our estimates of current climate generally come from observational measurements taken over the past several decades. Therefore, these estimates are not actually of the current climate but rather of a past that may no longer be relevant. Is the so-called 1-in-100 year hot spell, for instance, much more likely now than we think it is? Is the 1-in-100 year cold snap less probable? Obviously, the insurance industry is concerned about this.
  • Damage and liability: Who receives funding for adaptation to climate change? This is of particular importance for Africa, a supposed recipient of much of this money. International adaptation funding mechanisms are shifting in part to a "loss and damage" approach, whereby funding depends on the existence of damage attributable to greenhouse gas emissions. How is this mechanism to be informed?

A gap in information
Unfortunately, so far most climate research organisations have not studied this problem, focussing on the past and future rather than the present. In large part, this has arisen because of a focus on the mitigation problem, reducing emissions of greenhouse gases so that the future does not get too extreme, rather than the adaptation problem, dealing with the current climate. Adaptation activities have had to make do with climate products designed for informing mitigation activities.

Only recently have some research projects started examining the attribution of weather risk for a few recent iconic events, such as the record hot summer of 2003 in Europe. These have been all targeted studies, each examining one selected event in detail, but not considering events systematically.

The goals
This project has a several goals aimed at filling this gap in information needed for adaptation activities.
  • To develop a framework for producing information about how anthropogenic emissions are affecting our current weather;
  • To interact with users of the product and further develop its application;
  • To elucidate out how this sort of product is interpreted by potential users;
  • To determine the sensitivity of estimates to different forecast models and forecast configurations;
  • To provide indicative information required for adaptation, highlighting cases for more detailed study.

This is currently an experimental pilot product, provided as-is. No guarantee is given of its accuracy. Still, it should provide some qualitative indication of the effect of our greenhouse gases, we just do not know at the moment exactly how accurate it is. In that light, its main operational use should be in highlighting directions for more rigorous study.

What next?
So far we have just considered the probability of extreme weather events. But we would really like to live up to the name of the project and consider risk. To do that we are starting work on moving beyond climate variables and on examining the importance of the role of changes in probability due to climate change in the full risk context. This will involve quite a bit of work, so we welcome collaboration on all fronts.

In order to address our goals, we require feedback from visitors. Is the information provided here of interest to you? Is the information useful to you? How did you find the presentation of the "attribution forecast"? How could it be improved? (Also, what should we not do?!) Right now we only consider the effect of emissions of greenhouse gases: should we instead consider the effect all human emissions, for instance including the effect of aerosols (smog)? Please e-mail your comments to Dáithí Stone at

The team
Oliver Angélil, Chris Lennard, Dáithí Stone, Mark Tadross, Michael Wehner, Piotr Wolski


Last update: 12 October 2013
Contact: Dáithí Stone (

© Copyright 2009-2013 Contributors to the Weather Risk Attribution Forecast

Publications, etc.

Data portal
Much of this data is now available on the NERSC Earth System Grid Federation node (search for "C20C")

* NOAA's Climate Attribution page
* UCAR's Attribution of Climate Events page
*'s Seasonal attribution experiment