EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Towards multi-method and multi-scale attribution of global wildfire danger

Zhongwei Liu1, Jonathan Eden1, Bastien Dieppois1,2, and Matthew Blackett1
Zhongwei Liu et al.
  • 1Centre for Agroecology, Water and Resilience, Coventry University, UK (
  • 2Department of Oceanography, MARE Institute, University of Cape Town, RSA

Wildfires constitute a major natural hazard and pose huge risk to many regions of the world. The series of large fires across both hemispheres in recent years have led to inevitable questions about how human-induced climate change may be altering the character of such events. Providing answers to these questions is a crucial step to increasing resilience to major wildfires.

Long-term projections produced by state-of-the-art climate models, even when reliable, are not always a suitable means of communicating risk. Methodologies to attribute trends in meteorological phenomena associated with high-impact events to anthropogenic influence have the potential to better communicate risk and guide adaptation strategies. While the link between a warming world and heat-related extremes (e.g. heatwaves and droughts) is reasonably well-understood, there is a lack of consensus on the most appropriate and effective methodological approach for many variables, potentially impacted by warming climate, such as wildfire attribution. The link with climate change remains poorly understood and wildfires have been largely ignored by attribution studies to date.

As a first step towards the development of a seamless, globally-applicable framework for assessing past, present and future risk in wildfire danger, we present a global attribution analysis of wildfire danger. With initial focus on observational records, we use both established and novel empirical-statistical methods to attribute historical trends in episodes of extreme weather and climate conducive to wildfire ignition and spread. Particular consideration is given to the sensitivity of attribution findings to the spatial scale upon which the analysis is conducted. We also draw attention to a series of important, often overlooked, conceptual and technical challenges in event attribution, including validation and bias-correction of climate models and discuss the value of linking attribution of recent wildfire events with future risk assessment.

How to cite: Liu, Z., Eden, J., Dieppois, B., and Blackett, M.: Towards multi-method and multi-scale attribution of global wildfire danger, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20734,, 2020

Comments on the presentation

AC: Author Comment | CC: Community Comment | Report abuse

Presentation version 1 – uploaded on 07 May 2020 , no comments