Attributing impacts of external climate drivers on extreme weather in Africa

Given limited progress in reducing greenhouse gas emissions and uncertain potential for adaptation to many impacts, attention in vulnerable regions and sectors is turning to the question of "loss and damage". Who should bear the costs of human influence on climate that cannot be neutralized by adaptation? This debate is impeded by lack of robust estimates of what these costs are. Despite concerted efforts to compile inventories of emissions, we still have no agreed method of establishing how countries, companies or individuals are being adversely affected by anthropogenic climate change in the context of other drivers of regional environmental change.

Many of the most important impacts of climate change are related in some way to high-impact weather events (HIWEs), such as floods, storms, and droughts. Compiling an impact inventory requires documenting the impacts of individual events and how these events are affected by multiple climate drivers and internal climate variability. We will build on research into HIWEs and their impacts under THORPEX-Africa.

Studies assessing the link between climate change and extreme weather have so far focused primarily on mid-latitude phenomena and the impact of rising greenhouse gases. Yet in many tropical regions, short-lived climate forcings (SLCFs) such as sulphate, mineral and black carbon aerosols and tropospheric ozone may have played a larger role in changing patterns of weather risk to date. Substantial reductions in anthropogenic SLCFs could be achieved in only 20 years. Including measures already planned to reduce emissions of sulphate aerosol precursors, SLCFs may dominate near-term changes in weather risk. Climate impact assessments used for adaptation planning typically focus on net multi-decadal anthropogenic change, dominated by greenhouse-induced warming. Few address uncertainty in SLCF forcing and response. Hence relying on these and extrapolation of recent trends risks "adapting to yesterday's problem" as key drivers of regional weather are reversed.

Assessing the influence of external drivers on extreme weather is challenging because the most important events are typically rare. The only solution is to rely on simulation models, whose reliability can be tested and if necessary re-calibrated using well-established procedures developed for seasonal forecasting. We will also use the land-surface model JULES for indirect validation in regions with sparse meteorological data. Large ensembles of climate model simulations at relatively high resolution are required for robust statistics of extreme weather events, allowing for uncertainty in both external drivers and simulation models.

This project makes use of the climateprediction.net weatherathome worldwide volunteer computing project. We will quantify the role of various external climate drivers on changing risks of extreme weather in Africa by implementing a regional climate model over the CORDEX-Africa domain and simulate observed weather statistics over recent decades using multi-thousand-member ensembles, systematically excluding the influence of different climate drivers to quantify their effects.

Attribution studies of HIWEs to date have typically focussed on hydrometeorological events themselves, rather than modelling all the way through to their impacts. This can lead to "over-attribution": if a record-breaking weather event occurs that has been made more likely by some external driver, people tend to blame most of the impact of that event on that driver. But much of this impact might also have been caused by a lesser, non-record-breaking, event. Hence accurate assessment requires explicit modelling of changing impact risk, not simply weather risk, so a major focus of this project will using JULES to investigate various impacts and working with impact modellers across Africa to assess the implications of our weather simulations for changing impact risk in other sectors.

End-to-end attribution of climate change impacts is a pressing issue not only for the for the Intergovernmental Panel on Climate Change but especially for users of climate information in farming, decision makers for infrastructure development and adaptation. The interdisciplinary nature of the problem makes co-generation of knowledge essential: attribution of hydro-meteorological anomalies alone is insufficient to demonstrate impact. Calculation of actual harm requires a deep understanding of local confounding factors and changing vulnerability. We propose to address this through demonstration projects (attribution of floods in Kenya, droughts in East Africa and vegetation changes over parts of the continent) and providing funding for two workshop to initiate a network of collaborations within Africa using the established links within THORPEX Africa, the consortia of Regional Climate Outlook Forums (RCOF) in Eastern and Western Africa, and Oxfam projects in drought and flood prone regions of the African continent. At these workshops we aim to bring together a large group of impact modellers using the data provided by the proposed research in Oxford to start the process of compiling an inventory of climate impacts over Africa.

Valuing anthropogenic modification of weather and climate is the ultimate goal of this research. On the ten-year research horizon, there is a clear need for a global inventory of the impacts of climate change to complement current inventories of climate drivers such as national emissions data. This will provide an essential ingredient for evidence-based mitigation and adaptation policies, but will need a firm grounding in the link between climate and extreme weather, drawing on routine attribution assessments which address all weather events, not simply high-profile disasters. This is an especially pressing requirement in Africa which is expected to be hard hit by impacts of climate change in the very near future, is least resilient to changing risks of extreme weather events, and has an underdeveloped inventory at present as impact assessment depends on understanding of meteorological mechanisms and impacts of hydro-meteorological events on agriculture, infrastructure and livelihoods. This will require a sustained international effort: this project will develop and prove some of the necessary methods and demonstrate their feasibility and value in most vulnerable regions of Africa. The workshops are planned to initiate an inventory of climate change impacts from within the African continent itself.

With this close collaboration with established facilitators of climate information for and from Africa as THORPEX and RCOF and linking this to impact assessment in Oxford and provided by Oxfam, we develop also a bottom-up approach to quantifying the actual costs (and benefits) of climate change over Africa which represents a radical departure from the top-down approach of typical Integrated Assessment Models used to calculate the Social Cost of Carbon, which parameterise the net impact of extreme weather, providing limited scope for disaggregation of costs and benefits. In effect, we are addressing the question "what did human influence on climate actually cost me (my company or country) last year?" as opposed to "what would we expect the cost of anthropogenic climate change to be in a typical year of the current decade?"

Additionally we plan to further develop citizen science capacity through collaborations between members of climateprediction.net and modellers in the project team, supported by information provided on climateeducation.net, to understand how climate modelling works, examine evidence of the impact of human influence on climate and inform the attribution of extreme weather events to anthropogenic climate change.