Producing regional climate narratives to make better predictions of future climate

An interdisciplinary team of social scientists, physical climate scientists, statisticians and communication experts have developed a new method of uncertainty quantification for regional climate information that combines expert judgement with climate observations and simulations.

The Indian summer monsoon impacts the societal resilience of a large population in Southern India. However, many important processes are poorly reproduced by global climate models. To address this, a combination of expert-derived climate narratives with climate observations and simulations was used to understand and project how regional climate in Southern India could change in the future.

In July 2015, experts on the Indian summer monsoon from the UK and India were brought together to determine how the climate in Southern India could change in the near to mid-term future (2030s and 2050s). With a focus on the Cauvery River Basin and the mountain range of the Western Ghats (as its source), experts agreed that the most important climatic processes determining river flow were moisture availability over the Arabian Sea and strength of air flow perpendicular to the Western Ghats. From this, a number of narratives were developed with the experts describing possible future evolutions of the Indian summer monsoon and underlying plausible processes. The relationship between these two key processes (moisture availability and flow) and rainfall and river flow in the catchment were verified using data from observations and climate simulations. Using these relationships, time series of future regional climate information were created for each expert-based climate narrative.

• Expert-derived climate narratives can be combined into time series of regional future climate change that can be used as part of a climate risk assessment to inform long-term adaptation decision-making.

Suraje Dessai, Professor of Climate Change Adaptation

Cathryn Birch, University Academic Fellow in dynamical meteorology and high-resolution modelling of weather and climate

John Paul Gosling, Associate Professor of Statistics