Using new high-resolution ensembles to quantify uncertainty in projections of African climate processes

Background

Global models underpinning future projections struggle to capture African weather and climate change, to a large extent because the convective storms that generate tropical rainfall are below their coarse grid scale and have to be parameterised. Convection-permitting (CP) simulations using km-scale grids significantly reduce errors, but presently only cover short realisations of a single climate scenario. Quantifying risk requires a multi-model ensemble that covers a range of scenarios and accurately captures the atmospheric processes key to representing how weather evolves under climate change. Project partner the Met Office are thus running the first ever tropical CP ensemble, CP4A, driven by a range of global models. The open challenge is to understand how we can combine information from such models with global simulations, observations and theory to give accurate ranges of uncertainty on future projections of user-relevant quantities, such as extreme rain or monsoon onset.

PhD opportunity

This project will explore how physical understanding from high-resolution ensembles run for a decade or less can be combined with coarser-grid global or regional climate projections, run for many decades, to improve estimates of climate uncertainty. We will focus on West Africa, where deep convective storms dominate the regional climate and are often coupled to African Easterly Waves (AEWs), synoptic-scale features which influence Atlantic hurricane development. Rainfall intensity for such storms scales with environmental factors, such as wind-shear and moisture, themselves modulated by AEWs and other tropical wave modes. We will evaluate these mechanisms in CP4A, and quantify the spread in future environmental conditions and changes in wave properties (eg periodicity, magnitude, latitude). Together, this will enable a comprehensive process-based understanding of projected changes to rainfall and wave activity in the high-resolution ensemble, and of the couplings between them.

Building from our insights from the CP4A ensemble, we will interrogate the responsible mechanisms in global models, and synthesise the understanding derived from these complementary data sources to provide improved estimates of risks and uncertainties. This will enable updates, and we anticipate improvements, to user-relevant information, whose metrics (eg rainfall extremes, monsoon onset) can be chosen based on the project’s development. There is also scope to apply analysis from the high-resolution ensemble to explore and test methods such as machine learning climate downscaling techniques. The project offers an opportunity to combine fundamental atmospheric dynamics with a robust analysis of future climate uncertainty, building expertise at the intersection of climate and weather science. The results will directly feed into model evaluation processes at the Met Office, and contribute to improving tropical climate projections.