What are convection-permitting models and how can they improve understanding of extreme weather in Africa? This is the second blog in a series related to advances in convection-permitting models. In the first post, we examined how improvements in climate models can help decision makers. Users of weather and climate information are eager to know how the weather and climate will change in their specific locations. They are particularly interested in extreme events, such as thunderstorms, hailstorms, or heavy rainfall. These events are sometimes referred to as High Impact Weather (HIW) events as they may result in severe damages to critical infrastructure or crops within a short time period. For instance, in Kenya, 80% of the country is arid and semi-arid. Understanding and quantifying the magnitude and frequency of extreme rainfall events in the present and how they may change in the future is very important for sustainable agricultural and land and water resources management. Understanding the likely recurrence of extreme rainfall events at the local scale is particularly important because this is where most decisions are made – by communities. HIW events, such as hailstorms, are expected to increase in a warming climate. In many parts of Africa, high impact events are caused by local-scale processes, such as convection. Convection is the vertical ascent of warm moist air parcels that leads to development of storms. As noted in the previous blog, global and regional climate models are not good at simulating local-scale processes because they can’t simulate climate processes at a high enough resolution. Global and regional models can simulate large-scale climate processes that can be “seen” better at low or coarse spatial resolutions (>10 km²). An example of a large-scale climate process is the tropical cyclones typically with horizontal length of a few hundred to several thousand kilometres. However, the failure of global and regional models to capture local-level processes results in errors that increase uncertainty in future climate projections. We also know that models better represent average climate over a period of time when they can simulate very short-term or short-lived physical processes better. These very short-term physical processes occur within a single day and are referred to as “sub-daily”. An example of a sub-daily process is the thunderstorms (convective storms) in the lake Victoria basin that tend to form over land in the afternoons due to local upsurge of warm air and occur above the lake at night. Accurate simulation of rainfall is particularly sensitive to model resolution and capturing sub-daily processes. A growing body of evidence suggests that convection-permitting models do a good job of capturing short-lived weather extremes resulting from rapid convection of air. The table below shows the differences between traditional Global Circulation Models (GCM), Regional Circulation Models (RCM) and convection-permitting models. It highlights how convection-permitting models enhance our understanding of local HIW events. The potential added value of convection-permitting model is greatest (i) at short time scales, (ii) when convection is the dominant cause of rainfall, and (iii) in regions of complex landforms (e.g. mountains, valleys and lakes).   Because convection-permitting models are better at characterising local-scale rainfall generating processes, they promise improvements in estimating local impacts due to HIW events. Their capacity to represent convection provides an opportunity to study many local scale processes that have profound impacts on a country’ economy and citizens’ livelihoods. These include critical factors that influence water cycles, such as convective systems in the Congo Basin in Central Africa, where the world’s most intense thunderstorms occur; the Lake Victoria circulation system in East Africa; and the West African monsoon. Understanding these processes could, for example, improve flash flood early warning and help save lives and reduce destruction of property.Source: Prain et al. (2017) & Kendon et al. (2017) Convection-permitting models are already being utilised in other regions of the world, such as in the UK, for operational forecasting and climate research. This is not yet the case in Africa. Under FCFA, the IMPALA project is running the first ever simulations of a convection-permitting model for the African continent. This will no doubt demonstrate the value of convection-permitting models for improving our understanding of African climate. Critical gaps, such as the poor quality of basic climate data and lack of adequate computing infrastructure that have persisted in Africa, may however limit the benefits of convection-permitting models. High-quality, high-resolution observation data is required to evaluate model simulations for improvements necessary for operational weather forecasting use. Ways to get around these challenges need to be determined in order to benefit from this new tool as a vehicle for both research and operational use in Africa. This will be the topic for the next blog in this series. Further Reading Kendon, E. J. et al. (2017), Do convection-permitting regional climate models improve projections of future precipitation change? Bull. Am. Meteorol. Soc. 98, 79–93. Prein, A. F et al. (2017), Challenges and Advances in Convection-Permitting Climate Modeling, Bull. Am. Meteorol. Soc. 98, 1027 – 1030. This blog was written by Zablone Owiti, Suzanne Carter, Julio Araujo and Jean-Pierre Roux