Characterising Mesoscale Convective Systems and evaluating precipitation forecasts over West Africa

Mesoscale Convective Systems (MCSs) and their associated rainfall, winds and lightning pose threats to the lives of citizens and the economy of West African countries. These systems are poorly predicted as they are not well represented in Numerical Weather Prediction (NWP) models. To address this forecasting challenge, the (thermo-)dynamic environment associated with mature MCSs and its effect on the intensity of MCSs has been investigated. This analysis includes the effects of vertical wind shear, which plays a key role in the intensification of MCSs. The ability of current convection-permitting NWP models to represent rainfall and the effects of vertical wind shear on MCSs has also been evaluated by comparing the forecast skill of the Met Office 4.4 km Tropical Africa Model (TAM) to that of the Global model. Matured MCSs were mostly found in regions of strong surface convergence and high orography. These storms were also found to be associated with environments of strong vertical wind shear. Long-lived moderate speed storms had larger sizes, colder brightness temperatures (BTs) and associated with a strong vertical wind shear as compared to short-lived slow moving storms. Long-lived moderate speed storms were also associated with the presence of the African Easterly Waves. Oceanic storms were mostly slow-moving but had higher rain-rates as compared with land storms. A detailed investigation of the effects of vertical wind shear on MCSs has revealed that, a strong vertical wind shear results in colder BTs relative to their temperatures at the level of neutral buoyancies (LNBs). It is hypothesised that, vertical wind shear results in colder BTs by increasing updrafts and minimising entrainment. Oceanic storms could reach their LNBs compared to land storms, suggesting that, high entrainment rates over land, as well as entrainment dilution by drier air over land, prevents storms over land from reaching their LNBs. Over the Sahel, Southern West Africa (SWA), and the Coast, a strong vertical wind shear did not only result in colder BTs but also high rain-rates. The effects of vertical wind shear on MCSs was found to be better represented in the TAM compared to the Global model. The skill of the TAM is highest over coastal areas and mountainous regions suggesting the model's ability to simulate land-sea breeze and orographic effects more accurately. The TAM outperforms the Global model in forecasting sub-daily rainfall over West Africa, as well as in simulating the closest location of convective rainfall as compared to observations, although this location is still too coarse for forecasters to make useful decisions. The skill of the TAM is also significantly affected by the long spin-up period of the model, making Day 2 forecasts better than that of Day 1. The TAM also outperforms the Global model for higher rainfall intensities, demonstrating the model's ability to simulate organised convective systems more accurately. These findings are useful for forecasters, model developers and policy makers as a better representation of the (thermo-)dynamic environments of mature MCSs and the effect of vertical wind shear on rainfall, which is poorly represented in current convection-permitting climate models, will not only enhance better rainfall forecasts but also help to provide more accurate climate projections.