Modulation of ensemble perturbation growth over tropical Africa by the diurnal cycle

Abstract The mechanisms governing ensemble perturbation growth in tropical numerical weather prediction remain poorly understood, despite their importance for forecasting convection and hazardous weather. Here, we investigate practical predictability over tropical Africa using a convection‐permitting ensemble and its driving global counterpart, using kinetic and thermodynamic difference‐energy diagnostics, age‐of‐air diagnostics, and spectral analysis. Perturbation growth exhibits a pronounced diurnal cycle in the convection‐permitting ensemble, with maximum spread occurring when deep convection is most active. In contrast, the global ensemble displays a delayed phase and broader spatial distribution of uncertainty. Ensemble spread in the convection‐permitting system remains closely collocated with regions of organised convection and African Easterly Wave activity throughout the forecast period, whereas the global ensemble develops substantial upper tropospheric spread over the Indian Ocean and East Africa. Spectral analysis reveals rapid saturation of perturbations at wavelengths below approximately 200 km, while larger‐scale perturbations above 200 km continue to amplify throughout the 72‐hour forecast. These results support the view that tropical perturbation growth is governed primarily by moist convection and tropical‐wave dynamics, rather than following the classical three‐stage evolution commonly identified in midlatitude predictability studies. A further result is the identification of air‐mass residence time as an important control on realised ensemble spread within the limited‐area model. Regions of weak spread are associated with recently imported air masses that have experienced insufficient time for local perturbation growth. The results highlight fundamental differences between tropical and extratropical practical predictability and provide guidance for the design of future convection‐permitting ensemble prediction systems over tropical regions.