Turbulence forecasting over South Africa using diagnostics from the convective scale ensemble prediction system of the Unified Model : a pilot study

Abstract

Clear air turbulence is a hazardous micro-scale feature to the aviation industry, and the forecasting thereof remains a significant challenge. Clear air turbulence indices describe possible turbulent development mechanisms and are utilized as a forecasting tool. The selected indices for this research are the Ellrod Turbulence index and the Turbulence Kinetic Energy index converted into the Eddie Dissipation Rate available from the South African Weather Service Convective Scale Ensemble Prediction System. Through eleven 2018 case study investigations over 39 clear air turbulence events, the indices from the Convective Scale Ensemble Prediction System are evaluated against pilot reports of clear air turbulence. That is to establish which indices are better suited for aviation purposes over South Africa. Eleven case studies were identified for three different weather categories namely cut-off lows, upper air troughs and upper air zonal flow. The investigations established that the ensemble average and ensemble probability forecasts of the two indices are skilful in predicting clear air turbulence events. However, there is evidence of under-forecasting event severity over South Africa. In general, the EDR severity forecasts outperformed the Ellrod severity forecasts. However, Ellrod performed better when forecasting clear air turbulence associated with cut-off lows while the EDR outperformed the Ellrod in the upper air trough and upper air zonal flow weather categories. Ellrod had the most missed forecasts when dealing with the upper air trough weather category. The post-processing of the clear air turbulence maps would contribute to current forecasting methods of clear air turbulence by the conjoint display and interpretation of both indices over South Africa.