Evaluation of the Detection Efficiency of the Meteosat Lightning Imager using Ground-Truth Observations at the Sentech Tower in South Africa

Lightning is a severe atmospheric phenomenon that can cause human casualties, wildfires, infrastructure damages, and aviation disruptions. By timely prediction of lightning occurrences, preventive measures can be advised to counter the adverse effects of lightning. Prediction tasks require high quality data to ensure reliable predictions and avoid false alarms and unnecessary preventive measures. In this work, we analyzed the quality of lightning data collected by EUMETSAT's Meteosat satellite. The assessment is performed by comparing satellite's observations with high-speed camera recordings of lightning activity at the Sentech tower in Johannesburg, South Africa. The tower is chosen because its exact position is known, facilitating accurate matching between the satellite data and ground-based observations. We evaluated the detection efficiency to be 24% for all types of lightning processes, 38% for downward processes, and 12% for upward processes. These values are lower than expected when comparing to the values obtained for Europe (Switzerland and Austria). The reduced performance is attributed to the Sentech tower's position on the edge of the sensor's field of view, resulting in lower quality data and lower detection efficiency.