Filtering Algorithm of False Events in Lightning Detection by FY-4 Lightning Mapping Imager

The lightning detection by Lightning Mapping Imager (LMI) of the FY-4 geostationary satellite plays a significant role in monitoring strong convection in real time and provides continuous lighting measurements. An appropriate lightning filtering algorithm is proposed and described in this paper. The ghost, track and the shot are recognized as the primary non-lightning artifacts by analyzing the in-board lightning data of the LMI. The ghost is identified based on the mirror rules of the position and the energy measured in the laboratory. A line detection method based on the Hough transform is adopted to eliminate the track. The shot is filtered based on the event clustering principle. The lightning filter algorithm is applied to process two samples, sample one is obtained when a strong thunderstorm happened on the south and the southwest China from March 29th–30th, 2017, the other sample is obtained on June 21th, 2017 when a short-term serve storm happened in Beijing-Tianjin-Hebei region. The lightning data obtained by the LMI and synchronous ground based strokes was processed and compared. The results shows that the spatial distribution observed from LMI is in general agreement with that of ground-based monitoring result while the proposed filtering algorithm is applied, providing a well proof of the lightning detected accuracy of LMI.