Comparison study on the MTG LI and WWLLN lightning data sets

Tropospheric lightnings are fundamental sources of natural ultra-wide band (UWB) electromagnetic waves, utilised in various fields of exploration of the upper atmosphere and the plasma environment. Lightnings are observed by numerous detection systems, deployed as ground networks or operated on satellites/cubesats, all exhibiting specific limitations in detection efficiency (DE), sensibility, spatial coverage, and overall performance.

We have compared the flash database of the LI optical experiment onboard the Meteosat Third Generation (MTG) geostationary satellite (0° longitude, operational since July 2024) with the ground-based WWLLN VLF stroke data set in the period of July 2024 – May 2025, within the FoV of the MTG sensors. For that purpose the necessary correction of the location and time coordinates was first performed in the MTG data set. This step sets the spacecraft observation time backwards with the propagation time (approx. 120-140 ms), and also decreases the latitude and longitude coordinates towards the sub-satellite line due to the finite altitude of the observed optical phenomenon at cloud top. 

The ratio of the detected events, binned in a 1° by 1° raster in the African continental region, varying somewhat geographically, falls in the remarkable range of several hundreds. This difference can be explained partly by the known poor DE of the WWLLN over Africa, and by the fact that MTG LI detects total lightning (cloud-to-ground and intracloud/cloud-to-cloud), while WWLLN primarily detects strong CGs. A one-by-one matching of MTG flashes with detected WWLLN strokes, applying temporal and spatial windowing (±330 ms, and <25 km, respectively) was also completed. This analysis exhibited clear asymmetry in the distribution of the time offsets between matching events (time stamps of MTG flashes seem to precede the corresponding WWLLN time values by tens of ms). The distribution of spatial separation of matching pairs has a maximum at 8 km. Due to the reasonably strict conditions used in matching pair selection, the overwhelming number of detected lightnings in the MTG LI data set is not seen in the one-by-one comparison: 97.3 % of WWLLN to MTG LI matchings are single event pairs, a multiplicity factor of 2 is represented only by 2.5 % of matched events.