Enhancement of the South American Meteorological Hazards and Impacts Database (SAMHI) through Atmospheric Electrical Activity as a Proxy for Severe Weather Event Detection

Atmospheric electrical activity (AEA) reflects the microphysics and internal dynamics of storms, and abrupt variations in AEA, known as Lightning Jumps (LJ), can anticipate the occurrence of severe weather events up to 45 minutes in advance. In 2023, the South American Meteorological Hazards and Impacts Database (SAMHI) (https://samhi.cima.fcen.uba.ar/) was created to collect reports of hail, tornadoes, floods, and strong wind gusts, aiming to enhance the understanding, remote detection, and predictability of these phenomena in the context of climate change in South America.

Given that collecting reports, particularly in rural areas, remains a significant challenge, this study aims to enrich the SAMHI database and identify vulnerable regions that have been previously underestimated. To this end, reports of hail, tornadoes, and strong wind gusts from the 2018–2023 period were used, along with Lightning Jump data from 2009–2023. LJ data were derived by processing AEA data from the World Wide Lightning Location Network using the Georayos algorithm (https://georayos.citedef.gob.ar/), which applies the Density-Based Spatial Clustering of Applications with Noise  method to group lightning and detect significant increases in lightning rates.

The connection between severe weather events and LJ was analyzed across Argentina, Uruguay, and southern Brazil, a region recognized as one of the most active in the world for severe storms, frequently experiencing hail, tornadoes, floods, and intense wind gusts, with significant economic impacts on sectors such as agriculture, infrastructure, and energy. The analysis consisted of assessing the presence of LJ before and after each reported event, followed by the application of the K-Means clustering algorithm to identify the regions with the highest frequency of experiencing severe events.

The results show a significant correspondence between LJ occurrences and reported severe events, with an average lead time of 29.3 minutes. Over 60% of the LJ cases preceded severe events, with correspondence rates of 70% for hail, 60% for tornadoes, and 50% for strong wind gusts. In addition, Uruguay, northern and central Argentina and southern Brazil were identified as the regions most likely to experience adverse weather conditions. The use of LJ data also made it possible to characterize the different regions by type of most frequent severe events and to identify vulnerable areas previously underestimated, probably due to factors such as low population density, limited access to communication routes, among others.

These findings highlight the importance of integrating LJ data to enhance early detection of severe weather events across South America.