Simulations of Selective Seeding on Hailstorms in Northern Switzerland Using the COSMO Model: Effects on the Lightning Potential Index

Despite the challenges in understanding the relationship between lightning and convective cell dynamics, several approaches, among them the Lightning Potential Index (LPI), were developed to parameterize lightning activity. The LPI uses microphysical parameters to estimate the potential for charge separation in convective cells and is employed as an extension to the COSMO model (Consortium for Small-Scale Modelling) to simulate an observed convective storm on the 6 July 2019. A time-lagged approach with a control and three seeding simulations yields ten ensemble members each. In each seeding simulation, we inject silver iodide particles below cloud base in regions of updraft of a simulated thunderstorm cell. The concentrations of silver iodide are 0, 0.2, 20 and 2000 particles per cm 3 . In our data analysis, we consider the graupel, snow, ice crystal, raindrop, supercooled cloud droplet contents as well as the updraft and downdraft speeds. In the thunderstorm cell under investigation, enhanced updraft speeds in the seeding simulations lead to an increased potential for charge separation. The LPI increases by 15.3%, 61.1% and 11.2% with increasing seeding concentration, with respect to the control simulation. Our findings suggest that the medium seeding concentration possibly constitutes a Goldilocks situation - the sweet spot between increasing updraft speeds and still having sufficient supercooled cloud water. The results are tested for statistical significance with the Wilcoxon Signed-Rank test and we find a statistically significant seeded increase in the LPI in two of the three seeding simulations with p-values of 6.5%, 1% and 4.2%.