Abstract/Summary

Lightning presents a substantial hazard. Thunderstorm nowcasting requires a lightning strike to have already occurred, so advance warnings of lightning are currently only possible using numerical weather prediction models. This research aims to identify the processes within cumuliform clouds that can distribute charge prior to the first lightning strike. Precipitation can influence the potential gradient measured at the surface, masking the charge residing within an overhead cloud. An increase in negative charge was measured with increasing rain rate that was dependent on changes in surface conditions such as wind speed and surface dryness, but not cloud type. A field campaign was set up at Chilbolton Observatory by installing a field mill and Biral thunderstorm detector. Surface atmospheric electricity measurements were analysed for 653 cumuliform clouds observed over two years using a 35 GHz radar. Mixed-phase clouds with a high ice phase moisture content produced the largest electrification with increases in the size, density, and turbulence of the hydrometeors all increased cloud electrification, findings consistent with the relative diffusional growth rate theory. Ten radiosondes were flown from Reading University Atmospheric Observatory to provide in-situ measurements of charge (up to 550 pC m-3), cloud backscatter, turbulence (up to 10-1 m2 s-3), and supercooled liquid water (up to 2 g m-3) in cumuliform clouds. These measurements yield consistent findings with those from radar observations. The shallow depth of the charge centres within the cumuliform clouds suggests that the charge was typically insufficiently distributed to allow the required potential gradient enhancement for atmospheric breakdown leading to lightning. This is concluded as the fundamental reason cumuliform clouds in the UK are often charged but rarely produce lightning. This research could be used to improve probabilistic lightning forecasts and nowcasts through improvements to lightning parametrisation schemes in forecast models and the identification of potential thunderstorms using real-time existing surface- and satellite-based radar networks.