Characterization and Modeling of Upward Lightning at the Gaisberg Tower
MetadataShow full item record
The majority of cloud-to-ground lightning is downward. Upward lightning rarely occurs; less than 1% of cloud-to-ground lightning is upward. Upward lightning differs from its downward counterpart in how it is initiated. Upward lightning is initiated by an upward moving leader from a tall (> 100 m) object on ground that reaches into the charge regions of an overhead thundercloud resulting in a relatively-long-duration (tens to hundreds of milliseconds), low-level (tens to hundreds of amperes) current called the initial continuous current. It is of significant importance to understand in detail the processes associated with upward lightning as it affects ground-based infrastructure as well as aircrafts (during take-off) and spacecrafts (during launch). The majority of lightning occurring at the Gaisberg Tower in Salzburg, Austria, is upward, therefore, measurements made at this tower provide an ideal dataset for studying this type of lightning discharge. In this work, a detailed analysis of lightning current waveforms measured at the Gaisberg Tower (over nearly two decades) and electric fields measured at close and far distances is performed. Additionally, a model for upward leaders relating its charge density to the vertical component of electric field measured on perfectly conducting ground is presented. We found that 79% of the upward discharges at the Gaisberg Tower transferred only negative charge to ground, 4.3% transferred only positive charge to ground, and 17% transferred both negative and positive charge at different times during a flash (therefore they are called bipolar flashes). We extended the traditional categorization of bipolar flashes by classifying them into five types. Upward flashes at the Gaisberg Tower have a median flash-current duration of 311 ms and a median charge transfer of 35 C. We report the occurrence of bipolar current pulses and associated electric field charges at beginning of upward flashes which are likely related to inception-processes of upward leaders. Finally, comparing model computed and measured electric field changes for upward leaders we estimated the average line charge density and leader speed.