Association of ionospheric whistler waves with WWLLN lightning using a relaxed correlation method

This concerns the association of whistler features identified in the C/NOFS VEFI burst-mode VLF recordings, with causative lightning strokes located by the WWLLN global lightning network. Until now, we have pursued "strict correlations" based on temporal simultaneity (after correcting for propagation delays) between the WWLLN stroke and the VEFI whistler pulse. This requires that we limit our search to only those whistler strokes that, after adaptive dechirping, can be restored to a "pre-ionosphere" pulse width of ~ 0.5 millisec or less. Only these narrow pulses can be efficiently and accurately detected automatically and dechirped automatically. Whistlers of greater temporal width are not good candidates for such a "strict-correlation" approach. Unfortunately, the narrow pulses contain only ~ 0.25 of the total whistler energy content in the VEFI signals, and of these narrow pulses, only ~20% are successfully correlated and associated with a WWLLN stroke. Thus only 5% of the whistler energy content observed by VEFI is strictly associated with a terrestrial lightning stroke. However, the wider pulses (containing ~95% of the whistler energy) observed by VEFI still display an "eyeball" association with clusters of WWLLN activity. We will explore a more permissive approach (compared to strict correlation) to be able to attribute these wider whistler pulses to specific outbursts of terrestrial lightning.