Return to search

Practical considerations when inferring lightning current from electric field recordings with a high noise-floor

A dissertation submitted to the Faculty of Engineering and the
Built Environment, University of the Witwatersrand, in
ful lment of the requirements for the degree of Master of
Science in Engineering.
Johannesburg, 2015 / During a cloud to ground lightning event a charge centre within the storm
cloud is discharged. The discharge of a charge centre within the cloud leads to
a change in the electric eld radiated by the charge centre. It is theoretically
possible to infer the lightning current from the derivative of the electric eld.
It is only possible to infer the lightning current from the electric eld data
where the noise is comparatively much smaller than the electric eld data.
The changing electric elds for a lightning event that occurred on the 3rd
January 2015 13:15:13 were recorded by a
at plate electric eld sensor with a
passive integrator. The oscilloscope used to capture the electric eld data has
a relatively large measurement noise and a low resolution. A low pass digital
lter was applied to the recorded electric eld data to reduce the e ects from
the high frequency noise. The lightning strokes were recorded by the South
African Lightning Detection Network. The Lightning Detection Network
data is used to obtain the distance of the lightning event from the sensor,
to scale the inferred lightning current. The Lightning Detection Network
also provides a lightning peak current measurement to compare to the peak
current inferred from the electric eld data. The lightning stroke current
was inferred from the electric eld recording for various bandwidths of the
low pass lter. Inconsistent changes to the inferred lightning stroke current
as the lter bandwidth is changed shows that the frequency components
for each stroke di ers. The peak stroke current was not constant for any
lter bandwidth range implying that the measurement noise is relatively too
large. The case study presented demonstrates that with a relatively large
noise magnitude (3 to 4 discrete steps of the digital recording) compared to
the electric eld signal (21 discrete steps) it is di cult to accurately infer the
lightning current from the electric elds recorded. / MT2017

Identiferoai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:wits/oai:wiredspace.wits.ac.za:10539/22695
Date January 2015
CreatorsLange, Jarren Hilton
Source SetsSouth African National ETD Portal
LanguageEnglish
Detected LanguageEnglish
TypeThesis
FormatOnline resource (xii, 76 leaves), application/pdf, application/pdf

Page generated in 0.0016 seconds