A research project in cooperation with the Technical University Vienna and Arsenal Research

This project is sponsored by:

General

Based on the 1992 - 1997 data from the Austrian Lightning Location System ALDIS we performed a systematic analysis to reveal spots of high flash density in Austria.

One of the sites with an extraordinary high flash rate was at the "Gaisberg", a mountain near the city of Salzburg. We are assuming that most of the localized flashes actually did hit the radio tower at the top of the mountain. This was also confirmed by the local staff.

Similar numbers of increased high flash rates were observed in Austria at some other mountain tops with radio towers like Dobratsch and Kitzbühler Horn.

Considering some other aspects as the accessibility of the site during the entire year, the Gaisberg tower was selected for installation of equipment to perform lightning research on natural lightning.

Besides research on triggered lightning - as done since several years at Camp Blanding in Florida - measurements on lightning flashes to elevated objects are the only way to collect information about the lightning current waveform of natural lightning directly. Only very few sites are existing world wide, where direct lighting is measured:

CN Tower in Toronto, Canada, 550m
Chimney in Japan (Fukui), 200 m
Experimental tower in Brazil, Morro do Cachimbo, 60 m

Research Objectives

Measuring Equipment

To measure the current when lightning strikes the radio tower we have installed two sensors at the tower top - a shunt for current measurement and an inductive probe for di/dt recording.

When a lightning flash hits the lightning rod at the tower top, the lightning current flows directly through a current shunt of high bandwidth (0,25 Ohm) and the inductive probe and than it is carried by the lightning protection down conductors.

Sensor arrangement on the tower top

The output signal of the shunts are converted to optical signals and transmitted to the control building near the tower via fibre optical cables.

To be able to collect data over a wide range of peak current amplitudes two channels of distinct sensitivity are used:

Channel 1: 0,05 kA to 2,0 kA - to measure leader currents and continuing currents.

Channel 2: 1,6 kA to 40 kA - to measure high amplitude lightning pulse currents.

The measuring system applies GPS time stamping to all recorded data to be able to time correlate the lightning current data from the tower with the data from the Austrian Lightning Location System ALDIS.

All the equipment is located inside the office building of the ORF, the Austrian TV and the operator of this radio tower.

For data acquisition we are using PC based digitizer boards with a sampling rate of 20 MS/s for the lightning current recording and 100 MS/s for di/dt-data acquisition. At maximum sampling rate data over duration of up to 0.8 seconds are digitized. Dead time between triggers is about 20 seconds. This time is required to download the data (32 MB per trigger) from the onboard memory to the local hard disc.

The continuous data acquisition over 0.8 seconds allows collecting information about all different processes during a lightning flash, starting from the initialization of the upward leader till to the end of the last stroke of a flash.

The measuring system is configured in a way to be remotely controlled and operated using an ISDN modem access.

Some Results

On September 17th, 1998 the first flash to the tower was recorded by the measuring system. Until December 2003 we have collected data from 361 lightning events. A significant variation of the annual number of flashes to the tower is observed.

Based on the recorded current wave shape we could classify most of the events as upward initiated discharges as typical for elevated objects like the Gaisberg tower. The majority of discharges are of negative polarity and some currents actually show a bipolar wave shape. This means, that during a flash the current in the lightning channel reverses polarity.

Current Measurement
The following figure shows a typical wave shape of a upward lighting discharge. The discharge starts with a so called "Initial Continuous Current (ICC)" with amplitudes of several hundred amperes and duration of several 100 milliseconds. Frequently one or more Leader-Return Stroke sequences are observed following the ICC after a short period of no current in the lightning channel. These strokes are comparable to subsequent strokes in downward

Typical current wave shape of an upward initiated discharge

Superimposed on the ICC we observe current pulses of different wave shapes.

Example of two very different current pulses superimposed on the ICC

We have recorded numerous discharges showing only an ICC current without any significant superimposed pulses. Although the peak amplitudes of those ICC's are in the range of a few hundred amperes only, a significant amount of electric charge is drained from the thunder cloud to ground by this type of discharge.

High speed camera

Late evening June 14th, 2000 we succeeded the first time to record a tower strike with the installed high speed video system. This video system can take up to 1000 frames per second (resolution 240 x 512 pixels) and should provide sufficient optical information of the total lightning strike. For time correlation with the measured currents and the ALDIS data the video system is also synchronized by a GPS clock.

A Movie-File showing a lightning strike to the tower on June, 14th 2000, 20:30 (UTC) exhibits the change in luminosity depending on the current flow in the lighting channel. This movie has been created from several hundreds of bitmap files taken by the high speed video system. The left to right movement of the lightning channel in the movie is a result of strong wind blowing on Gaisberg when the tower strike occurred.

We have analyzed in detail the correlation between measured current at the tower top and the lightning channel brightness right above the tower top. We observe a strong linear correlation between these two parameters [Diendorfer et al., 2003]. Based on this correlation we could estimate the fraction of the total lightning current carried by the individual branches of the lighting channel.

'
Lightning flash to the tower with 3 channel branches (Download MPEG movie of the flash (637kB))

We can only measure the total lightning current i_m at the tower top. We can estimate the current fraction in the individual branches (i₁, i₂ und i₃) based on the function of brightness versus time of the individual branches.

Lightning current as a function of time in the 3 channel branches

Field mill

For the measurement of the long term electrostatic field changes in the vicinity of the tower a field mill is installed. The field mill is provided by VAISALA, the manufacturer of the lightning location system used in Austria.

Also during periods of fine-weather there exists an electric field of about 130 V/m as a result of the positive space charge in the atmosphere. When a thunderstorm is approaching or locally developing this field first changes polarity and then increases to values of up to -10.000 V/m. Any significant change in the electrostatic field measured by the field mill is an indication of cloud electrification. The recording of the field mill data on August, 2nd 2000, shows maximum electrostatic fields in the range from -5.5 kV to +3kV during occurrence of a storm in the region of Salzburg.

Lightning flashes either intracloud or cloud-to-ground cause rapid changes in the electrostatic field plot as a result of the rearrangement of charges in the overhead thunderstorm.

Correlated Far Field Measurements

An additional station to measure remote (distance of several 10 kilometres) electromagnetic fields radiated by the strokes to the Gaisberg Tower was installed in Wels in 2007. For operational and maintenance reasons this station was moved to Neudorf, Upper Austria, in 2008. Distance to the Gaisberg Tower site is 108.7 km. The data collected by this experimental setup will allow validating some return stroke models, used to calculate the radiated fields, as in this case correlated current and field waveforms at a known distance are available. The only remaining unknown will be the return stroke velocity. The following figure shows the fast antenna mounted at the roof of a building in Neudorf.