356 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 43, NO.

3, AUGUST 2001

Current and Electromagnetic Field Associated With

Lightning–Return Strokes to Tall Towers
Farhad Rachidi, Wasyl Janischewskyj, Fellow, IEEE, Ali M. Hussein, Senior Member, IEEE, Carlo Alberto Nucci,
Silvia Guerrieri, Behzad Kordi, and Jen-Shih Chang, Senior Member, IEEE

Abstract—In this paper, an analysis of electric and magnetic I. INTRODUCTION

fields radiated by lightning first and subsequent return strokes
to tall towers is presented. The contributions of the various
components of the fields, namely, static, induction, and radiation
for the electric field, and induction and radiation for the magnetic
I NFORMATION about lightning current parameters either
comes from direct measurements, using, for example tall
instrumented towers, or is based on quantification of lightning
field are illustrated and discussed. It is shown in particular that electromagnetic fields from which lightning currents are
the presence of a tower tends, in general, to increase substantially
the electric and magnetic field peaks and their derivatives. This
inferred adopting some empirical [1], [2] or theoretical [3] rela-
increase is mainly caused by the presence of two oppositely tions. Experimental observations and theoretical investigations
propagating current wavefronts originating from the tower top have shown that the presence of an elevated strike object, such
and by the very high-propagation velocity of current pulses as a tall tower, could affect substantially lightning currents and
within the tower (practically at the speed of light), and depends their radiated electromagnetic fields (e.g., [4]–[14]). Accurate
essentially on the wavefront steepness of the channel-base current.
Because of the last factor, the increase of the field magnitudes is
knowledge of lightning electromagnetic fields is essential
found to be significantly higher for subsequent return strokes, for achieving an efficient insulation design of electric-power
which are characterized by much faster risetimes compared networks and for determining electromagnetic compatibility
to first return strokes. Furthermore, the presented results are requirements of telecommunication systems, as many light-
shown to be consistent with recent experimental observations ning-caused disturbances are due to lightning electromagnetic
of current in lightning strokes to the Toronto CN Tower and
of the associated electric and magnetic fields measured 2 km
fields (e.g., lightning-induced overvoltages [15]).
away. These findings partially explain the fact that subsequent In this paper, an analysis of electric and magnetic fields ra-
return strokes characterized by lower current peaks but higher diated by lightning return strokes to a tall tower is presented.
front steepnesses and return stroke speeds may result in higher Using the modified transmission line (MTL) model1 Reference
field peaks. The results obtained in this study have important [16], [17], recently extended to take into account the presence
implications in electromagnetic compatibility. It is found that
lightning strikes to tall metallic objects (towers, rods, etc.) lead
of an elevated strike object [9], radiated electric and magnetic
to increased electromagnetic field disturbances. Also, subsequent field waveforms are computed and analyzed. In the analysis, two
return strokes are to be considered an even more important source channel-base current waveshapes, corresponding respectively to
of electromagnetic interferences than first return strokes. Indeed, typical first and subsequent return strokes, are utilized. The ef-
electromagnetic fields from subsequent strokes are characterized fect of the presence of the tower on the magnitude and shape of
by faster fronts and additionally, they may reach greater peaks
than first strokes. Lastly, findings of this study emphasize the
2-km distant fields is investigated, and the contributions of var-
difficulty of extracting reliable lightning return stroke current ious field components, namely, static, induction and radiation
information from remote electromagnetic field measurements for the electric field, and induction and radiation for the mag-
using oversimplified formulae. netic field are illustrated and discussed. Finally, electromagnetic
Index Terms—Lightning, lightning electromagnetic pulse, light- compatibility implications of the theoretical findings are consid-
ning at tall structures, lightning-return stroke current. ered, and some explanations are given on recent experimental
observations of lightning flashes to the CN Tower.
Manuscript received April 20, 2000; revised December 10, 2000. This paper
was supported in part by the Swiss National Science Foundation, in part by II. COMPUTATION OF RADIATED ELECTRIC AND
the Natural Sciences and Engineering Research Council of Canada, and in part MAGNETIC FIELDS
by the Italian Ministry of University and Scientific Research. This paper was
presented in part at the 24th International Conference on Lightning Protection A. Electromagnetic Field Radiated by a Vertical Antenna
(ICLP), Birmingham, U.K., September 1998.
F. Rachidi is with the Swiss Federal Institute of Technology, Department of Above Ground
Electrical Engineering, Lausanne 1015, Switzerland. In our computations, we consider both, the lightning-return
W. Janischewskyj is with the University of Toronto, Department of Electrical
and Computer Engineering, Toronto, ON M5S 3G4, Canada. stroke channel and the elevated strike object, as vertical
A. M. Hussein is with the Ryerson Polytechnic University, Department of antennas above a perfectly conducting ground. At a general
Electrical and Computer Engineering, Toronto, ON M5S 3G4, Canada. point in free space, the expressions for the vertical electric and
C. A. Nucci and S. Guerrieri are with the University of Bologna, Department
of Electrical Engineering, Bologna 40136, Italy. azimuthal magnetic-field components, originated by a vertical
B. Kordi is with the Shahed University, Tehran 15875–5794, Iran.
J.-S. Chang is with the McMaster University, Department of Engineering
Physics, Hamilton, ON L8S 4MI, Canada. 1The MTL model proposed in [16], [17] has been recently referred to as the
Publisher Item Identifier S 0018-9375(01)07130-7. MTLE model [18].

0018–9375/01$10.00 © 2001 IEEE

RACHIDI et al.: CURRENT AND ELECTROMAGNETIC FIELD ASSOCIATED WITH LIGHTNING-RETURN STROKES TO TALL TOWERS 357

dipole of infinitesimal length along the axis and at a reflected at the top of the object (and part of it is transmitted to
height [19] are given by the lightning channel). This multiple-reflection process along
the elevated strike object, assumed lossless, continues until the
energy of the pulse dissipates in the ground and in the light-
ning channel [see (3)]. Concerning the current distribution in the
lightning channel above the strike object, we assume that it can
be described by the MTL return stroke current model, in that the
current pulse propagates upward in the channel with expo-
nentially decreasing amplitude. This attenuation is not viewed
as due to losses in the channel but is rather to take account of the
(1) effect of the distribution of charge stored in the corona sheath of
the leader and subsequently discharged during the return stroke
phase [25]. Clearly, the expression of the spatial-temporal dis-
(2) tribution of the current along the channel has also to account for
the portions of the pulses refracted into the channel following
the multiple reflections at the ground and at the object top [see
in which are the cylindrical coordinates of the observa-
(4)].5
tion point, is the distance between the dipole and the observa-
The expressions for the current distribution along the elevated
tion point, , is the dipole current,
strike object and the lightning return stroke channel, as derived
is the speed of light, and, is the permittivity of free space.
in [9], [11], are given by
The first term in (1) is the electrostatic field, the second the elec-
tric induction or intermediate field, and the third the electric ra-
diation field. In (2), the first term is the magnetic induction and
the second the radiation field.

B. Lightning Return Stroke Model Including the Presence of

an Elevated Strike Object for (3)
The problem of lightning-return strokes to tall towers was the
subject of numerous recent studies (e.g., [4]–[14], [20], [21]).
Different lightning-return stroke models, which, in their orig-
inal forms, assume the return stroke to start at ground level,
have been extended to take into account the presence of an ele-
for (4)
vated strike object [6], [9]–[11], [20]–[22]. In the present anal-
ysis, we will use the extended MTL model as described in [9]. where
For convenience, the main features of such a model are here and reflection coefficients at the top of the object
summarized. As the leader descends toward the ground and and at ground, respectively;
reaches and attaches to an elevated object, an abrupt potential “noncontaminated” return-stroke current
change generates a wave front (return stroke) that propagates pulse at the top of the object [11];
toward the cloud. The current pulse associated with the decay constant of the MTL model [16], [17].
return-stroke phase starts at the top of the elevated object and In the above expressions, possible reflections of the upward-
propagates upward at the return stroke velocity and with an propagating current pulses at the return-stroke wavefront [22],
exponentially-decreasing current amplitude, in accordance with [23] are ignored, even though it is shown by Shostak et al. [26]
the MTL model. Simultaneously, a similar current pulse that a better agreement with experimental data is found when
is assumed to be injected by the lightning channel at the top considering such reflections.
of the object. The object is modeled as an ideal transmission Electromagnetic fields computed using the extended MTL
line characterized by a height and reflection coefficients model have been compared with experimental data and a sat-
and , at its top and ground respectively. These reflection co- isfactory agreement has been found [9].
efficients are assumed to be constant and independent of fre-
quency.2 The “noncontaminated”3 return-stroke current pulse C. Non-Contaminated Channel-Base Current and Return
travels downward the object at the speed of light and is Stroke Speed
first reflected at the bottom of the object (and part of it is trans- In this study, we have considered two channel-base current
mitted to ground).4 The reflected wave travels upward and is waveforms corresponding, respectively, to typical first and
2Note that the reflection at ground level can be considered also in absence of
5Note that the MTL model can be seen also as the result of the contribu-
any elevated strike object for return strokes initiating at ground level [23]. tion of a number of current sources distributed along the channel [25]. Each of
3Free of reflections from the struck object.
these distributed sources delivers charges moving downward as the return stroke
4While ground is assumed to be perfectly conducting for field computation wavefront reaches its altitude. In this paper, however, we have disregarded for
(an approximation which is reasonable for the vertical electric and azimuthal simplicity the reflections of these source contributions at the object top, an as-
magnetic fields [24]), its finite conductivity is taken into account when deter- sumption that we consider will not alter qualitatively the obtained results, as it
mining current reflections. will be shown later.
358 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 43, NO. 3, AUGUST 2001

TABLE I
PARAMETERS OF THE TWO HEIDLER’S FUNCTIONS USED TO REPRODUCE THE CHANNEL-BASE CURRENT WAVESHAPE

subsequent return strokes, based on observations of Berger et

al.6 [27].
The “noncontaminated” channel-base currents are repro-
duced by means of a sum of two Heidler’s functions [28]. The
first return-stroke channel-base current is characterized by a
peak value of 30 kA and a maximum steepness of 12 kA/ s,
whereas the subsequent return stroke current has a peak value of
12 kA and a maximum steepness of 40 kA/ s. Note that recent
experimental data from triggered lightning indicate values for
the maximum steepness of about twice as large. However,
the main results of this study remain qualitatively unchanged
adopting a larger value than 40 kA/ s. The parameters of the
Heidler’s functions are given in Table I. Fig. 1 presents both
channel-base current waveforms and their time-derivatives.
The values adopted for first and subsequent return-stroke
speed within the lightning channel, are, respectively, 1.7
m/s and 1.9 m/s, corresponding to the average
values for natural lightning over the bottom 500 m of the
channel [29]. The decay constant in the MTL model is as-
sumed to be equal to 2 km, a value which has been determined
using experimental data [17].

III. SIMULATION RESULTS AND DISCUSSION

A. Lightning Return Stroke Initiated at Ground Level
Consider first the case of a return stroke initiated at ground
level. Fig. 2(a) and (b) represents the first and the subsequent
stroke vertical electric field computed at 2 km from the lightning
channel. Such a value has been used since the distance from the
CN tower to the E- and H-field measuring systems is 2 km (see
[30]). The contributions of the static, induction and radiation
components of the field are also included. It can be seen that the
first-stroke field [Fig. 2(a)] is characterized by a fast rising part
mainly due to the radiation component, followed by a ramp in-
crease due to the static term. The rapid decrease of the radiation
term for the subsequent-stroke field is responsible for the initial
Fig. 1. (a) Idealized channel-base current. (b) Current derivative for typical
first peak [Fig. 2(b)]. first and subsequent strokes.
Fig. 3(a) and (b) represents the azimuthal magnetic field
waveforms for the same distance. The induction and radiation
terms are separately included. In case of the magnetic field, occurs at an earlier instant of time. For both waveforms, the
the waves for both the first and the subsequent stroke show typical azimuthal H-field is characterized by a fast rising part,
an increase to a peak and then a gradual decay, although again due to the radiation term, followed by a slight “hump”
for the subsequent stroke the peak is more pronounced and resulting from the induction component. The zero-crossing of
the radiation term occurring at about 15 s, is responsible for
6It is worth noting that Berger’s data are also obtained using instrumented the faster decay of the total magnetic field in comparison to
towers. Hence, even though the various reflections are not discernible in mea- that of the induction term.
surement records due to the relatively short instrumented tower, some contami- It can also be seen in Figs. 2 and 3 that the contribution of the
nation might yet occur (see [11]). However, despite the above, in this study, we
consider the Berger’s current waveshapes as reference and exempt from tower radiation term to the total electric and magnetic field is more
contamination. significant for the subsequent stroke. This is essentially due to
RACHIDI et al.: CURRENT AND ELECTROMAGNETIC FIELD ASSOCIATED WITH LIGHTNING-RETURN STROKES TO TALL TOWERS 359

Fig. 2. Vertical electric field at 2 km from the lightning channel. Return stroke Fig. 3. Azimuthal magnetic field at 2 km from the lightning channel. Return
initiated at ground level. (a) First return stroke. (b) Subsequent return stroke. stroke initiated at ground level. (a) First return stroke. (b) Subsequent return
stroke.

the fact that subsequent strokes are characterized by faster re- for a more accurate representation of the tower, three or four
turn stroke speeds and higher front steepnesses of the current transmission-line sections in cascade are to be considered [10],
waveform. [33]. The reflection coefficients at the bottom and at the top
It is worth mentioning that the contribution of the various of the tower are assumed to be 0.48 and 0.50, respectively.
components of the electric and magnetic fields depends strongly These values have been derived analyzing the fine structure of
on the distance to the observation point. At closer observation the lightning return stroke current measured at 474 m above
points (less than a few hundred meters), the radiation-term con- ground level [10].
tribution to the field magnitude is small compared to contribu- Current and Current Derivative Along the Tower: Fig. 4
tions from the static and induction terms. However, at distances shows the waveforms of current and current derivative eval-
beyond some tens of kilometers, the contribution of the radia- uated at the top (553 m), the middle (276.5 m), and the base
tion term will become predominant and the static and induction of the tower (0 m), starting from the typical first return stroke
terms will have a negligible effect [31]. current7 presented in Fig. 1, as the “noncontaminated current”
at the top of the object and taking into account reflections at its
B. Lightning Return Stroke to a Tall Tower two extremities.
Elevated Strike Object: Consider an elevated strike object It can be seen that moving toward the ground, the current
characterized by a height of 553 m above ground, corresponding experiences a higher peak value and a shorter time to its peak
to the actual height of the CN Tower [32]. In this paper, the tower due to the contribution of the reflected wave at the ground level.
is modeled as a single, uniform and lossless transmission line (as 7Note that in the case of lightning strikes to a tall tower, the adopted first and
described in Section II-B). Such an assumption does not alter subsequent stroke-current waveshapes represent, respectively, slow-front and
the conclusions of our analysis, although it has been shown that fast-front strokes.
360 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 43, NO. 3, AUGUST 2001

Fig. 4. (a) Current; and (b) current derivative along the strike object for a Fig. 5. (a) Current; and (b) current derivative along the strike object for a
typical first return stroke. typical subsequent return stroke.

Fig. 5 presents similar results, using the typical subsequent Electromagnetic Fields: Figs. 6 and 7 present the waveforms
return-stroke current. It can be seen that in case of subsequent of the first and subsequent return stroke vertical electric field
strokes, both the current and the current time-derivative are more and azimuthal magnetic field at 2 km from the tower. In the
significantly affected by the presence of the tower. Furthermore, same figures, the contributions of various electromagnetic field
since the time to current peak is shorter than the wave traveling components (static, induction and radiation) are also illustrated.
time along the tower, the current reflections can be clearly dis- Note that current reflections along the tower can be identified,
tinguished on the waveforms. especially on the radiation term of the subsequent stroke field.
The main parameters of the current waveform at various ob- A comparison between Figs. 2, 3 and Figs. 6, 7 shows that,
servation points along the tower are summarized in Table II. It for both first and subsequent return strokes, the presence of the
can be seen that the current peak value and maximum steepness tower results in a significant increase of the electromagnetic
at the bottom of the tower are 30% to 50% larger than those field peak, and also in the appearance of subsidiary peaks.
associated with the injected current. Furthermore, for the first For the subsequent return stroke, however, the effect of the
stroke, the time to current peak decreases as the observation tower is much more pronounced [compare Figs. 2(b), 3(b) with
point moves toward the ground while for the subsequent stroke, Figs. 6(b), 7(b)]. For the considered case, the peak field associ-
it remains constant and equal to the time to peak of the injected ated with tower strokes is about 2 times (for the first stroke) to 3
current. This result can be explained considering that for the first times (for the subsequent stroke) as large as that corresponding
stroke, the propagation time along the tower is smaller than the to return strokes initiated at ground level. Furthermore, it can
risetime of the injected current. The situation is reversed for the be seen that the radiation term of both electric and magnetic
subsequent stroke where the propagation time along the tower fields, responsible for the initial peak, is the one most affected
is larger than the risetime of the injected current. by the presence of the tower.
RACHIDI et al.: CURRENT AND ELECTROMAGNETIC FIELD ASSOCIATED WITH LIGHTNING-RETURN STROKES TO TALL TOWERS 361

TABLE II
PARAMETERS OF LIGHTNING RETURN STROKE CURRENT ALONG THE TOWER

Fig. 6. Vertical electric field at 2 km from the 553-m high tower. (a) First return
stroke. (b) Subsequent return stroke. Fig. 7. Azimuthal magnetic field at 2 km from the 553-m high tower. (a) First
return stroke. (b) Subsequent return stroke.

Table III-A and -B summarizes the main parameters of the the tower affects also the values for the maximum steepness.
electric and magnetic field waveforms for both lightning strike Note further that, in accordance with the results of Table II, the
to ground and to the tower. It can be seen that the presence of time to peak of subsequent stroke fields is not affected by the
362 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 43, NO. 3, AUGUST 2001

TABLE III
(A) PARAMETERS OF LIGHTNING RETURN STROKE “NON-CONTAMINATED” CHANNEL-BASE CURRENT AND THE CORRESPONDING ELECTRIC AND MAGNETIC
FIELDS AT 2 km. LIGHTNING STRIKE TO GROUND. (B) PARAMETERS OF LIGHTNING RETURN STROKE “NON-CONTAMINATED” CHANNEL-BASE CURRENT AND THE
CORRESPONDING ELECTRIC AND MAGNETIC FIELDS AT 2 km. LIGHTNING STRIKE TO THE TOWER

tower and is determined by the time to peak of the originating

current. On the other hand, the presence of the tower results in
a decrease by a factor of two in the time to peak of first-stroke
fields.
The contribution of the tower to the total electric field caused
by a subsequent stroke is presented in Fig. 8. This figure shows
that the tower, which in this case represents the bottom 553 m
of the radiating antenna, has therefore a dominant effect upon
the electromagnetic field at 2 km.
Similar results have been presented by Diendorfer [20] who
extended the DU model [34] to take into account the presence of
an elevated strike object. Although Diendorfer did not consider
reflections at the top and at the bottom of the strike object, and
while the strike objects considered in his study were not taller
than 20 m, the theoretical results showed that the increase in
the field peak and in the field derivative peak was more appre-
Fig. 8. Contribution of the tower to the vertical electric field calculated at 2
ciable for lightning return stroke currents with faster risetimes. km from the tower (subsequent stroke).
For a 20-m high strike object and a current 10%–90% risetime
of 0.25 s, an increase in the field peak by a factor of 2.4 was
ment with recent experimental observations of lightning strikes
observed [20].
to the CN Tower presented in [30], [35]. It is found in [30]
that subsequent return stroke magnetic field peak is on average
IV. FIRST VERSUS SUBSEQUENT RETURN STROKE greater than that of first return stroke. This is illustrated in Figs. 9
ELECTROMAGNETIC FIELD PEAK
and 10, in which two sets of simultaneous data (current deriva-
Computed results presented in Section III have shown that the tive, current, E-field, H-field) corresponding to the first (Fig. 9)
presence of the tower tends in general to increase the electro- and to the third stroke (Fig. 10) of the same flash to the CN
magnetic field peak and maximum steepness. This increase de- Tower are presented. It can be seen that despite a smaller peak
pends essentially on the wavefront steepness of the channel-base current, the subsequent stroke electric and magnetic fields are
current, and it has been found to be significantly higher for sub- larger than those from the first stroke.
sequent return strokes, which are characterized by shorter rise- The values of the average ratio of subsequent-to-first mag-
times compared to first return strokes. This finding is in agree- netic field peak to subsequent-to-first current peak are re-
RACHIDI et al.: CURRENT AND ELECTROMAGNETIC FIELD ASSOCIATED WITH LIGHTNING-RETURN STROKES TO TALL TOWERS 363

Fig. 9. Lightning current derivative (top left), lightning current (top right), vertical electric field at 2 km (bottom left), azimuthal magnetic field at 2 km (bottom
right). First return stroke of a CN Tower lightning flash recorded on August 8, 1996.

Fig. 10. As in Fig. 9. Third return stroke.

ported in Table IV. Although the comparison between com- well with the value 1.67 which is obtained from measured
puted results in Section III and experimental data in [30], [35] fields and currents.
should be considered as only qualitative, it can be seen, how- Similar observations, showing that the subsequent-stroke
ever, that the computed value of the ratio of the field peaks field peak might be greater than the first-stroke peak, have
to current peaks , agrees reasonably been made on natural lightning in Florida [36]. In [36], electric
364 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 43, NO. 3, AUGUST 2001

TABLE IV V. PEAK CURRENT ESTIMATES USING REMOTE

RATIO OF SUBSEQUENT-TO-FIRST MAGNETIC FIELD PEAK H =H TO
ELECTROMAGNETIC FIELD DATA
SUBSEQUENT-TO-FIRST CURRENT PEAK I =I
Lightning channel-base currents are established by direct
measurements using either instrumented towers (e.g., [4], [5],
[7], [27], [38]–[41]) or artificially-initiated lightning by small
rockets (e.g., [42]–[45]). Estimates of various lightning current
parameters can also be obtained from the measurements of
lightning electromagnetic fields assuming one or more empir-
ical [1], [2] or theoretical [3] relations between electromagnetic
field and lightning current.
The indirect estimation of lightning current parameters from
measured fields has grown in importance in the last years due
to the widespread dispersion of the lightning location systems
(LLS). The basic aim of such systems is to provide density maps
of lightning flashes. However, more recently, LLS have also
field waveforms from 46 multiple-stroke flashes were analyzed been used to estimate lightning current parameters (e.g., [37])
and it has been found that 15 flashes (33%) had at least one using empirical formulae in which return stroke current peaks
subsequent stroke whose initial electric field peak was greater are assumed to be simply proportional to the remote electric
than that of the first return stroke. and/or magnetic field peaks. Because of the enormous amount
Subsequent-stroke field peak greater than the first-stroke of data they can provide and the possibility of obtaining local
peak is generally attributed to subsequent-stroke current peak statistical data, it is expected that LLS will become more and
greater than the first-stroke peak (e.g., [36], [37]). Indeed, more important in the near future. However, the discussion given
although on average first return stroke current peaks are in Section IV supports the concept already presented in [6],
higher than subsequent stroke ones, observed lightning data [11] that such simple relations assuming a direct proportion-
in Switzerland [38] and in Canada [35] have shown that a ality between currents and fields are not sufficient to extract
nonnegligible number of flashes contain subsequent strokes reliable lightning current statistics. There are several reasons
with higher current peaks than those of their first strokes. which make the application of this type of relations worth of
more careful investigations. It is not only the peak value of the
— Five (15%) of the 33 negative downward multiple- lightning current, but also the parameters discussed in the pre-
stroke flashes striking instrumented towers in Switzer- vious section, namely, the steepness of the current wavefront
land contained one or two subsequent strokes with re- and the return stroke velocity, along with other parameters such
turn stroke peak currents greater than their respective as the tortuosity of the lightning channel trajectory [46], that are
first stroke peak currents. The percentage of subse- affecting the corresponding electromagnetic field. Furthermore,
quent strokes with greater current peaks than the first the presence of an elevated strike object influences profoundly
stroke peak was about 7% (8 strokes out of 115) [36], the magnitude of the electromagnetic field, since lightning cur-
[38]. rent propagates within an elevated strike object at practically the
— Data recorded at the CN Tower in Toronto (about ten speed of light, and this must be taken into account when current
times as high as the towers used in Switzerland) have data directly measured from instrumented towers are used to cal-
shown that 40% of subsequent strokes had current ibrate LLS performances relevant to lightning current statistics.
peaks ranging from one to four times the value of the
first stroke peak [35].
VI. CONCLUSIONS
On the other hand, it is important to realize that even a sub-
sequent return stroke with a lower current peak than the first, In this paper, an analysis of the electric and magnetic fields
could result in a greater field peak. This can be due to several radiated by lightning-return strokes to tall towers has been pre-
factors. One of these, is the fact that the subsequent return stroke sented. Electric and magnetic field waveforms corresponding
velocity is in general greater than the first return stroke speed to typical first and subsequent strokes have been computed and
[29]. The electromagnetic field radiated by the lightning channel analyzed making use of the MTL model extended to take into
depends strongly on the value of the return-stroke velocity and account the presence of an elevated strike object. The effect of
a larger value for the return-stroke velocity results in a greater the presence of a tower on the magnitude and shape of the fields
electromagnetic field peak. The other factor is the steepness of has been investigated. The contributions of the various compo-
the current wavefront. Since subsequent strokes are associated nents of the fields, namely, static, induction and radiation for
with shorter risetimes than first return strokes, their increased the electric field, and induction and radiation for the magnetic
steepness will make a larger contribution to the electromagnetic field have been illustrated and discussed. It was shown in par-
field. Finally, the mere presence of an elevated strike object in- ticular that the radiation term of the electromagnetic field is the
creases the electromagnetic field and that influence is larger for one most affected by the presence of the tower. Furthermore,
currents with steeper wavefronts, as shown by the results ob- it was found that the presence of the tower tends, in general,
tained in the present study. to increase substantially electromagnetic field peaks and their
RACHIDI et al.: CURRENT AND ELECTROMAGNETIC FIELD ASSOCIATED WITH LIGHTNING-RETURN STROKES TO TALL TOWERS 365

derivatives. This increase is mainly caused by the presence of [11] S. Guerrieri, C. A. Nucci, F. Rachidi, and M. Rubinstein, “On the in-
two oppositely propagating current wavefronts originating from fluence of elevated strike objects on directly measured and indirectly
estimated lightning currents,” IEEE Trans. Power Delivery, vol. 13, pp.
the tower top and by the very high velocity (practically speed of 1543–1555, Oct. 1998.
light) of current propagation within the tower, and depends es- [12] F. Fuchs, “On the transient behavior of the telecommunication tower
sentially on the wavefront steepness of the channel-base current. at the mountain Hoher Peissenberg,” in Proc. 24th Int. Conf. Lightning
Protection, Birmingham, U.K., Sept. 1998.
For that reason, the electromagnetic field peak has been found [13] G. Diendorfer, M. Mair, W. Schultz, and W. Hadrian, “Lightning cur-
to be significantly higher for subsequent return strokes, which rent measurements in Austria—Experimental setup and first results,” in
are characterized by faster risetimes compared with first return Proc. 25th Int. Conf. Lightning Protection, Rhodes, U.K., Sept. 18–22,
2000.
strokes. [14] V. A. Rakov, “Transient response of a tall object to lightning,” IEEE
The obtained results have been shown to be consistent with Trans. Electromagn. Compat., 2001, submitted for publication.
recent experimental observations of lightning strokes to the CN [15] C. A. Nucci, “Lightning-induced voltages on overhead power lines. Part
II: Coupling models for the evaluation of the induced voltages,” Electra,
Tower and they qualitatively explain the fact that, because of no. 162, pp. 121–145, 1995.
other parameters, subsequent return strokes with lower current [16] C. A. Nucci, C. Mazzetti, F. Rachidi, and M. Ianoz, “On lightning return
peaks could result in higher field peaks. This observation is to be stroke models for LEMP calculations,” in Proc. 19th Int. Conf. Lightning
Protection, Graz, Apr. 1988.
taken into account when data from tall instrumented towers are [17] C. A. Nucci and F. Rachidi, “Experimental validation of a modification
used to calibrate LLS performances relevant to lightning current to the transmission line model for LEMP calculations,” in Proc. 8th Int.
statistics. Symp. Electromagnetic Compatibilty, Zurich, Switzerland, March 7–9,
1989, pp. 389–394.
[18] V. Rakov and M. A. Uman, “Review and evaluation of lightning return
stroke models including some aspects of their application,” IEEE Trans.
ACKNOWLEDGMENT Electromagn. Compat., vol. 40, pp. 403–426, Nov. 1998.
[19] M. A. Uman, D. K. McLain, and E. P. Krider, “The electromagnetic
The authors wish to thank the Swiss Defense Agency for pro- radiation from a finite antenna,” Amer. J. Phys., vol. 43, pp. 33–38, 1975.
viding them with electromagnetic-field measuring equipment, [20] G. Diendorfer, “Effect of an elevated strike object on the lightning elec-
Dr. V. Shostak for his helpful contributions to the manuscript, tromagnetic fields,” in Proc. 9th Int. Symp. Electromagnetic Compati-
bility, Zurich, Switzerland, Mar. 1991, pp. 235–238.
and Prof. V. Rakov for his valuable comments. [21] F. Rachidi, M. Ianoz, C. A. Nucci, and C. Mazzetti, “Modified Trans-
mission Line Model for LEMP calculations. Effect of the return stroke
velocity decreasing and elevated strike objects on close fields,” in Proc.
REFERENCES 9th Int. Conf. Atmospheric Electricity, St. Petersburg, Russia, June 1992.
[22] W. Janischewskyj, V. Shostak, and A. M. Hussein, “Comparison of light-
[1] J. C. Willett, J. C. Bailey, V. P. Idone, R. E. Orville, A. Eybert-Be- ning electromagnetic field characteristics of first and subsequent return
rard, and L. Barret, “Submicrosecond intercomparison of radiation fields strokes to a tall tower: I magnetic field,” in Proc. 24th Int. Conf. Light-
and currents in triggered lightning return strokes based on the transmis- ning Protection (ICLP), Birmingham, U.K., Sept. 1998, pp. 245–251.
sion-line model,” J. Geophys. Res., vol. 94, no. D11, pp. 13 275–13 286, [23] F. Heidler and Ch. Hopf, “Lightning current and lightning electromag-
1989. netic impulse considering current reflection at the earth’s surface,” in
[2] V. A. Rakov, R. Thottappillil, and M. A. Uman, “On the empirical for- Proc. 22nd Int. Conf. Lightning Protection, Budapest, Hungary, 1994,
mula of Willett relating lightning return-stroke peak current and peak Paper R4-05.
electric field,” J. Geophys. Res., vol. 97, no. D11, pp. 11 527–11 533, [24] F. Rachidi, C. A. Nucci, M. Ianoz, and C. Mazzetti, “Influence of a lossy
1992. ground on lightning-induced voltages on overhead lines,” IEEE Trans.
[3] F. Rachidi and R. Thottapillil, “Determination of lightning currents Electromagn. Compat., vol. 38, Aug. 1996.
from far electromagnetic fields,” J. Geophys. Res., vol. 98, no. D10, pp. [25] F. Rachidi and C. A. Nucci, “On the Master, Lin, Uman, Standler and
18 315–18 321, 1993. the modified transmission line lightning return stroke current models,”
[4] B. N. Gorin and A. V. Shkilev, “Measurements of lightning currents at J. Geophys. Res., vol. 95, pp. 20 389–20 394, Nov. 1990.
the Ostankino tower” (in Russian), Elektrichestvo, vol. 8, pp. 64–65, [26] V. Shostak, W. Janischewskyj, A. M. Hussein, J. S. Chang, and B. Kordi,
1984. “Return stroke current modeling of lightning striking a tall tower ac-
[5] O. Beierl, “Front shape parameters of negative subsequent strokes mea- counting for reflections within the growing channel and for upward-con-
sured at the Peissenberg tower,” in Proc. 21st Int. Conf. Lightning Pro- necting discharges,” in Proc. Int. Conf. Atmospheric Electricity (ICAE),
tection, Berlin, Sept. 21–25, 1992. Guntersville, AL, June 1999.
[6] W. Janischewskyj, A. M. Hussein, V. Shostak, P. Dziurewicz, and W. A. [27] K. Berger, R. B. Anderson, and H. Kroninger, “Parameters of lightning
Chisholm, “Analysis of electromagnetic fields from lightning strokes flashes,” Electra, no. 41, 1975.
to the Toronto CN Tower and from lightning in the surrounding area,” [28] F. Heidler, “Analytische Blitzstromfunktion zur LEMP- Berechnung,”
in Proc. CIGRE Symp. Power System Electromagnetic Compatibility, in Proc. 18th Int. Conf. Lightning Protection, Munich, Sept. 16–20,
Lausanne, Switzerland, Oct. 18–20, 1993, Paper 100-08. 1985, paper 1.9, pp. 63–66.
[7] E. Montandon and B. Beyeler, “The lightning measuring equipment on [29] D. M. Mach and W. D. Rust, “Photoelectric return stroke velocity and
the Swiss PTT telecommunications tower at St. Chrischona, Switzer- peak current estimates in natural and triggered lightning,” J. Geophys.
land,” in Proc. 22nd Int. Conf. Lightning Protection, Budapest, Hungary, Res., vol. 94, no. D11, pp. 13 237–13 247, 1989.
Sept. 19–23, 1994. [30] M. Abdel-Rahman, W. Janischewskyj, A. M. Hussein, F. Rachidi, and J.
[8] H. Motoyama, W. Janischewskyj, A. M. Hussein, R. Rusan, W. A. S. Chang, “Statistical analysis of magnetic field due to CN Tower mul-
Chisholm, and J. S. Chang, “Electromagnetic field radiation model for tistroke flashes,” in Proc. 24th Int. Conf. Lightning Protection, Birm-
lightning strokes to tall structures,” IEEE Trans. Power Delivery, vol. ingham, U.K., Sept. 1998.
11, pp. 1624–1632, July 1996. [31] C. A. Nucci, C. Mazzetti, F. Rachidi, and M. Ianoz, “Analyze du champ
[9] S. Guerrieri, F. Heidler, C. A. Nucci, F. Rachidi, and M. Rubinstein, “Ex- électromagnétique dû à une décharge de foudre dans les domaines tem-
tension of two return stroke models to consider the influence of elevated porel et fréquentiel,” Ann. Télécommun., vol. 43, no. 11/12, 1988.
strike objects on the lightning return stroke current and the radiated elec- [32] W. Janischewskyj, A. M. Hussein, V. Shostak, I. Rusan, J. X. Li, and J.
tric field,” in Proc. Int. Conf. EMC, Sept. 1996. S. Chang, “Statistics of lightning strikes to the Toronto Canadian Na-
[10] R. Rusan, W. Janischewskyj, A. M. Hussein, and J. S. Chang, “Com- tional tower (1978–1995),” IEEE Trans. Power Delivery, vol. 12, pp.
parison of measured and computed electromagnetic fields radiated from 1210–1221, July 1997.
lightning strikes to the Toronto CN Tower,” in Proc. 23rd Int. Conf. [33] W. Janischewskyj, A. M. Hussein, and V. Shostak, “Propagation of light-
Lightning Protection (ICLP), Florence, Italy, Sept. 23–27, 1996, pp. ning current within the CN Tower,” in Proc. Int. CIGRE Colloq. Insula-
297–303. tion Coordination, Toronto, Canada, Sept. 1997, paper 33-2.10.
366 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 43, NO. 3, AUGUST 2001

[34] G. Diendorfer and M. A. Uman, “An improved return stroke model with Wasyl Janischewskyj (F’77) was born in Prague,
specified channel-base current,” J. Geophys. Res., vol. 95, no. D9, pp. Czechoslovakia, in 1925. He received the B.A.Sc.
13 621–13 644, 1990. and M.A.Sc. degrees in electrical engineering
[35] W. Janischewskyj, A. M. Hussein, and J. S. Chang, “Characteristics of from the Ukrainian Technical-Husbandry Institute,
CN Tower multistroke flashes,” in Proc. 10th Int. Symp. High Voltage Regensburg, Germany, the Technical University
Engineering, Montreal, Canada, Aug. 1997, pp. 29–34. of Hanover, Hanover, Germany and the University
[36] R. Thottappillil, V. A. Rakov, and M. A. Uman, “Lightning subsequent- of Toronto, in 1952 and 1954, respectively, and
stroke electric field peak greater than the first stroke peak and multiple the Honorary Doctorate Degree from he National
ground terminations,” J. Geophys. Res., vol. 97, no. D7, pp. 7503–7509, Technical University of Ukraine “Kyiv Polytechnic
May 20, 1992. Institute”, Kiev, Ukraine, in 1998.
[37] K. L. Cummins, E. P. Krider, and M. D. Malone, “The US national From 1955 to 1959, he was with the Aluminum
lightning detection network and applications to cloud-to-ground light- Laboratories Ltd., Kingston, ON, Canada. Since 1959, he has been with the
ning data by electric power utilities,” IEEE Transactions Electromagn. University of Toronto, Toronto, ON, Canada, where he has been a Lecturer, a
Compat., vol. 40, pp. 465–480, Nov. 1998. Professor, the Assistant Chairman, and later the Associate Dean, and since 1990,
[38] K. Berger, “Methoden und Resultate der Blitzforschung auf dem Monte Professor Emeritus. From 1960 to 1965, he was with Ontario Hydro, Toronto,
San Salvatore bei Lugano in den Jahren 1963–1971,” Bull. SEV, vol. 63, ON, Canada, including the Coldwater Project, and has been on sabbatical leaves
pp. 1403–1422, 1972. at the University of Liverpool, Liverpool, U.K., the High Voltage Institute of the
[39] E. Garbagnati and G. B. Lo Piparo, “Parameter von Blitzstromen,” Technical University of Munich, Munich, Germany, the Federal University of
ETZa), vol. 103, no. 2, 1982. Rio de Janeiro, Rio de Janeiro, Brazil, the High Voltage Laboratory of Elec-
[40] A. J. Eriksson, “Lightning and tall structures,” Trans. South Afr. IEE, pt. tricite de France, Clamart, France, and Trench Electric, Toronto, ON, Canada.
8, vol. 69, pp. 238–252, 1978. He has authored and co-authored some 150 papers in the areas of electric-power
[41] A. M. Hussein, W. Janischewskyj, J. S. Chang, V. Shostak, W. A. transmission, corona, eletromagnetic interference, lightning and power system
Chisholm, P. Dziurewych, and Z. I. Kawasaki, “Simultaneous mea- stability.
surement of lightning parameters for strokes to the Toronto Canadian He is involved with many International Technical Committees including IEC
National Tower,” J. Geophys. Res., vol. 100, no. D5, pp. 8853–8861, TC42, CIGRE SC 33 and 36, EPRI, IEEE, COMPIMERA, Canadian Electricity
1995. Association and Canadian Standards Association, and is a member of the IEEE
[42] J. C. Willett, V. P. Idone, R. E. Orville, C. Leteinturier, A. Eybert-Be- Power Engineering Society, IEEE Working Group on Estimating the Lightning
rard, L. Barret, and E. P. Krider, “An experimental test of the “transmis- Performance of Transmission Lines, IEEE Working Group on Lightning Perfor-
sion-line model” of electromagnetic radiation from triggered lightning mance of Distribution Lines, and the IEEE Electromagnetic Compatibility So-
return stroke,” J. Geophys. Res., vol. 93, no. D4, pp. 3867–3878, 1988. ciety, a member and a former Chairman of the IEEE Subcommittee on Corona
[43] J. C. Willett, J. C. Bailey, V. P. Idone, R. E. Orville, A. Eybert-Be- and Field Effects, and a Registered Professional Engineer in the Province of
rard, and L. Barret, “Submicrosecond intercomparison of radiation fields Ontario, Canada.
and currents in triggered lightning return strokes based on the transmis-
sion-line model,” J. Geophys. Res., vol. 94, no. D11, pp. 13 275–13 286,
1989.
[44] C. Leteinturier, C. Wideman, and J. Hamelin, “Current and electric field
derivatives in triggered lightning return strokes,” J. Geophys. Res., vol.
95, no. D1, pp. 811–828, Jan. 1990.
[45] E. P. Krider, C. Leteinturier, and J. C. Willet, “Submicrosecond fields
radiated during the onset of first return strokes in cloud-to-ground light-
ning,” J. Geophys. Res., vol. 101, no. D1, pp. 1589–1597, Jan. 20, 1996.
[46] W. Janischewskyj, V. Shostak, A. M. Hussein, and W. Chisholm, “Esti-
mation of lightning location system accuracy using CN Tower lightning
data,” in 23rd Int. Conf. Lightning Protection (ICLP), Florence, Italy,
Sept. 23–27, 1996, pp. 215–223.

Ali M. Hussein (M’80–SM’90) received the B.Sc.

Farhad Rachidi was born in Geneva, Switzerland, degree from Alexandria University, Alexandria,
in 1962. He received the M.S. and the Ph.D. degrees, Egypt, in 1967, the M.Sc. degree from Ain-Shams
both in electrical engineering, from the Swiss Fed- University, Cairo, Egypt, in 1972, and the Ph.D.
eral Institute of Technology, Lausanne, Switzerland, degree from the University of Toronto, Toronto,
in 1986 and 1991, respectively. Canada, in 1979, all in electrical engineering.
He worked at the Power Systems Laboratory of the From 1967 to 1974, he was an Assistant Lecturer
Swiss Federal Institute of Technology until 1996. In and Demonstrator at Ain-Shams University. From
1997, he joined the Lightning Research Laboratory of 1979 to 1980, he was a Research Associate at the
the University of Toronto, Toronto, Canada and from April 1998 until September National Research Council of Canada, Ottawa,
1999, he was with Montena EMC, Rossens, Switzerland. He is currently “Maître Canada. From 1980 to 1982, he was an Assistant
d’Enseignement et de Recherche” at the Swiss Federal Institute of Technology. Professor at Riyadh University, Riyadh, Saudi Arabia. From 1982 to 1986,
His research interests concern EMC and in particular lightning and EMP interac- he was a Research Associate at the Electrical Engineering Department of the
tions with transmission lines. Dr. Rachidi is author or coauthor of more than 80 University of Toronto. From 1986 to 1988, he was a member of the Scientific
scientific papers published in reviewed journals and presented at international Staff, the Electromagnetics Technology Division, Bell-Northern Research,
conferences. Ottawa, Canada. Since 1989, he has been a Professor at Ryerson University,
He is a member of the CIGRE Working Groups 33.01 “Lightning” and 36.07 Toronto, Canada, and an Adjunct Associate Professor at the University of
“Power Quality Indices and Objectives,” and of IEEE Working Group on Light- Toronto. Dr. Hussein has authored and co-authored over 90 publications in
ning Performances of Distribution Lines, the Scientific Committee of the Inter- the areas of Microwave Ferrites, Microwave Acoustics, Electromagnetic Field
national Conference on Lightning Protection, and Chairman of the tutorials and computations and, more recently, measurement and analysis of fast transients
workshops of the International Zurich Symposium on EMC. (Microgap Discharge, Electromagnetic Pulse, and Lightning Discharge).
RACHIDI et al.: CURRENT AND ELECTROMAGNETIC FIELD ASSOCIATED WITH LIGHTNING-RETURN STROKES TO TALL TOWERS 367

Carlo Alberto Nucci was born in Bologna, Italy, in Behzad Kordi was born in Tehran, Iran in 1969. He
1956. He received the degree with honors, and the received the B.S., M.S., and Ph.D. (with distinction)
Ph.D. degree, both in in electrical engineering, from degrees in electrical engineering from Amirkabir
the University of Bologna, Bologna, Italy, in 1981 University of Technology (Tehran Polytechnic),
and 1986, respectively. Tehran, Iran, in 1992, 1995, and 2000, respectively.
In 1982, he joined the Power Electrical En- He joint the Lightning Studies Group of the Uni-
gineering Institute, University of Bologna, as a versity of Toronto, Toronto, Canada in 1998 where
Researcher, and becam an Associate Professor in he was awarded a Graduate Research Grant from the
1992, and a full Professor in Power Systems, in 2000. Electrical and Computer Engineering Department.
He is author or co-author of more than 100 scientific Dr. Kordi has recently joined Shahed University,
papers published on reviewed journals or presented Tehran, Iran as an Assistant Professor and is also
at international conferences. He has been responsible for several International collaborating with the Electromagnetics Lab of Amirkabir University. His
projects supported from the Italian side by the National Research Council (the research interest focuses on numerical methods in electromagnetics and various
Italian National Science Foundation) and involving foreign universities, such electromagnetic compatibility (EMC) topics, especially lightning and EMP.
as the Swiss Federal Institute of Technology, Lausanne, Switzerland, (Power Dr. Kordi was the recipient of the Amirkabir University of Technology Stu-
network laboratory), the University of Florida, Gainesville, the University of dent of the Year Award in 1995.
Illinois, Urbana Champaign, and the University of Sydney, Sydney, Australia.
His research interests concern power systems transients and dynamics, with
particular reference to lightning and nuclear EMP impact on power lines and to
voltage collapse, power station simulators and the study of power components
including medium voltage capacitors and traction batteries. Jen-Shih Chang (M’90–SM’96) received B. Eng.
He is member of the IEEE Working Group “Lightning” performance of Dis- and M. Eng. degrees in electrical engineering
tribution lines, of the CIGRE Working group 33.01 “Lightning” (of which he is from the Musashi Institute of Technology, Tokyo,
also secretary) and has been member of some technical committees of the Italian Japan, and the Ph.D. degree in Experimental Space
Electrical Commission in charge of producing technical standards. Sciences from York University, Toronto, Canada.
During 1973–1974, he was a Researcher at the
Centre de Recherches en Physique de l’Environ-
nement (CNRS), France. From 1975 to 1979, he
Silvia Guerrieri was born in Modena, Italy, in 1968. was a project Scientist/Assistant Professor with the
She received a degree with honors in electrical engi- Department of Physics and Center for Research
neering and the Ph.D. degree from the University of in Experimental Space Sciences, York University,
Bologna, Bologna, Italy, in 1993 and 1997, respec- York, U.K.. From 1979 to 1986, he was an Assistant/Associate Professor with
tively. the Department of Engineering Physics, McMaster University, Hamilton, ON,
Her doctoral work includes research at the Power Candada. During 1985 to 1996, he was a visiting professor with the Musashi
Systems Laboratory, Swiss Federal Institute of Tech- Institute of Technology, Tokyo, Japan; Tokyo Denki University, Tokyo,
nology, Lausanne, Switzerland, and participation Japan; Tokyo University, Tokyo, Japan; University of Sevilla, Sevilla, Spain;
in the 1995 Triggered Lightning Campaign, Camp Joseph Fourier University, Grenoble, France; University of Poitiers, Poitiers,
Blanding, FL, coordinated by the University of France; Oita University, Oita, Japan; and Tokyo University of Agriculture and
Florida, Gainesville. From April 1997, she is with Technology, Tokyo, Japan. Since 1987, he has been a Professor at Mc Master
University of Modena and Reggio Emilia, Modena, Italy. Her main scientific University, and is involved in research on applied electrostatics, lightning, air
interest concerns lightning and nuclear electromagnetic pulse effects on power pollution control, solid and liquid waste destruction plasma technologies.
systems. She is author or co-author of about 20 scientific papers published on Dr. Chang is currently the chair of IEEE DEIS Electrohydrodynamics Tech-
reviewed journals or presented at international conferences. nical Committee.