On the proportion of upward flashes to lightning research towers

Atmospheric Research 129–130 (2013) 110–116

Contents lists available at SciVerse ScienceDirect

Atmospheric Research journal homepage: www.elsevier.com/locate/atmos

On the proportion of upward flashes to lightning research towers Alexander Smorgonskiy a,⁎, Farhad Rachidi a, Marcos Rubinstein b, Gerhard Diendorfer c, Wolfgang Schulz c a b c

EMC Laboratory, Ecole polytechnique fédérale de Lausanne, SCI-STI-GR-FR, Station 11, CH-1015, Lausanne, Switzerland University of Applied Sciences of Western Switzerland, Route de Cheseaux 1, 1400 Yverdon-les-Bains, Switzerland Austrian Electrotechnical Association (OVE), Dept. ALDIS (Austrian Lightning Detection & Information System), Kahlenberger Str. 2a, A-1190, Vienna, Austria

a r t i c l e

i n f o

Article history: Received 7 March 2012 Received in revised form 8 August 2012 Accepted 10 August 2012 Keywords: Upward lightning Downward lightning Towers Lightning incidence Mountains

a b s t r a c t We compare in this paper direct measurements obtained on the towers on San Salvatore Mountain (Switzerland) and on the Gaisberg Mountain (Austria). They are situated in similar topographical environments but in different lightning activity zones. Direct measurements of lightning currents on these towers have revealed a major difference in terms of the number of downward flashes. While measurements made by Berger and co-workers revealed a significant number of downward flashes on the two towers on San Salvatore Mountain, more recent observations on the Gaisberg and Peissenberg towers were essentially composed of upward flashes. We use in this paper a new method to estimate the proportion of upward/ downward flashes to a given tower, based on the data from lightning location systems. The analysis using the proposed method explains the discrepancy in terms of the measured number of downward flashes in the Gaisberg and San Salvatore towers. The analysis presented reveals also that in the evaluation of the percentage of upward flashes initiated from a tall structure, different parameters should be carefully examined, namely (i) the value of the ground flash density, (ii) the topographical conditions, and (iii) the presence of other tall structures in the region from which upward flashes might be initiated. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Long-term observations and measurements of lightning discharges to tall structures have been conducted since the 1930s following the construction of high-rise buildings and the development of radio and television (Hagenguth and Anderson, 1952). Skyscrapers and telecommunications towers have been widely used for lightning research, while some towers have been built and used exclusively for the measurement of lightning parameters (see Chapter 6 by Rakov and Uman, 2003).

⁎ Corresponding author. Tel.: +41 21 693 48 18, +41 78 934 78 91 (Mob.); fax: +41 21 693 46 62. E-mail addresses: alexander.smorgonskiy@epfl.ch (A. Smorgonskiy), farhad.rachidi@epfl.ch (F. Rachidi), [email protected] (M. Rubinstein), [email protected] (G. Diendorfer), [email protected] (W. Schulz). 0169-8095/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.atmosres.2012.08.014

The design, installation and maintenance of a lightning recording system on a tower present a very complex, expensive and challenging task. Among thousands of tall structures erected in Europe, there are only a few that have been equipped with a lightning measurement system for lightning research that is still in operation (e.g. Gaisberg Tower in Austria (Diendorfer et al., 2009), Peissenberg Tower in Germany (Heidler et al., 2001), and, most recently, the Säntis Tower in Switzerland which was instrumented in 2010 (Romero et al., 2012)). Earlier in the mid-20th century, long-term observations of lightning currents using two instrumented towers on the Mountain San Salvatore in Switzerland resulted in the most complete statistical characterization to date of lightning current parameters (Berger et al., 1975). One important difference in the measurement results obtained on the Mountain San Salvatore towers and more recent data obtained on the Gaisberg and Peissenberg towers

A. Smorgonskiy et al. / Atmospheric Research 129–130 (2013) 110–116

concerns the downward/upward flash ratio. While downward flashes were rarely observed at the Gaisberg and Peissenberg towers and constitute only about 1% of all recorded flashes (Diendorfer et al., 2009), nearly 30% of the recorded flashes on the Mountain San Salvatore were of downward type (Berger, 1967). The evaluation of the lightning incidence to tall towers situated in mountainous regions is more difficult than on flat ground due to the fact that topological factors play a major role in the enhancement of the electric field at the top of the mountain. To account for topological effects such as the height and geometrical morphology of the mountain, the structure's actual height above ground level is replaced with an effective height heff, whose evaluation is a complex task (Theethayi et al., 2004). Different techniques for the estimation of heff were summarized by Zhou et al. (2010). For the Mountain San Salvatore towers (70‐m tall located on a 910-m high mountain), the resulting effective heights range from 198 m to 350 m. For the Gaisberg Tower (100-m tall on a 1287-m high mountain), the obtained values for the effective heights range from 274 m to 1000 m. In this paper, which constitutes an extended version of Smorgonskiy et al. (2011a), we will discuss the differences in the percentage of upward flashes observed on the Mountain San Salvatore towers and the Gaisberg Tower analyzing the data from lightning location systems (LLS) and comparing it to the direct measurements performed at the selected towers. 2. Direct lightning incidence observations for the selected towers San Salvatore Mountain (Monte San Salvatore in Italian) is located in the Southern Alps near the city of Lugano in

A

B

111

Switzerland. The mountain is 915 m high on the shore of Lake Lugano. Significantly higher mountains encircle the lake: Mount Bré (925 m) on the North-East, Mount Sighignola (1302 m) on the East, Mount Generoso on the South-East (1701 m), Mount San Giorgio (1097 m) on the South, and Mount Lema (1621 m) on the West as shown in Fig. 1a. The distances from San Salvatore Mountain to the above listed summits range from 4 to 11 km. Therefore, thunderstorms usually develop and move above all of these mountains. The first tower for lightning measurements was constructed on the summit of San Salvatore Mountain in 1943. It was replaced by a radio and television tower in 1958. Still, the measurement of lightning discharges continued on this new tower. In 1950, a second lightning research tower was constructed 400 m to the North from the first one. Both towers were 70 m tall. Later, the second tower was demolished and nowadays only the TV and radio tower is present on the summit. For this study, we have selected the results of the direct measurements made during the period 1955–1963, for which the flashes to both towers are clearly distinguished (Berger, 1967). Gaisberg is a 1287 m tall mountain in the Northern Alps situated to the East of the city of Salzburg in Austria. Other summits located nearby are Gurlspitze (1158 m), Schwarzenberg (1334 m) and Ochsenberg (1487 m), all of them South-East of the Gaisberg Mountain. An elevation map showing their locations is given in Fig. 2a. A 100-m tall radio and TV transmitter was built on the top of the Gaisberg Mountain in 1953. Since 1998, lightning flashes to the tower are recorded by the ALDIS group and measurements still continue today. In our analysis, we have used the data for the period 2000–2007 reported by Diendorfer et al. (2009).

C

Fig. 1. San Salvatore Mountain region, the North is on the top. Each figure shows the same area in geographic projection: latitude 45.882–46.072 N, longitude 8.852–9.042 E. (a) Topographic map. (b) Lightning flash density map obtained from the EUCLID LLS data during 2000–2012. Square dots indicate locations of upward flashes registered with a photo camera by Berger (1967) from structures listed in Section 3. (c) Locations of downward flashes observed by Berger (1967). In the grayed areas the ground was not visible from the observation point, therefore only the total number of downward flashes in those areas is given in the figure (see Section 3).

112

A. Smorgonskiy et al. / Atmospheric Research 129–130 (2013) 110–116

a

b

c

Fig. 2. Gaisberg Mountain region, the North is on the top. Each figure shows the same area in geographic projection: latitude 47.709–47.899 N, longitude 13.015– 13.205 E. (a) Topographic map. Square dot indicates the location of the tower. (b) Lightning flash density map. (c) Locations of cloud-to-ground lightning flashes registered by EUCLID LLS during five years: 2007–2011.

A summary of the observed annual incidence of upward and downward flashes for the two towers on San Salvatore Mountain and that on the Gaisberg Mountain based on the data for the selected time periods is presented in Table 1. The towers are situated in regions with different lightning activities: the ground flash density near Salzburg is Ng = 2.5 flashes km−2 y−1, while the region of the Lugano lake is characterized by the higher value of Ng =3.8 flashes km−2 y−1. The values of the ground flash densities around the towers were derived from the LLS data as discussed further in Section 5. Note that, in general, the higher the value of Ng, the higher the number of downward flashes attracted to a tower. However, the total number of flashes (including both upward and downward) recorded annually at the Gaisberg Tower is two to three times higher than for each of the towers on San Salvatore Mountain. A structure with the same height as one of the towers on San Salvatore Mountain but situated on a flat surface (in an area where Ng = 4 flashes km −2 y −1) would be struck approximately once every two years, as estimated using the equation by Eriksson and Meal (1984). The same object on a mountaintop could get struck 17 to 25 times annually (see Table 1). As mentioned earlier, this is mainly due to the

increment of the electric field introduced by the mountain resulting in the initiation of upward flashes. As can be seen from Table 1, measurements made by Berger and co-workers revealed a significant number of downward flashes on the San Salvatore towers (24% on Tower 1 and 29% on Tower 2), whereas observations on the Gaisberg Tower were essentially composed of upward flashes (99%). This difference resulted in some doubts about the used procedure of classification of Berger and co-workers. In this article we propose to use an independent method to estimate this proportion and compare it with the results of the direct measurements. Figs. 1b and 2b will be further discussed in Section 4. 3. Photographic records of lightning activity in the San Salvatore Region In addition to the lightning current measurements at the San Salvatore towers, photographic records of lightning flashes around the tower were performed in 1955–1965. It was possible to take photos only during the night. Therefore, only about one quarter of the total number of flashes was recorded.

Table 1 Lightning incidence to the selected towers. Tower

Tower 1 on Mt. San Salvatore Tower 2 on Mt. San Salvatore Gaisberg Tower a

Average lightning incidence, flashes y−1

Upward flashes

Upward negative flashesa

Downward

Upward

%

ICCRS, p, %

ICConly, %

6 5 b1

19 12 57

76 71 99

36

64

52

48

References

Berger (1967) Berger (1967) Diendorfer et al. (2009)

ICCRS: flashes composed of an initial continuous current with or without superimposed pulses followed by return strokes, ICCp: flashes composed of an initial continuous current with fast superimposed pulses, ICConly: flashes containing an initial continuous current.

A. Smorgonskiy et al. / Atmospheric Research 129–130 (2013) 110–116

The locations of more than a thousand downward flashes were carefully estimated and plotted (Berger, 1967) and they are represented in Fig. 1c. In the areas shown in gray, it was impossible to determine the terminations of the flashes and, therefore, only the total number of downward flashes observed within the approximate area is given. Downward flashes were observed everywhere within the area of study, including the lake, the valleys and the mountains. In contrast, upward flashes were only observed at a few locations shown in Fig. 1b, namely from metal structures on the mountain peaks. They include the towers on San Salvatore Mountain itself, a 49-m tall transmitter on Mount Generoso, a flagpole on Mount Sighignola, a bell-tower in the village of Bré on the slope of Mount Boglia and a church in the Biogno village. The presence of these objects creates the conditions necessary for upward lightning initiation from the mountain peaks on which they are situated. 4. Determination of lightning incidence to the towers using a lightning location system Lightning location systems (LLS) are excellent tools for studying the lightning distribution in different regions. Several tens of sensors forming the EUCLID network (Diendorfer, 2002) record lightning flashes over the whole of Europe. The assessment of the performance of lightning location systems can be evaluated by means of directly measured events provided by either instrumented towers (Schulz and Diendorfer, 2004) or rocket-triggered lightning (Jerauld et al., 2005). Direct observation on the Gaisberg Tower was used to evaluate the performance of the EUCLID network (Diendorfer, 2010). The flash detection efficiency was found to be more than 97% for peak currents greater than 5 kA. The median location accuracy was 368 m with a standard deviation of 768 m. Nevertheless, it is important to realize that lightning location systems are not able to capture all lightning flashes due to several reasons. Upward lightning flashes could be composed of an initial continuous current only (ICConly type) with amplitude up to several hundred amperes. These flashes do not produce significant electromagnetic radiation and they cannot be detected by sensors far away from the tower. Therefore, they cannot be detected remotely. Sometimes, fast pulses are superimposed on the initial continuous current (ICCp type) or the initial continuous current with or without superimposed pulses can also be followed by return strokes (ICCRS type). The detection efficiency for ICCp flashes depends strongly on the rise time and peak current of the superimposed pulses. As can be seen in Table 1, a considerable fraction of the upward flashes is of the ICConly type. In San Salvatore Mountain, only 36% of the upward flashes were of ICCp and/or ICCRS type. This fraction was found to be higher for the Gaisberg Tower (52%). For this study, we have used the lightning locations obtained by the EUCLID network around the selected towers during the time period 01/01/2000–31/12/2011. An example of these data is shown in Fig. 2c. The areas of study were divided into cells of 0.01 by 0.01° and, for each cell, the value of the lightning flash density was calculated. The map showing the distribution of cloud-to-ground (CG) flashes around San Salvatore Mountain is shown in Fig. 1b. Although the overall

113

ground flash density is around 4 flashes km−2 y−1, some cells exhibit much higher flash densities. A notable example is the central cell, where the tower on the Mountain San Salvatore is located. As suggested by Smorgonskiy et al. (2011b), this difference is mainly due to the fact that an object placed on the mountain top initiates additional upward flashes. Indeed, we observe the increase of lightning flash density in the cells where the objects mentioned in Section 3 are situated, from which upward flashes are initiated. On other summits like Mount Lema, Mount Bré and Mount San Giorgo, there are also small structures or crosses installed. There is, however, no photographic evidence of upward flashes from them (it is possible that in 1950–60s there were no objects there). Nevertheless, since the distribution of lightning flash density is similar to the cells where upward flashes were recorded, we attribute this difference to the initiation of upward flashes as well. Observations from several tall objects located closely together. The map of CG lightning flash density around the Gaisberg Tower is shown in Fig. 2b. It could be observed that the lightning flash density close to the tower differs greatly from the flash density at a distance from it. We assume that the increase in the flash density is essentially due to the upward flashes initiated from the Gaisberg Tower. Comparing Figs. 1 and 2, we can see two notable differences between the Gaisberg and the San Salvatore towers. First, the ground flash density in the San Salvatore tower region is 50% higher than in the Gaisberg region. Second, other tall structures are present on the mountains surrounding the San Salvatore tower whose locations coincide with an increase in the ground flash density. As we will discuss in Section 5, this increase is likely to be due to upward flashes created by the structures themselves. Similar observations of multiple upward flashes from tall structures located closely together have been recently reported by Warner (2012). Contrary to what is observed for the San Salvatore tower, the mountains in the vicinity of the Gaisberg Tower (Fig. 2) do not exhibit a significant increase in the ground flash density since they do not have structures at their mountaintops. 5. Analysis of the proportion of upward flashes To evaluate the number of upward flashes associated with a given object, we propose to use the method described in Smorgonskiy et al. (2011b), which is based on the assumption that the number of downward lightning flashes in the vicinity of a tall structure is only marginally affected by the presence of the object. The method consists of considering two concentric areas around the tall object as shown in Fig. 3. The inner one is a circle with a radius r = 1 km enclosing the object and the second area is a ring with inner radius r and outer radius R, typically 8–10 km. The increment of the flash density in the inner circle is assumed to be mainly due to the upward lightning discharges which were initiated from the tower and, thus, could be observed only in that central circle. At the same time, some downward flashes within this circle are attracted to the tower (depending on their charge and distance from the tower) while some are not. All of these downward flashes, however, are assumed to be detected by the LLS within the inner circle and, therefore, the presence of the tower does not increase the density of downward flashes.

114

A. Smorgonskiy et al. / Atmospheric Research 129–130 (2013) 110–116

Fig. 3. Areas used for the evaluation of the number of upward flashes from a tall structure.

need to apply a correction factor of 100/(ICCRS, p, %). This will give us the total number of upward flashes, including both ICConly and ICCRS, p, see last column of Table 3. For example, for the case of the Tower on San Salvatore Mountain, the average number of flashes annually registered by the EUCLID LLS within the circle of r = 1 km radius is 16.4. According to the value of Ng and Eq. (3), there are on average 11.8 downward flashes recorded within this circle every year. Therefore, the number of upward flashes is 4.6 and all of them are assumed to be initiated from the tower. Since only 36% of flashes are of ICCRS, p type, we need to increase the estimated number of upward flashes by a factor of 2.8. However, our goal is to compare the percentage of upward flashes for the selected towers. Therefore, we need to find the number of downward flashes that were attracted to the towers. This number Ndown_tower is usually determined by the attractive radius Ra which can be estimated using different models (Rakov and Uman, 2003). Here, we will not use those models to calculate the value of Ndown_tower. Instead, we will consider Ndown_tower as a variable and we will study the variation of the proportion of upward flashes as a function of the number of downward flashes attracted to the structure. This allows us to avoid additional assumptions used in the model calculation of the attractive radius. For the San Salvatore Tower, this implies that the number of downward flashes to the tower Ndown_tower is between 0 and 11.8 flashes y −1. Hence, the percentage of upward flashes can be calculated by way of,

To estimate the number of upward flashes to the selected towers, we first need to calculate the value of the ground flash density Ng from the distribution of CG flashes in the outer ring (see Fig. 3, R = 8 km, r = 1 km). The overall number of flashes recorded by the LLS within the central circle (see Fig. 3) is 2

Nall ¼ D1 ⋅S1 ¼ D1 ⋅π⋅r ;

ð1Þ

where D1 is the lightning flash density in the central circle with the surface S1. The value of Nall is composed of the number of upward flashes Nup_tower initiated from the tower and downward flashes Ndown: Nall ¼ N up

tower

þ Ndown :

ð2Þ

We assume that the number of downward flashes within the central circle is essentially unaffected by the presence of the tower. Under this assumption, the number of downward flashes is determined by the value of Ng around the tower. However part of these flashes is attracted to the tower Ndown_tower, the remaining part of downward flashes is not attached to the tower Ndown_out: 2

Ndown ¼ S1 ⋅Ng ¼ π⋅r ⋅Ng ¼ Ndown

tower

þ Ndown

out :

ð3Þ

Finally, since we know the total number of downward flashes within the central circle, we can estimate the number of upward flashes from the tower: Nup

2

tower

¼ Nall −π⋅r ⋅N g :

Nup ; % ¼

Nup

NupXtower þ Ndown

tower

tower

ð5Þ

⋅100

ð4Þ and it varies from 100 to 52%. The obtained results for this example for five values of Ndown_tower from 0 to 11.8 are given in Table 3. The same calculation procedure was also applied to the lightning density data around the Gaisberg Tower. Fig. 4 shows the variation of the percentage of the upward flashes

The results of the estimations using Eqs. (1)–(4) are given in Table 2 for Tower 1 on San Salvatore and for the Gaisberg Tower. The value of Nup_tower includes only flashes with ICCRS, p since only they are detected by the LLS. In order to compare these results with the direct measurements from Table 1, we

Table 2 Estimation of the number of flashes for the selected towers from LLS data. Tower

flashes km Tower 1 on Mt. San Salvatore Gaisberg Tower

Nall

Ng

3.8 2.1

−2 −1

y

flashes y 16.4 33.8

Ndown

Nup_tower (ICCRS, p)

Nup_tower (ICConly, ICCRS, p)

11.8 6.6

4.6 27.2

12.8 52.3

−1

A. Smorgonskiy et al. / Atmospheric Research 129–130 (2013) 110–116 Table 3 Percentage of upward flashes as a function of the number of downward flashes attracted to the tower, example of San Salvatore Tower. Nup_tower

Ndown_tower

Nup, %

12.8 12.8 12.8 12.8 12.8

11.8 9 6 3 0

52 56 68 86 100

as a function of the number of downward flashes for both, the Gaisberg Tower and San Salvatore towers. In the figure, we have used a dot to represent the value obtained from the direct measurements for each tower (see Table 1). When the exact proportion of upward flashes for a tower is unknown (for instance for non-instrumented towers), a plot of Eq. (5) as a function of Ndown_tower varying from Ndown_tower = 0 to π ⋅ r 2 ⋅ Ng can be used to characterize the tower. It can be seen from Fig. 4 that the value of Nup, % for the Gaisberg Tower is indeed much higher than that for the Tower on San Salvatore Mountain irrespective of the value of Ra and the number of downward flashes attracted to the tower. This observation by means of the new method based on the lightning flash densities obtained from LLS data explains the discrepancy in terms of the measured number of downward flashes to the towers on the Gaisberg Mountain and on San Salvatore Mountain. Note that the measured data in Fig. 4 (the dots) were recorded when both towers, separated by a distance of 400 m, were still standing on the Mountain San Salvatore, while the curve labeled “San Salvatore” in the same figure is based on lightning location data obtained when one of the towers had already been removed. In other words, we have assumed that the presence of the second tower does not have a significant influence on the ratio of upward to downward flashes.

115

The number of upward flashes is influenced at least by three factors: the ground flash density, the topographical conditions, and the number of tall structures in the region. It seems reasonable to assume that the number of upward flashes to a given test tower increases with the ground flash density. One way to test this hypothesis would be to compare regions with similar topographical conditions and different ground flash densities. The other parameters can be tested in a similar manner, by comparing regions for which the two other parameters are similar. At present, the relative influence of each one of these effects is unknown. If we assume that the number of upward flashes to a given test tower indeed increases with the ground flash density and decreases with the presence of other tall structures in the region, we can offer the following explanation: other tall objects diminish the charge available for upward flashes from the test tower.

6. Conclusions Direct measurements of lightning currents obtained at the towers on San Salvatore Mountain (Switzerland) and the Gaisberg Tower (Austria) have revealed a major difference in terms of the number of upward flashes. While Berger and co-workers obtained a significant number of downward flashes, more recent observations on the Gaisberg Tower were essentially composed of upward flashes. Due to this difference, there was a need to assess Berger's upward/ downward classification of flashes to the San Salvatore Tower. An independent method, based on the data from LLS was applied to find the number of upward flashes initiated from the towers. The results of our analysis are in agreement with Berger's results. The analysis revealed that in the evaluation of the percentage of upward flashes initiated from a tall structure, different parameters should be carefully examined, namely (i) the value

Fig. 4. Proportion of upward flashes as a function of the number of downward flashes to the structure using Eq. (5) and data from EUCLID LLS.

116

A. Smorgonskiy et al. / Atmospheric Research 129–130 (2013) 110–116

of the ground flash density, (ii) the topographical conditions, and (iii) the presence of other tall structures in the region from which upward flashes might be initiated. The ground flash density around the San Salvatore Mountain is 50% higher compared to the Gaisberg region. The San Salvatore Mountain is not the dominant summit in its region. On each of the peaks surrounding the San Salvatore Mountain, there are structures also producing upward flashes. Even though the Gaisberg Mountain is also surrounded by high mountains, there are no other structures on their summits and there are no areas where we observe upward flashes. The relative influence of each one of the three abovementioned effects is currently unknown and calls for further research. Appendix A. Supplementary data Supplementary data associated with this article can be found in the online version, at http://dx.doi.org/10.1016/j. atmosres.2012.08.014. These data include Google maps of the most important areas described in this article. References Berger, K., 1967. Novel observations on lightning discharges—results of research on Mount San Salvatore. J. Franklin Inst. 283, 478–525 http:// dx.doi.org/10.1016/0016-0032(67)90598-4. Berger, K., Anderson, R.B., Kroninger, H., 1975. Parameters of lightning flashes. Electra 41, 23–37. Diendorfer, G., 2002. EUCLID—technical structure and performance of the European wide lightning location system. Proc. Int. Conf. on Grounding and Earthing and 3rd Brazilian Workshop on Atmospheric Electricity, Rio de Janeiro, Brazil. Diendorfer, G., 2010. Lightning location systems performance validation using lightning to towers. Proc. 21st Int. Lightning Detection Conf. (ILDC) and 3rd Int. Lightning Meteorology Conf. (ILMC), Orlando, Florida.

Diendorfer, G., Pichler, H., Mair, M., 2009. Some parameters of negative upward initiated lightning to the Gaisberg Tower (2000–2007). IEEE Trans. Electromagn. Compat. 51, 443–452 http://dx.doi.org/10.1109/ TEMC.2009.2021616. Eriksson, A.J., Meal, D.V., 1984. The incidence of direct lightning strikes to structures and overhead lines. Proc. Int. Conf. on Lightning and Power Systems, London, UK, pp. 67–71. Hagenguth, J.H., Anderson, J.G., 1952. Lightning to the Empire State Building. Part III. AIEE Trans. 71, 641–649 http://dx.doi.org/10.1109/ AIEEPAS.1952.4498521. Heidler, F., Wiesinger, J., Zischank, W., 2001. Lightning currents measured at a telecommunication tower from 1992 to 1998. Proc. 14th Int. Zurich Symposium on Electromagnetic Compatibility, Zurich, Switzerland. Jerauld, J., Rakov, V.A., Uman, M.A., Rambo, K.J., Jordan, D.M., 2005. An evaluation of the performance characteristics of the U.S. National Lightning Detection Network in Florida using rocket-triggered lightning. J. Geophys. Res. 110, D19106 http://dx.doi.org/10.1029/2005JD005924. Rakov, V.A., Uman, M.A., 2003. Lightning: Physics and Effects. Cambridge University Press. Romero, C., Paolone, M., Rubinstein, M., Rachidi, F., Rubinstein, A., Diendorfer, G., Schulz, W., Daout, B., Kälin, A., Zweiacker, P., 2012. A system for the measurements of lightning currents at the Säntis Tower. Electr. Power Syst. Res. 82, 34–43 http://dx.doi.org/10.1016/ j.epsr.2011.08.011. Schulz, W., Diendorfer, G., 2004. Lightning peak currents measured on tall towers and measured with lightning location systems. Proc. 18th Int. Lightning Detection Conf. ILDC 2004, Helsinki, Finland. Smorgonskiy, A., Rachidi, F., Rubinstein, M., Diendorfer, G., Schulz, W., 2011a. On the proportion of upward flashes to lightning research towers. Proc. 7th Asia-Pacific Int. Conf. on Lightning (APL), Chengdu, China, pp. 858–862 http://dx.doi.org/10.1109/APL.2011.6110248. Smorgonskiy, A., Rachidi, F., Rubinstein, M., Diendorfer, G., Schulz, W., Korovkin, N., 2011b. A new method for the estimation of the number of upward flashes from tall structures. Proc. XIth Int. Symp. on Lightning Protection (SIPDA), Fortaleza, Brazil, pp. 97–100 http://dx.doi.org/ 10.1109/SIPDA.2011.6088466. Theethayi, N., Diendorfer, G., Thottappillil, R., 2004. On determining the effective height of Gaisberg Tower. Proc. European Electromagnetics EUROEM 2004, Magdeburg, Germany. Warner, T., 2012. Observations of simultaneous upward lightning leaders from multiple tall structures. J. Atmos. Res. 117, 45–54. Zhou, H., Theethayi, N., Diendorfer, G., Thottappillil, R., Rakov, V.A., 2010. On estimation of the effective height of towers on mountaintops in lightning incidence studies. J. Electrostat. 68, 415–418 http:// dx.doi.org/10.1016/j.elstat.2010.05.014.