- Email: [email protected]
An analysis of the initiation of upward flashes from tall towers with particular reference to Gaisberg and Säntis Towers
Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
Contents lists available at ScienceDirect
Journal of Atmospheric and Solar-Terrestrial Physics journal homepage: www.elsevier.com/locate/jastp
An analysis of the initiation of upward flashes from tall towers with particular reference to Gaisberg and Säntis Towers Alexander Smorgonskiy a,n, Alaleh Tajalli a, Farhad Rachidi a, Marcos Rubinstein b, Gerhard Diendorfer c, Hannes Pichler c a
EMC Laboratory, EPFL, Lausanne, Switzerland IICT, HEIG-VD, Yverdon-les-Bains, Switzerland c ALDIS, OVE, Vienna, Austria b
art ic l e i nf o
a b s t r a c t
Article history: Received 4 March 2015 Received in revised form 18 June 2015 Accepted 23 June 2015 Available online 17 July 2015
In this paper, we present an analysis of the lightning events preceding initiation of upward lightning flashes from the Gaisberg and the Säntis Towers. It is found that the majority of upward lightning discharges from both towers are initiated without any preceding lightning activity. We show also that the results of the presented studies on the initiation of upward flashes from tall structures might be affected by the selected parameters of the study, namely the time and distance intervals used to identify the triggering events. Preceding events had the same polarity as triggered flashes in the case of the Säntis Tower and had opposite polarities in the case of the Gaisberg Tower. The effect of seasonal and temperature variations have been also analyzed. & 2015 Elsevier Ltd. All rights reserved.
Keywords: Lightning Upward lightning Downward lightning Self-initiated flashes Other-triggered flashes Tower Lightning location system
1. Introduction Understanding the mechanisms of initiation of upward lightning flashes is an important issue in lightning research. It has been suggested in early studies (e.g., Berger et al., 1975; McEachron, 1939) that upward flashes originated from tall towers could be actually initiated by preceding lightning activity in the surrounding area. Recently, the interest in understanding the mechanism of the initiation of upward flashes has increased, essentially because of the continuous growth in the number of tall structures, such as wind turbines. On the other hand, the dramatic improvement of the lightning research equipment, including instrumented towers for direct lightning current measurements, video observations, and large-scale availability of data from lightning location systems (LLS) has made it possible to gather valuable experimental data. Several studies on this topic have been performed at various locations around the world (Heidler et al., 2014; Manhardt et al., 2012; Smorgonskiy et al., 2014; Wang and Takagi, 2012; Warner et al., 2012; Zhou et al., 2012), with sometimes contradictory n
Corresponding author. Fax: þ41 21 693 46 62. E-mail address: alexander.smorgonskiy@epfl.ch (A. Smorgonskiy).
http://dx.doi.org/10.1016/j.jastp.2015.06.016 1364-6826/& 2015 Elsevier Ltd. All rights reserved.
conclusions. In some cases, upward lightning discharges from tall structures were found to be initiated without any preceding lightning activity, while in other cases, preceding events seem to play a significant role in their initiation. In the literature, the latter are referred to as ‘other-triggered’ or ‘nearby-lightning-triggered’ flashes, while the former are called ‘self-triggered’ or ‘self-initiated’ flashes. In other words, the word triggered (or initiated) is used in the literature and in this paper to refer to the relation between flashes initiated on tall objects and other lightning activity that precedes them within given spatial and temporal limits. It is important to note, however, that the causality implied by the word triggered has not been established at this time. Very recently, (Rubinstein et al., 2015) used a probabilistic model to show that it is possible to explain at least some of the lightning activity prior to tower flashes as being the result of chance. In this paper, which is an extended version of (Smorgonskiy et al., 2014), we present further analyses and comparisons aiming at better understanding the underlying mechanisms of the initiation of upward flashes from tall structures, with particular reference to the data obtained at two instrumented towers: the Gaisberg Tower in Austria (Diendorfer et al., 2009) and the Säntis Tower in Switzerland (Romero et al., 2012).
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
47
190 m away from the tower. Unlike the observations made in Rapid City, 90% of the 41 analyzed upward flashes were found to be ‘self-initiated’.
2. Previous studies 2.1. Uchinada, Japan The objects of this study were a wind turbine and its lightning protection tower with heights of 100 and 105 m, respectively, and separated by a distance of 45 m. Both of them are located on a hill of 40 m above the sea level. Lightning strikes to the wind turbine and the protection tower were recorded during 6 winter seasons. Lightning currents were measured at the bottom of both structures using Rogowski coils. Electric fields were measured using a field mill as well as slow and fast capacitive antennas. This research was initially focused on studying the reduction of lightning incidence to the wind turbine due to the protection tower, but further analysis of the data allowed observing different mechanisms of upward lightning initiation (Wang and Takagi, 2012). Half of the upward flashes (actually, initiated from both the wind turbine and the tower) were self-initiated, whereas the other half were preceded by lightning activity (‘other-triggered’ or ‘nearbylightning-triggered’). No classification of the triggering events was given in the published study. 2.2. Rapid City, USA Observations made at Rapid City were focused on upward lightning discharges from ten towers with heights varying from 91 to 191 m, located on a ridge of 180 m above the surrounding terrain. The towers are distributed over a distance of 8 km on a line stretched from North to South. The presence of so many towers within a relatively small area and the fact that multiple simultaneous upward flashes from different towers were observed, makes this location unique among the other studies and should be taken into consideration when comparing the observed results. Two methods were used in (Warner et al., 2012) to distinguish self-initiated and other-triggered upward flashes. The first one was based on the data from the US National Lightning Detection Network (NLDN) within a circle of 200 km radius centered on the towers. It was found that the majority (83%) of upward flashes were preceded by nearby lightning activity. To overcome the limitations of NLDN in detecting all the flashes (Cummins and Murphy, 2009), optical observations were also used. These revealed that the percentage of the upward flashes preceded by nearby lightning activity was even higher (99%).
2.4. Gaisberg Tower, Austria The 100-m tall Gaisberg Tower is located on a 1287-m high mountain. The tower has been the object of two separate studies on lightning events preceding upward flashes from the tower. The first study (Zhou et al., 2012) was based on a combination of the electric field variations measured at a distance of 170 m from the tower and observations from the EUCLID lightning detection network (Schulz et al., 2005). In contrast with the Rapid City study and in agreement with Peissenberg, most upward flashes from the Gaisberg Tower (87%) were not preceded by any nearby lightning activity. Positive cloud-to-ground (þCG) strokes were the most frequent triggering event, in agreement with the previous study.
3. Procedure used in this study In this study, we focus on upward flashes from two towers: the Gaisberg Tower in Austria (Diendorfer et al., 2009) and the Säntis Tower in Switzerland (Romero et al., 2012). The Säntis Tower is characterized by a height of 124 m and it is located on the top of Mount Säntis (2502 m high above sea level). Such high altitude is a unique characteristic among other considered tall structures. The analysis method which was adopted in the present study is as follows: 1. A list of lightning flashes directly measured at the tower (either Gaisberg or Säntis) and an extract from the European Lightning Detection Network EUCLID database with lightning strokes preceding each flash from the tower were merged together. 2. A circle of a given radius around the tower and a maximum time interval between preceding lightning events and observed upward flashes are selected. 3. Upward flashes without preceding lightning events observed within the limits specified in step 2 are categorized as ‘self-initiated’. Other upward flashes are labeled as ‘other-triggered’ and are further classified depending on the polarity and type of the preceding lightning event. Two types of lightning events are reported by EUCLID and are used in this study: cloud-to-ground (CG) and intracloud (IC) lightning discharges.
2.3. Peissenberg Tower, Germany The Peissenberg Tower (160-m tall) used in this study is located on the mountain 'Hoher Peissenberg' in the South of Germany. The tower is instrumented for lightning current measurements (Manhardt et al., 2012). The altitude above mean sea level is about 940 m. The electric field was measured at a distance of about
4. Results and analysis In Section 2 and Table 1, a brief comparison of the percentage of upward flashes with preceding lightning events was presented. A significant difference was observed between the locations in
Table 1 Comparison of the ratio of self-initiated and other-triggered flashes observed at different locations. Location
Uchinada, Japan
Tower height, m 100, 105 Location altitude, m 40 Study period 2005–2010 (winter) Method Electric field Total upward 53 flashes Other-triggered 25 (47%) Self-initiated 28 (53%) Reference Wang and Takagi (2012)
Rapid City, USA
Peissenberg, Germany
Gaisberg, Austria
Säntis, Switzerland
91 to 191 1165–1340 2004–2010 (Apr– Sept) NLDN Optic 81 81
160 940 1996–1999 (all year)
2000–2013 (all year)
124 2502 2011–2012 (all year)
Electric field 41
100 1287 2005–2009 (all year) Electric field 205
EUCLID 759
EUCLID 326
4 (10%) 37 (90%) Heidler et al. (2014) and Manhardt et al. (2012)
26 (13%) 179 (87%) Zhou et al. (2012)
121 (16%) 638 (84%) Smorgonskiy et al. (2014)
48 (15%) 278 (85%) Smorgonskiy et al. (2014)
67 (83%) 80 (99%) 14 (17%) 1 (1%) Warner et al. (2012)
48
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
Fig. 1. Locations of the Säntis, Peissenberg and Gaisberg Towers in the northern Alps.
different continents. At the same time, however, a strong similarity between the towers in Europe has been found. The results of our study for the Gaisberg and the Säntis Towers are presented in the last two columns in Table 1. Unlike Rapid City, but similarly to the Peissenberg Tower, it can be seen that the majority of upward lightning discharges from the Gaisberg and the Säntis Towers (84% and 85%, respectively) are initiated without any preceding lightning activity. Interestingly, while upward flashes from the Gaisberg Tower were mostly triggered by positive lightning events, flashes from the Säntis Tower were mostly preceded by negative lightning events. It is worth noting that the three European towers considered in this study, Gaisberg (Austria), Peissenberg (Germany) and Säntis (Switzerland) are located in a similar geographical area, namely in the northern Alps as shown in Fig. 1. This might explain the similarities obtained in these three different sites, although, as we will see in the next section, notable differences also exist. 4.1. Polarity of other-triggered upward flashes and lightning events preceding them A major difference between the towers was observed when the polarities of the preceding lightning events and the triggered upward flashes were analyzed. As can be seen from Table 2, negative upward flashes from the Säntis Tower were triggered by negative CG or negative IC flashes, while positive and bipolar flashes were triggered by preceding positive flashes. Conversely, upward flashes from the Gaisberg Tower were triggered by preceding events of opposite polarity. The reasons for these differences are currently unknown. A more detailed analysis over a more extended time period should Table 2 Comparison of the polarities of the triggered upward flashes and lightning events preceding them. Tower
Observation period
Gaisberg 2005–2009
Säntis
2011–2012
Preceding events
Triggered upward flashes
15 Positive CGa 1 Negative CG 10 Positive and negative ICb
15 1 8 1 1 1 1 2 16 1 25
1 Positive CG 3 Positive IC 17 Negative CG 25 Negative IC
a b
CG – cloud-to-ground lightning flash. IC – intracloud lightning flash.
Negative Positive Negative Positive Bipolar Positive Positive Bipolar Negative Positive Negative
be carried out to clarify this issue. 4.2. Distances and time intervals between the upward flash initiated from the tower and preceding lightning events We have also compared statistically the spatial distances and time intervals between the upward flash initiated from the tower and preceding lightning events (Table 3), for Rapid city, Gaisberg and Säntis. From this Table, it can be noted that the statistical parameters of the time intervals are very similar for all three towers. The distances between the tower and preceding events for the case of the Gaisberg Tower and Rapid City are very similar. However, lightning triggering events at the Säntis Tower occur at significantly closer distances to the tower. 4.3. Seasonal variations Seasonal variations are known to influence the occurrence of lightning discharges (Rakov and Uman, 2007). We have analyzed and compared the occurrence of self-initiated and other-triggered lightning strikes for the five locations listed in Table 1 during summer and winter seasons. The results are summarized in Fig. 2. The measurements at Uchinada were reported for the winter period only (Wang and Takagi, 2012). Between 2004 and 2010, there have been no direct observations of upward lightning from the towers in Rapid City during the winter months (Warner et al., 2012). The three towers in Europe show, despite their similar high proportion of self-initiated flashes, different seasonal characteristics. All self-initiated flashes at the Peissenberg Tower were recorded during the winter period. Winter self-initiated flashes at the Gaisberg Tower represent 58% of the total number of upward flashes. However, this value reduces to only 8% for the case of the Säntis Tower. The majority (91%) of upward flashes occur in summer time at the Säntis Tower, like in Rapid City. Nevertheless, the majority of the Säntis flashes are self-triggered, as opposed to Rapid City. As it has been described in Section 4, the Peissenberg, Gaisberg and Säntis Towers are situated on the top of mountains, Table 3 Comparison of the statistical parameters for the time and distance intervals between upward flashes and preceding events. Tower
Distance between the tower and Time from preceding event to preceding event, km upward leader, ms Min
Rapid City 3.5 Gaisberg 0.3 Säntis 0.05
Mean
Median
Max
Min
Mean
Median
Max
17.5 18.9 7.2
14.8 16.3 1.5
50.5 48.3 46.8
o1 o1 o1
72 102 109
46 33 49
404 819 600
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
49
70 Other−triggered Self−initiated 60
Number
50 40 30 20 Fig. 2. Seasonal proportions of self-initiated and other-triggered upward flashes at different sites.
10
rising to 940 m, 1287 m and 2502 m above sea level, respectively. We suppose that those seasonal variations could be attributed to the differences in the altitudes, but further research is needed to study this hypothesis.
0
4.4. Temperature variations Seasonal variations mentioned above imply that air temperature also affects the mechanism of upward flash initiation. Indeed, in Zhou et al. (2014), it was shown that, at the Gaisberg Tower, relatively low ambient temperatures (mean value 2.8 °C) facilitate the initiation of upward flashes without any nearby preceding lightning activity. On the other hand, relatively high ambient temperatures (mean value þ11.1 °C) seem to be correlated with the initiation of upward other-triggered flashes. The distribution of the air temperatures measured at the top of the Gaisberg Tower at the moments when upward flashes were initiated is shown in Fig. 3. For the Gaisberg Tower, the temperature measurements were taken every minute. Other parameters, like wind speed, relative humidity and pressure were found to have no direct influence on upward flash initiation. Meteorological parameters at the Säntis Tower are measured at the tower base and not at the top of the tower, as is the case for the Gaisberg Tower. Also, temperature measurements are reordered every ten minutes. An analysis similar to the reported Gaisberg study has been performed for the case of the Säntis Tower. The obtained temperature distribution is shown in Fig. 4. Since the majority of upward flashes of both types, self- and other-triggered, 35 Other−triggered Self−initiated
30
Number
25 20
−11 −9 −7 −5 −3 −1 1 3 5 7 Temperature, °C
9
11 13 15
Fig. 4. Distribution of temperatures associated with occurrence of self-initiated and other-triggered upward flashes at the Säntis Tower.
occur at the Säntis Tower during the summer, mean temperatures when these flashes are observed are found to be positive and very close: þ4.2 and þ 4.6 °C, respectively.
5. Practical considerations A number of practical issues can arise when applying the general procedure used in this study or similar methodologies used in previous studies for the analysis of initiation of upward flashes from tall structures. Some of these issues can be handled by applying common rules. They are briefly described in what follows. 5.1. Time and distance intervals When correlating the occurrence of upward lightning flashes on a tower with preceding lightning strokes or flashes from a lightning location system database, the time window and distance to the tower should be limited. A comparison of the time and distance intervals used in previous studies is presented in Table 4. It can be seen that the study in Rapid City used a circle of 200 km radius while, in the Gaisberg study, the circle’s radius was 10 times smaller. On the other hand, the time interval was smaller in the Gaisberg study. Clearly, by increasing the time and distance intervals, it is more likely to find a preceding lightning event. The resulting ratio of self-initiated to other-triggered upward flashes will therefore be affected by the choice of these two parameters. These intervals are strongly linked to the mechanism of upward flash initiation. Several processes can lead to the formation of an upward discharge Table 4 Comparison of the time and distance intervals used to find preceding events for the other-triggered upward flashes from the towers.
15 10
Location
Rapid City
Gaisberg
Study period
2004– 2010 200
2005–2009 2000–2012 2010– 2012 20 50 50
2
5
5 0
−13−11 −9 −7 −5 −3 −1 1 3 5 7 Temperature, °C
9 11 13 15
Fig. 3. Distribution of temperatures associated with the occurrence of self-initiated and other-triggered flashes at the Gaisberg Tower (Fig. 2 in Zhou et al. (2014)).
Radius of observation circle (km) Duration of observation interval (s)
Gaisberg
2
Säntis
2
50
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
20%
15%
10%
5%
0%
Fig. 5. Percentage of other-triggered upward flashes for the Gaisberg Tower as a function of time interval and distance (circle radius).
from the tower. Among the most important are: slow electrostatic field variations and faster radiated field change. However, it is still not clear which one of them plays the dominant role. We have demonstrated this effect for the case of the Gaisberg Tower in Fig. 5, in which the variation of the percentage of othertriggered upward flashes is calculated based on the choice of different time intervals and circle radii. The choice of these intervals will have some effect on all the studies reported in Section 4 since it will change the ratio of selfto other-triggered flashes. Especially, it would be noticeable for the results reported in Table 3. By analyzing the rate of growth of the percentage of upward flashes from Fig. 5, it can be seen that this rate reduces significantly in the circles greater than 50 km and time intervals above 1 s. Therefore, these values can be used as reference values for future studies. Otherwise, parametric study should be performed to examine the influence of these parameters on the resulting percentage of other-triggered upward flashes.
from both towers are initiated without any preceding lightning activity. The observations at the European towers (Gaisberg, Peissenberg and Säntis) differ greatly from the Rapid City study in which most of the flashes were preceded by nearby lightning activity. The study revealed two major differences between the data obtained at the Gaisberg and the Säntis Towers concerning flashes with preceding lightning activity: (1) Upward flashes from the Gaisberg Tower were mostly triggered by positive lightning events, while flashes from the Säntis Tower were mostly preceded by negative flashes, (2) preceding events had the same polarity as triggered flashes in the case of the Säntis Tower and had opposite polarities in the case of the Gaisberg Tower. We showed that the mechanism of upward flash initiation in different sites features a strong seasonal variation, even within Europe. Therefore, the observed differences cannot only be explained to the temperature variations as it has been suggested before. Other parameters such as the altitude of the site and meteorological conditions above the tower should also be taken into account. We showed also that the results of the presented studies on the initiation of upward flashes from tall structures might be affected by the selected parameters of the study and the performance of the LLS to detect IC discharges. Although our results do not enable us to propose optimum values for the radius and the time interval, we recommend the use of a time interval of at least 1 s and an area around the tower of at least 50 km radius. Studies from Rapid City, Gaisberg and Säntis were found to have similar statistical parameters of the time intervals between the upward flashes and their preceding events. However, lightning triggering events at the Säntis Tower were found to occur at significantly closer distances to the tower, compared to Gaisberg and Rapid City. It is worth noting again that the causality implied by the word triggered has not been established at this time and calls for further investigations.
5.2. Sequence of preceding lightning events
Acknowledgments
Some rare upward flashes from the towers were preceded by a series of lightning events, where it was not evident which preceding stroke to consider as a triggering event.
Financial support from the BKW Ecology Fund (Grant number 110005F) is acknowledged.
5.3. Limited detection efficiency of LLS for IC discharges Lightning location systems (LLS) have limited capabilities to detect intracloud (IC) discharges, especially those with small peak fields, and the detection efficiency may depend on the main orientation of the IC discharge channel (mostly vertical similar to CG, or mostly horizontal).
6. Conclusions We presented an analysis of the initiation of upward lightning flashes for the Gaisberg and the Säntis Towers solely based on the data from the EUCLID lightning location system. We have updated a previous study for the Gaisberg Tower, with an increased observation period of 14 years. The proportion of self-initiated and other-triggered flashes for the Gaisberg Tower remained almost unchanged with respect to the previous study, which was carried out over an observation period of 5 years. We presented a similar study for the Säntis Tower, for which we obtained a behavior analogous to that of the Gaisberg Tower and the Peissenberg Tower, namely that the majority of upward lightning discharges
References Berger, K., Anderson, R.B., Kroninger, H., 1975. Parameters of lightning flashes. Electra 41, 23–37. Cummins, K.L., Murphy, M.J., 2009. An overview of lightning locating systems: history, techniques, and data uses, with an in-depth look at the U.S. NLDN. IEEE Trans. Electromagn. Compat. 51, 499–518. http://dx.doi.org/10.1109/ TEMC.2009.2023450. Diendorfer, G., Pichler, H., Mair, M., 2009. Some parameters of negative upwardinitiated lightning to the Gaisberg Tower (2000–2007). Electromagn. Compat. IEEE Trans. 51, 443–452. http://dx.doi.org/10.1109/TEMC.2009.2021616. Heidler, F., Manhardt, M., Stimper, K., 2014. Self-initiated and other-triggered positive upward lightning measured at the Peissenberg Tower, Germany. IEEE, 157–166. http://dx.doi.org/10.1109/ICLP.2014.6973113. Manhardt, M., Heidler, F., Stimper, K., 2012. The electric field of negative upward lightning strikes at the Peissenberg tower, Germany. In: Presented at the Lightning Protection (ICLP), 2012 International Conference on, IEEE, pp. 1–9. 10. 1109/ICLP.2012.6344205. McEachron, K.B., 1939. Lightning to the Empire State Building. J. Frankl. Inst. 227, 149–217. http://dx.doi.org/10.1016/S0016-0032(39)90397-2. Rakov, V.A., Uman, M.A., 2007. Lightning: Physics and Effects. Cambridge University Press, United Kingdom. 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. Rubinstein, M., Smorgonskiy, A., Rachidi, F., Zuber, J., 2015. On the classification of
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51 tower flashes as self-initiated and other-triggered. In: Presented at the Asia Electromagnetics International Symposium ASIAEM, Jeju, Republic of Korea. Schulz, W., Cummins, K., Diendorfer, G., Dorninger, M., 2005. Cloud-to-ground lightning in Austria: a 10-year study using data from a lightning location system. J. Geophys. Res. 110, D09101. http://dx.doi.org/10.1029/2004JD005332. Smorgonskiy, A., Tajalli, A., Rachidi, F., Rubinstein, M., Diendorfer, G., Pichler, H., 2014. Analysis of lightning events preceding upward flashes from Gaisberg and Säntis Towers. In: Lightning Protection (ICLP), 2014 International Conference O, Presented at the Lightning Protection (ICLP), 2014 International Conference o, pp. 1382–1385. 10.1109/ICLP.2014.6973346. Wang, D., Takagi, N., 2012. Characteristics of winter lightning that occurred on a windmill and its lightning protection tower in Japan. IEEJ Trans. Power Energy 132, 568–572. http://dx.doi.org/10.1541/ieejpes.132.568.
51
Warner, T.A., Cummins, K.L., Orville, R.E., 2012. Upward lightning observations from towers in Rapid City, South Dakota and comparison with National Lightning Detection Network data, 2004–2010. J. Geophys. Res. 117, D19109. http://dx.doi. org/10.1029/2012JD018346. Zhou, H., Diendorfer, G., Thottappillil, R., Pichler, H., Mair, M., 2012. Measured current and close electric field changes associated with the initiation of upward lightning from a tall tower. J. Geophys. Res. 117, 9 9 pp., 201210.1029/ 2011JD017269. Zhou, H., Diendorfer, G., Thottappillil, R., Pichler, H., Mair, M., 2014. The influence of meteorological conditions on upward lightning initiation at the Gaisberg Tower. In: Lightning Protection (ICLP), 2014 International Conference O. Presented at the Lightning Protection (ICLP), 2014 International Conference o, pp. 1162–1165. 10.1109/ICLP.2014.6973303.
Contents lists available at ScienceDirect
Journal of Atmospheric and Solar-Terrestrial Physics journal homepage: www.elsevier.com/locate/jastp
An analysis of the initiation of upward flashes from tall towers with particular reference to Gaisberg and Säntis Towers Alexander Smorgonskiy a,n, Alaleh Tajalli a, Farhad Rachidi a, Marcos Rubinstein b, Gerhard Diendorfer c, Hannes Pichler c a
EMC Laboratory, EPFL, Lausanne, Switzerland IICT, HEIG-VD, Yverdon-les-Bains, Switzerland c ALDIS, OVE, Vienna, Austria b
art ic l e i nf o
a b s t r a c t
Article history: Received 4 March 2015 Received in revised form 18 June 2015 Accepted 23 June 2015 Available online 17 July 2015
In this paper, we present an analysis of the lightning events preceding initiation of upward lightning flashes from the Gaisberg and the Säntis Towers. It is found that the majority of upward lightning discharges from both towers are initiated without any preceding lightning activity. We show also that the results of the presented studies on the initiation of upward flashes from tall structures might be affected by the selected parameters of the study, namely the time and distance intervals used to identify the triggering events. Preceding events had the same polarity as triggered flashes in the case of the Säntis Tower and had opposite polarities in the case of the Gaisberg Tower. The effect of seasonal and temperature variations have been also analyzed. & 2015 Elsevier Ltd. All rights reserved.
Keywords: Lightning Upward lightning Downward lightning Self-initiated flashes Other-triggered flashes Tower Lightning location system
1. Introduction Understanding the mechanisms of initiation of upward lightning flashes is an important issue in lightning research. It has been suggested in early studies (e.g., Berger et al., 1975; McEachron, 1939) that upward flashes originated from tall towers could be actually initiated by preceding lightning activity in the surrounding area. Recently, the interest in understanding the mechanism of the initiation of upward flashes has increased, essentially because of the continuous growth in the number of tall structures, such as wind turbines. On the other hand, the dramatic improvement of the lightning research equipment, including instrumented towers for direct lightning current measurements, video observations, and large-scale availability of data from lightning location systems (LLS) has made it possible to gather valuable experimental data. Several studies on this topic have been performed at various locations around the world (Heidler et al., 2014; Manhardt et al., 2012; Smorgonskiy et al., 2014; Wang and Takagi, 2012; Warner et al., 2012; Zhou et al., 2012), with sometimes contradictory n
Corresponding author. Fax: þ41 21 693 46 62. E-mail address: alexander.smorgonskiy@epfl.ch (A. Smorgonskiy).
http://dx.doi.org/10.1016/j.jastp.2015.06.016 1364-6826/& 2015 Elsevier Ltd. All rights reserved.
conclusions. In some cases, upward lightning discharges from tall structures were found to be initiated without any preceding lightning activity, while in other cases, preceding events seem to play a significant role in their initiation. In the literature, the latter are referred to as ‘other-triggered’ or ‘nearby-lightning-triggered’ flashes, while the former are called ‘self-triggered’ or ‘self-initiated’ flashes. In other words, the word triggered (or initiated) is used in the literature and in this paper to refer to the relation between flashes initiated on tall objects and other lightning activity that precedes them within given spatial and temporal limits. It is important to note, however, that the causality implied by the word triggered has not been established at this time. Very recently, (Rubinstein et al., 2015) used a probabilistic model to show that it is possible to explain at least some of the lightning activity prior to tower flashes as being the result of chance. In this paper, which is an extended version of (Smorgonskiy et al., 2014), we present further analyses and comparisons aiming at better understanding the underlying mechanisms of the initiation of upward flashes from tall structures, with particular reference to the data obtained at two instrumented towers: the Gaisberg Tower in Austria (Diendorfer et al., 2009) and the Säntis Tower in Switzerland (Romero et al., 2012).
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
47
190 m away from the tower. Unlike the observations made in Rapid City, 90% of the 41 analyzed upward flashes were found to be ‘self-initiated’.
2. Previous studies 2.1. Uchinada, Japan The objects of this study were a wind turbine and its lightning protection tower with heights of 100 and 105 m, respectively, and separated by a distance of 45 m. Both of them are located on a hill of 40 m above the sea level. Lightning strikes to the wind turbine and the protection tower were recorded during 6 winter seasons. Lightning currents were measured at the bottom of both structures using Rogowski coils. Electric fields were measured using a field mill as well as slow and fast capacitive antennas. This research was initially focused on studying the reduction of lightning incidence to the wind turbine due to the protection tower, but further analysis of the data allowed observing different mechanisms of upward lightning initiation (Wang and Takagi, 2012). Half of the upward flashes (actually, initiated from both the wind turbine and the tower) were self-initiated, whereas the other half were preceded by lightning activity (‘other-triggered’ or ‘nearbylightning-triggered’). No classification of the triggering events was given in the published study. 2.2. Rapid City, USA Observations made at Rapid City were focused on upward lightning discharges from ten towers with heights varying from 91 to 191 m, located on a ridge of 180 m above the surrounding terrain. The towers are distributed over a distance of 8 km on a line stretched from North to South. The presence of so many towers within a relatively small area and the fact that multiple simultaneous upward flashes from different towers were observed, makes this location unique among the other studies and should be taken into consideration when comparing the observed results. Two methods were used in (Warner et al., 2012) to distinguish self-initiated and other-triggered upward flashes. The first one was based on the data from the US National Lightning Detection Network (NLDN) within a circle of 200 km radius centered on the towers. It was found that the majority (83%) of upward flashes were preceded by nearby lightning activity. To overcome the limitations of NLDN in detecting all the flashes (Cummins and Murphy, 2009), optical observations were also used. These revealed that the percentage of the upward flashes preceded by nearby lightning activity was even higher (99%).
2.4. Gaisberg Tower, Austria The 100-m tall Gaisberg Tower is located on a 1287-m high mountain. The tower has been the object of two separate studies on lightning events preceding upward flashes from the tower. The first study (Zhou et al., 2012) was based on a combination of the electric field variations measured at a distance of 170 m from the tower and observations from the EUCLID lightning detection network (Schulz et al., 2005). In contrast with the Rapid City study and in agreement with Peissenberg, most upward flashes from the Gaisberg Tower (87%) were not preceded by any nearby lightning activity. Positive cloud-to-ground (þCG) strokes were the most frequent triggering event, in agreement with the previous study.
3. Procedure used in this study In this study, we focus on upward flashes from two towers: the Gaisberg Tower in Austria (Diendorfer et al., 2009) and the Säntis Tower in Switzerland (Romero et al., 2012). The Säntis Tower is characterized by a height of 124 m and it is located on the top of Mount Säntis (2502 m high above sea level). Such high altitude is a unique characteristic among other considered tall structures. The analysis method which was adopted in the present study is as follows: 1. A list of lightning flashes directly measured at the tower (either Gaisberg or Säntis) and an extract from the European Lightning Detection Network EUCLID database with lightning strokes preceding each flash from the tower were merged together. 2. A circle of a given radius around the tower and a maximum time interval between preceding lightning events and observed upward flashes are selected. 3. Upward flashes without preceding lightning events observed within the limits specified in step 2 are categorized as ‘self-initiated’. Other upward flashes are labeled as ‘other-triggered’ and are further classified depending on the polarity and type of the preceding lightning event. Two types of lightning events are reported by EUCLID and are used in this study: cloud-to-ground (CG) and intracloud (IC) lightning discharges.
2.3. Peissenberg Tower, Germany The Peissenberg Tower (160-m tall) used in this study is located on the mountain 'Hoher Peissenberg' in the South of Germany. The tower is instrumented for lightning current measurements (Manhardt et al., 2012). The altitude above mean sea level is about 940 m. The electric field was measured at a distance of about
4. Results and analysis In Section 2 and Table 1, a brief comparison of the percentage of upward flashes with preceding lightning events was presented. A significant difference was observed between the locations in
Table 1 Comparison of the ratio of self-initiated and other-triggered flashes observed at different locations. Location
Uchinada, Japan
Tower height, m 100, 105 Location altitude, m 40 Study period 2005–2010 (winter) Method Electric field Total upward 53 flashes Other-triggered 25 (47%) Self-initiated 28 (53%) Reference Wang and Takagi (2012)
Rapid City, USA
Peissenberg, Germany
Gaisberg, Austria
Säntis, Switzerland
91 to 191 1165–1340 2004–2010 (Apr– Sept) NLDN Optic 81 81
160 940 1996–1999 (all year)
2000–2013 (all year)
124 2502 2011–2012 (all year)
Electric field 41
100 1287 2005–2009 (all year) Electric field 205
EUCLID 759
EUCLID 326
4 (10%) 37 (90%) Heidler et al. (2014) and Manhardt et al. (2012)
26 (13%) 179 (87%) Zhou et al. (2012)
121 (16%) 638 (84%) Smorgonskiy et al. (2014)
48 (15%) 278 (85%) Smorgonskiy et al. (2014)
67 (83%) 80 (99%) 14 (17%) 1 (1%) Warner et al. (2012)
48
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
Fig. 1. Locations of the Säntis, Peissenberg and Gaisberg Towers in the northern Alps.
different continents. At the same time, however, a strong similarity between the towers in Europe has been found. The results of our study for the Gaisberg and the Säntis Towers are presented in the last two columns in Table 1. Unlike Rapid City, but similarly to the Peissenberg Tower, it can be seen that the majority of upward lightning discharges from the Gaisberg and the Säntis Towers (84% and 85%, respectively) are initiated without any preceding lightning activity. Interestingly, while upward flashes from the Gaisberg Tower were mostly triggered by positive lightning events, flashes from the Säntis Tower were mostly preceded by negative lightning events. It is worth noting that the three European towers considered in this study, Gaisberg (Austria), Peissenberg (Germany) and Säntis (Switzerland) are located in a similar geographical area, namely in the northern Alps as shown in Fig. 1. This might explain the similarities obtained in these three different sites, although, as we will see in the next section, notable differences also exist. 4.1. Polarity of other-triggered upward flashes and lightning events preceding them A major difference between the towers was observed when the polarities of the preceding lightning events and the triggered upward flashes were analyzed. As can be seen from Table 2, negative upward flashes from the Säntis Tower were triggered by negative CG or negative IC flashes, while positive and bipolar flashes were triggered by preceding positive flashes. Conversely, upward flashes from the Gaisberg Tower were triggered by preceding events of opposite polarity. The reasons for these differences are currently unknown. A more detailed analysis over a more extended time period should Table 2 Comparison of the polarities of the triggered upward flashes and lightning events preceding them. Tower
Observation period
Gaisberg 2005–2009
Säntis
2011–2012
Preceding events
Triggered upward flashes
15 Positive CGa 1 Negative CG 10 Positive and negative ICb
15 1 8 1 1 1 1 2 16 1 25
1 Positive CG 3 Positive IC 17 Negative CG 25 Negative IC
a b
CG – cloud-to-ground lightning flash. IC – intracloud lightning flash.
Negative Positive Negative Positive Bipolar Positive Positive Bipolar Negative Positive Negative
be carried out to clarify this issue. 4.2. Distances and time intervals between the upward flash initiated from the tower and preceding lightning events We have also compared statistically the spatial distances and time intervals between the upward flash initiated from the tower and preceding lightning events (Table 3), for Rapid city, Gaisberg and Säntis. From this Table, it can be noted that the statistical parameters of the time intervals are very similar for all three towers. The distances between the tower and preceding events for the case of the Gaisberg Tower and Rapid City are very similar. However, lightning triggering events at the Säntis Tower occur at significantly closer distances to the tower. 4.3. Seasonal variations Seasonal variations are known to influence the occurrence of lightning discharges (Rakov and Uman, 2007). We have analyzed and compared the occurrence of self-initiated and other-triggered lightning strikes for the five locations listed in Table 1 during summer and winter seasons. The results are summarized in Fig. 2. The measurements at Uchinada were reported for the winter period only (Wang and Takagi, 2012). Between 2004 and 2010, there have been no direct observations of upward lightning from the towers in Rapid City during the winter months (Warner et al., 2012). The three towers in Europe show, despite their similar high proportion of self-initiated flashes, different seasonal characteristics. All self-initiated flashes at the Peissenberg Tower were recorded during the winter period. Winter self-initiated flashes at the Gaisberg Tower represent 58% of the total number of upward flashes. However, this value reduces to only 8% for the case of the Säntis Tower. The majority (91%) of upward flashes occur in summer time at the Säntis Tower, like in Rapid City. Nevertheless, the majority of the Säntis flashes are self-triggered, as opposed to Rapid City. As it has been described in Section 4, the Peissenberg, Gaisberg and Säntis Towers are situated on the top of mountains, Table 3 Comparison of the statistical parameters for the time and distance intervals between upward flashes and preceding events. Tower
Distance between the tower and Time from preceding event to preceding event, km upward leader, ms Min
Rapid City 3.5 Gaisberg 0.3 Säntis 0.05
Mean
Median
Max
Min
Mean
Median
Max
17.5 18.9 7.2
14.8 16.3 1.5
50.5 48.3 46.8
o1 o1 o1
72 102 109
46 33 49
404 819 600
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
49
70 Other−triggered Self−initiated 60
Number
50 40 30 20 Fig. 2. Seasonal proportions of self-initiated and other-triggered upward flashes at different sites.
10
rising to 940 m, 1287 m and 2502 m above sea level, respectively. We suppose that those seasonal variations could be attributed to the differences in the altitudes, but further research is needed to study this hypothesis.
0
4.4. Temperature variations Seasonal variations mentioned above imply that air temperature also affects the mechanism of upward flash initiation. Indeed, in Zhou et al. (2014), it was shown that, at the Gaisberg Tower, relatively low ambient temperatures (mean value 2.8 °C) facilitate the initiation of upward flashes without any nearby preceding lightning activity. On the other hand, relatively high ambient temperatures (mean value þ11.1 °C) seem to be correlated with the initiation of upward other-triggered flashes. The distribution of the air temperatures measured at the top of the Gaisberg Tower at the moments when upward flashes were initiated is shown in Fig. 3. For the Gaisberg Tower, the temperature measurements were taken every minute. Other parameters, like wind speed, relative humidity and pressure were found to have no direct influence on upward flash initiation. Meteorological parameters at the Säntis Tower are measured at the tower base and not at the top of the tower, as is the case for the Gaisberg Tower. Also, temperature measurements are reordered every ten minutes. An analysis similar to the reported Gaisberg study has been performed for the case of the Säntis Tower. The obtained temperature distribution is shown in Fig. 4. Since the majority of upward flashes of both types, self- and other-triggered, 35 Other−triggered Self−initiated
30
Number
25 20
−11 −9 −7 −5 −3 −1 1 3 5 7 Temperature, °C
9
11 13 15
Fig. 4. Distribution of temperatures associated with occurrence of self-initiated and other-triggered upward flashes at the Säntis Tower.
occur at the Säntis Tower during the summer, mean temperatures when these flashes are observed are found to be positive and very close: þ4.2 and þ 4.6 °C, respectively.
5. Practical considerations A number of practical issues can arise when applying the general procedure used in this study or similar methodologies used in previous studies for the analysis of initiation of upward flashes from tall structures. Some of these issues can be handled by applying common rules. They are briefly described in what follows. 5.1. Time and distance intervals When correlating the occurrence of upward lightning flashes on a tower with preceding lightning strokes or flashes from a lightning location system database, the time window and distance to the tower should be limited. A comparison of the time and distance intervals used in previous studies is presented in Table 4. It can be seen that the study in Rapid City used a circle of 200 km radius while, in the Gaisberg study, the circle’s radius was 10 times smaller. On the other hand, the time interval was smaller in the Gaisberg study. Clearly, by increasing the time and distance intervals, it is more likely to find a preceding lightning event. The resulting ratio of self-initiated to other-triggered upward flashes will therefore be affected by the choice of these two parameters. These intervals are strongly linked to the mechanism of upward flash initiation. Several processes can lead to the formation of an upward discharge Table 4 Comparison of the time and distance intervals used to find preceding events for the other-triggered upward flashes from the towers.
15 10
Location
Rapid City
Gaisberg
Study period
2004– 2010 200
2005–2009 2000–2012 2010– 2012 20 50 50
2
5
5 0
−13−11 −9 −7 −5 −3 −1 1 3 5 7 Temperature, °C
9 11 13 15
Fig. 3. Distribution of temperatures associated with the occurrence of self-initiated and other-triggered flashes at the Gaisberg Tower (Fig. 2 in Zhou et al. (2014)).
Radius of observation circle (km) Duration of observation interval (s)
Gaisberg
2
Säntis
2
50
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51
20%
15%
10%
5%
0%
Fig. 5. Percentage of other-triggered upward flashes for the Gaisberg Tower as a function of time interval and distance (circle radius).
from the tower. Among the most important are: slow electrostatic field variations and faster radiated field change. However, it is still not clear which one of them plays the dominant role. We have demonstrated this effect for the case of the Gaisberg Tower in Fig. 5, in which the variation of the percentage of othertriggered upward flashes is calculated based on the choice of different time intervals and circle radii. The choice of these intervals will have some effect on all the studies reported in Section 4 since it will change the ratio of selfto other-triggered flashes. Especially, it would be noticeable for the results reported in Table 3. By analyzing the rate of growth of the percentage of upward flashes from Fig. 5, it can be seen that this rate reduces significantly in the circles greater than 50 km and time intervals above 1 s. Therefore, these values can be used as reference values for future studies. Otherwise, parametric study should be performed to examine the influence of these parameters on the resulting percentage of other-triggered upward flashes.
from both towers are initiated without any preceding lightning activity. The observations at the European towers (Gaisberg, Peissenberg and Säntis) differ greatly from the Rapid City study in which most of the flashes were preceded by nearby lightning activity. The study revealed two major differences between the data obtained at the Gaisberg and the Säntis Towers concerning flashes with preceding lightning activity: (1) Upward flashes from the Gaisberg Tower were mostly triggered by positive lightning events, while flashes from the Säntis Tower were mostly preceded by negative flashes, (2) preceding events had the same polarity as triggered flashes in the case of the Säntis Tower and had opposite polarities in the case of the Gaisberg Tower. We showed that the mechanism of upward flash initiation in different sites features a strong seasonal variation, even within Europe. Therefore, the observed differences cannot only be explained to the temperature variations as it has been suggested before. Other parameters such as the altitude of the site and meteorological conditions above the tower should also be taken into account. We showed also that the results of the presented studies on the initiation of upward flashes from tall structures might be affected by the selected parameters of the study and the performance of the LLS to detect IC discharges. Although our results do not enable us to propose optimum values for the radius and the time interval, we recommend the use of a time interval of at least 1 s and an area around the tower of at least 50 km radius. Studies from Rapid City, Gaisberg and Säntis were found to have similar statistical parameters of the time intervals between the upward flashes and their preceding events. However, lightning triggering events at the Säntis Tower were found to occur at significantly closer distances to the tower, compared to Gaisberg and Rapid City. It is worth noting again that the causality implied by the word triggered has not been established at this time and calls for further investigations.
5.2. Sequence of preceding lightning events
Acknowledgments
Some rare upward flashes from the towers were preceded by a series of lightning events, where it was not evident which preceding stroke to consider as a triggering event.
Financial support from the BKW Ecology Fund (Grant number 110005F) is acknowledged.
5.3. Limited detection efficiency of LLS for IC discharges Lightning location systems (LLS) have limited capabilities to detect intracloud (IC) discharges, especially those with small peak fields, and the detection efficiency may depend on the main orientation of the IC discharge channel (mostly vertical similar to CG, or mostly horizontal).
6. Conclusions We presented an analysis of the initiation of upward lightning flashes for the Gaisberg and the Säntis Towers solely based on the data from the EUCLID lightning location system. We have updated a previous study for the Gaisberg Tower, with an increased observation period of 14 years. The proportion of self-initiated and other-triggered flashes for the Gaisberg Tower remained almost unchanged with respect to the previous study, which was carried out over an observation period of 5 years. We presented a similar study for the Säntis Tower, for which we obtained a behavior analogous to that of the Gaisberg Tower and the Peissenberg Tower, namely that the majority of upward lightning discharges
References Berger, K., Anderson, R.B., Kroninger, H., 1975. Parameters of lightning flashes. Electra 41, 23–37. Cummins, K.L., Murphy, M.J., 2009. An overview of lightning locating systems: history, techniques, and data uses, with an in-depth look at the U.S. NLDN. IEEE Trans. Electromagn. Compat. 51, 499–518. http://dx.doi.org/10.1109/ TEMC.2009.2023450. Diendorfer, G., Pichler, H., Mair, M., 2009. Some parameters of negative upwardinitiated lightning to the Gaisberg Tower (2000–2007). Electromagn. Compat. IEEE Trans. 51, 443–452. http://dx.doi.org/10.1109/TEMC.2009.2021616. Heidler, F., Manhardt, M., Stimper, K., 2014. Self-initiated and other-triggered positive upward lightning measured at the Peissenberg Tower, Germany. IEEE, 157–166. http://dx.doi.org/10.1109/ICLP.2014.6973113. Manhardt, M., Heidler, F., Stimper, K., 2012. The electric field of negative upward lightning strikes at the Peissenberg tower, Germany. In: Presented at the Lightning Protection (ICLP), 2012 International Conference on, IEEE, pp. 1–9. 10. 1109/ICLP.2012.6344205. McEachron, K.B., 1939. Lightning to the Empire State Building. J. Frankl. Inst. 227, 149–217. http://dx.doi.org/10.1016/S0016-0032(39)90397-2. Rakov, V.A., Uman, M.A., 2007. Lightning: Physics and Effects. Cambridge University Press, United Kingdom. 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. Rubinstein, M., Smorgonskiy, A., Rachidi, F., Zuber, J., 2015. On the classification of
A. Smorgonskiy et al. / Journal of Atmospheric and Solar-Terrestrial Physics 136 (2015) 46–51 tower flashes as self-initiated and other-triggered. In: Presented at the Asia Electromagnetics International Symposium ASIAEM, Jeju, Republic of Korea. Schulz, W., Cummins, K., Diendorfer, G., Dorninger, M., 2005. Cloud-to-ground lightning in Austria: a 10-year study using data from a lightning location system. J. Geophys. Res. 110, D09101. http://dx.doi.org/10.1029/2004JD005332. Smorgonskiy, A., Tajalli, A., Rachidi, F., Rubinstein, M., Diendorfer, G., Pichler, H., 2014. Analysis of lightning events preceding upward flashes from Gaisberg and Säntis Towers. In: Lightning Protection (ICLP), 2014 International Conference O, Presented at the Lightning Protection (ICLP), 2014 International Conference o, pp. 1382–1385. 10.1109/ICLP.2014.6973346. Wang, D., Takagi, N., 2012. Characteristics of winter lightning that occurred on a windmill and its lightning protection tower in Japan. IEEJ Trans. Power Energy 132, 568–572. http://dx.doi.org/10.1541/ieejpes.132.568.
51
Warner, T.A., Cummins, K.L., Orville, R.E., 2012. Upward lightning observations from towers in Rapid City, South Dakota and comparison with National Lightning Detection Network data, 2004–2010. J. Geophys. Res. 117, D19109. http://dx.doi. org/10.1029/2012JD018346. Zhou, H., Diendorfer, G., Thottappillil, R., Pichler, H., Mair, M., 2012. Measured current and close electric field changes associated with the initiation of upward lightning from a tall tower. J. Geophys. Res. 117, 9 9 pp., 201210.1029/ 2011JD017269. Zhou, H., Diendorfer, G., Thottappillil, R., Pichler, H., Mair, M., 2014. The influence of meteorological conditions on upward lightning initiation at the Gaisberg Tower. In: Lightning Protection (ICLP), 2014 International Conference O. Presented at the Lightning Protection (ICLP), 2014 International Conference o, pp. 1162–1165. 10.1109/ICLP.2014.6973303.