1. Introduction
In recent years, how an upward positive leader (UPL) initiates has become an important problem in lightning physics research. However, due to the randomness of natural lightning, it is difficult to make close observations. Until now, the initial process of UPL has not been fully understood. However, the discharge process of triggered lightning is very similar to that of natural lightning, and the timing and location of triggered lightning can be planned, making it easy to observe at close range. Therefore, triggered lightning has become an important method by which to study the initiation of a leader.
In the field of high voltage technology, a long air gap discharge can be used to simulate lightning [
1,
2,
3,
4,
5,
6]. Some conversion threshold parameters during the evolution process of leader are given, such as local electric field, charge per unit length and critical length, but the scale of a long air gap discharge is much smaller than that of a natural discharge process, leading to great differences in parameters between the two. At the same time, because the scale of a long air gap discharge is too small and the discharge duration is too short, it is difficult to obtain a clear observation of the initial stage. Therefore, detailed research on UPL can only rely on triggered lightning.
Classic triggered lightning is usually realized by a rocket carrying metal wires. When the negative ground-level quasi-static electric field is dominant, a negative triggered lightning will be generated, which starts from a UPL. UPL includes two stages: an initial stage and a self-sustaining development stage. The evolution of a UPL process is as follows. As the rocket rises, discontinuous corona discharges are generated at the rocket′s tip, forming some isolated current pulses at the bottom of the metal wire (called precursor current pulses (PCPs), each precursor current pulse called (PCP)). With the increase of rocket height, the discharges become frequent, with PCP appearing in groups (called PCP clusters). When the rocket rises to a certain height, a strong PCP cluster appears, indicating that the UPL has started self-sustaining development. After that, the current at the bottom of the metal wire no longer appears in the form of pulses, but instead changes to a continuous current form. The strong PCP cluster is referred to as the initial precursor current pulses (IPCPs), and the current in the self-sustaining development stage is called the initial continuous current (ICC). In this paper, each IPCPs and preceding PCPs constitute an initial process. It should be noted that although an IPCPs is a sign of the beginning of self-sustaining development, it is still considered part of the initial stage for analysis in this paper. At present, the research on UPLs of triggered lightning have mainly focused on the self-sustaining development stage. The characteristics of current and channel development have been obtained by observation of coaxial shunts, high speed video and radiation source positioning. For example, Jiang et al. [
7] found that the luminous intensity at the tip of leader steps was stronger than in the channel behind it in the early portion of the self-sustaining development stage. The average 2-D speed of the leader was 1.0 × 10
5 m/s, with partial speeds ranging from 2.0 × 10
4 to 1.8 × 10
5 m/s between 130 and 730 m above ground. Sun et al. [
8] found that the UPL was mapped immediately from the tip of the metal wire during the early portion of self-sustaining development stage, developing at a speed of about 10
4 m/s without branches by using a very high frequency (VHF) radiation source positioning system. Biagi et al. [
9] presented high-speed video images (recorded at 300 kfps) of a UPL developing stepwise from the top of a grounded triggering wire. The leader’s upward development speed increased with height, from 5.5 × 10
4 m/s between heights of 123 and 134 m, to 2.1 × 10
5 m/s at a height of 350 m.
It is relatively difficult to research the initial stage before self-sustaining development stage due to weak radiation and weak luminous intensity. Some limited studies have been based mainly on the data of directly measured currents. Horii [
10] first introduced PCPs, and found that PCPs appeared in the form of a single or group of pulses with an amplitude within 100 A. Lalande et al. [
11] observed the currents and electric fields of PCPs, and found that the time intervals between PCPs were roughly tens of milliseconds, and that the pulse intervals in PCP clusters were about 25 μs. Willet et al. [
12] found that PCPs appeared about every 10 ms, and that the pulse intervals within clusters were about 30 μs. The related research also reported that PCPs were generated by the discharge near the tip of the rocket [
9,
11,
12,
13,
14]. Zhang et al. [
15] analyzed the current and radiation of PCPs, and found that individual PCPs were produced by weak upward positive breakdowns over a meters-scale distance, followed by more energetic and faster downward negative breakdowns over a distance several tens of meters in scale. Zhang et al. [
16] found that the geometric mean values of current peaks, 10~90% rise times, durations, half peak widths, transferred charges and pulse intervals for PCPs were 28 A, 0.33 μs, 2.3 μs, 0.73 μs, 27 μC and 25 μs, respectively. Although the above studies have given some basic characteristics of PCPs, the development and evolution during the initial stage have rarely been reported. As a result, further research is needed. In this paper, we study the evolution characteristics of UPLs during the initial stage based on directly measured current data of triggered lightning in 2019.
2. Experiments and Data
Since 2006, the Chinese Academy of Meteorological Sciences lightning research group has carried out triggered lightning experiments in Conghua District, Guangzhou City. More information about this experiment can be found in [
17]. This paper analyzes the directly measured current data of 14 triggered lightning events in 2019, which were measured using a 1 mΩ coaxial shunt. As shown in
Figure 1, the lower part of the coaxial shunt was connected to ground and the upper part was connected to a lightning rod. The current waveform was digitized by an HBM digitizer and transmitted via optical fiber to an HBM transient recorder in the control room. Then, the converted signal by O-E (optical-electro) conversion was output to a DL850 oscilloscope for recording. In order to realize the complete recording of the large current signal and weak current signal at the same time, we set up two HBM digitizers with ranges of ±50 and ±2 kA, respectively. At that point, the minimum current resolution was about 5 A. In addition, the ground-level quasi-static electric field was detected by a CS110 electric meter that was installed 85 m away from the lightning rod. The positive electric field indicated negative charge overhead.
Some parameters are defined as follows. Pulse duration is the time interval when the amplitude of the front and back edges of a pulse reach the background noise level. Pulse transferred charge is the integration of pulse current and duration. The duration of the PCP cluster or IPCPs is the time interval from the first pulse to the last pulse. The duration of ICC process is defined as the time interval between the end of the preceding IPCPs and the end of the ICC. The duration of the current quiet period is the time interval from the end of the previous ICC to the first PCP in the following initial process. The duration of the initial process is defined as the time interval between the first PCP and the first pulse of the IPCPs during each initial process. The duration of the initial stage is the time interval from the first PCP of the first initial process to the first pulse in the IPCPs during the last initial process. For triggered lightning in a single initial process form, there is only one initial process, so the duration of the initial process is the same as the duration of the initial stage.
4. Conclusions and Discussion
Using the directly measured current data of 14 triggered lightning events in 2019, this paper systematically studies the evolution process of discharges during the initial stage of the UPL, in particular the process before self-sustaining development. A detailed conclusion follows.
(1) The initial stage of triggered lightning can be divided into two types: a single initial process form and a multiple initial process form, with percentages of 64.29% and 35.71%, respectively. Compared with the former, the latter usually lasts longer, and the corresponding lightning is often triggered under a lower ground-level quasi-static electric field. The critical value of the initial stage duration between the two types is in the range of 1314.44–1634.35 ms. In addition, the duration of the first initial process in a multiple initial process form is generally longer than the duration of the initial stage of the triggered lightning in a single initial process form.
(2) Regardless of whether the triggered lightning is in the form of a single initial process or multiple initial processes, the pulse durations and transferred charges of PCPs generally increase linearly with time during each initial process. However, the linear increases are not continuous during multiple initial processes. The pulse durations and transferred charges of PCPs at the beginning of the next initial process are less than those of the end of the previous initial process. This may be due to the occurrence of an ICC process during failed initial process, resulting in a reduction in the local electric field that is not conducive to the formation of the next new initial process.
(3) In each initial process, when the pulse duration and transferred charge of a PCP gradually increase to a certain extent, PCP clusters and IPCPs will appear. According to the research results, when a PCP cluster appears, the average value of the pulse durations and transferred charges of PCPs are 3.48 μs and 19.53 μC, respectively. When IPCPs appears, the corresponding average values are 4.69 μs and 27.23 μC, respectively.
(4) Compared with the closest PCP cluster, IPCPs contain more pulses, with a critical value range of 6–7. Furthermore, the duration of the IPCPs is also longer, with a critical value range of 138–198 μs. Whether the statistics include all 21 initial processes or each initial process individually, the average pulse durations and average pulse transferred charges of IPCPs are larger than those of their closest PCP cluster. On the other hand, IPCPs and PCP clusters have no significant difference in current pulse peak.
The initial process of triggered lightning is very similar to that of long sparks observed in laboratory. Gallimberti et al. [
3] reported the fundamental processes in long air gap discharges in detail. A positive discharge is initiated by the formation of the first corona. The corona consists of a number of streamer channels that develop from a common root. After one or more coronas, if the electric field is high enough, the leader channel inception itself takes place, starting from the common root of the coronas. The UPL of the triggered lightning should also be initiated and sustained by the formation in virgin air of ‘streamer corona’ and ‘leader’ discharges. It can be speculated that PCPs and IPCPs correspond to corona discharges and ‘leader’ discharges, respectively. Increases in the durations and transferred charges of PCPs indicate that the corresponding corona discharges intensifies. Gallimberti et al. [
3] also mentioned that the leader propagates almost continuously, unless the rate of increase of the voltage is too low, in which case sudden re-illuminations associated with channel elongation appear during the propagation (‘restrikes’). In our results, the multi-initial process was similar to that case, as the condition with a slow increase of the local electric field led to an occurrence of multiple initial processes.
The evolution characteristics given in this paper can provide a reference for the development of a leader model in which the charge per unit length is necessary to realize the thermal transition from the diffuse glow to the leader channel [
4]. In this paper, although we have given the transferred charge of PCPs when IPCPs appear, there is still a lack of the development scale. Therefore, we will obtain the development scale through analysis of VHF positioning results in the next step, so as to obtain the key model parameters.