应用气象学报

2020, 31(2)

[FullText](4818) [PDF 643KB](62)

Abstract:

Advances of Observation and Study on Tall-object Lightning in Guangzhou over the Last Decade

Lü Weitao, Chen Lüwen, Ma Ying, Qi Qi, Wu Bin, Jiang Ruijiao

2020, 31(2): 129-145. DOI: 10.11898/1001-7313.20200201

[FullText HTML] (1722) [PDF 1857KB](140)

Abstract:
Comparing to the ground or low-object, tall-object is more likely to reach the threshold for the initiation of leader on its top due to the distortion and enhancement of electric field, therefore tall-object is not only easier to get struck by downward lightning, but also can initiate upward lightning. A field experiment, mainly focusing on the observation of lightning flashes terminating on tall structures, has been conducted since 2009 at the Tall-Object Lightning Observatory in Guangzhou (TOLOG), which is the important part of the Field Experiment Base on Lightning Sciences, China Meteorological Administration (CMA_FEBLS). Hundreds of tall-object lightning flashes have been captured during 2009-2018.For the lightning attachment process, tall-object will play the role of "magnifier":The TOLOG high-speed optical observation discovers the connection of the downward negative leader to the lateral surface of the upward connecting leader for the first time, and shows two basic types of the leader connection behaviors during the attachment process in negative cloud-to-ground lightning; the fine structure of negative stepped leader in natural lightning at close distance is revealed by using high-speed video records; the 2D/3D propagation characteristics of downward and upward leaders are analyzed; and the striking distances of lightning flashes to tall-object with different heights are also estimated.Tall-object plays an "amplifier" role on lightning electromagnetic field:Statistical analysis of the TOLOG data show that the magnetic field peak values induced by the first return stroke of lightning flashes to objects higher than 200 m is 2.4 times of that of lightning flashes to objects lower than 200 m; the higher the tall-object is, the larger the lightning location system inferred peak current of strokes are recorded in the vicinity of the tall-object; and the numberical simulation of the tall-object on electromagnetic field of lightning return stroke also show that the height of tall-object has significant enhancing effects.Tall-object is the "hot spot" of downward and upward lightning:Attraction effects of tall-object on downward lightning can protect other objects near tall-object from lightning strikes; the upward negative lightning from the tall-object can be triggered by the return stroke, the continuing current or the discharging process in cloud of positive cloud-to-ground lightning; and in upward lightning, abrupt extension is found at the positive end of the recoil leader which propagates bidirectionally.Using data of the TOLOG, the detection efficiency, the location error and the systematic bias of lightning location systems in Guangdong are evaluated, showing that the observation area of the TOLOG can be used as a "calibration field" for ground-based or space-based lightning monitoring system.

High-speed Video Observations on Abrupt Elongations of the Positive End of Bidirectional Leader

Wu Bin, Lü Weitao, Qi Qi, Ma Ying, Chen Lüwen, Jiang Ruijiao

2020, 31(2): 146-155. DOI: 10.11898/1001-7313.20200202

[FullText HTML] (1626) [PDF 1750KB](53)

Abstract:
One of the most important advances in lightning physics research in recent decades is the introduction of the the bidirectional developmengt of leaders and its observation and verification, but this theory is not paid much attention until more and more observation results of natural lightning, long gap spark discharge in laboratory and artificially triggered lightning have proved the correctness of the concept.The theory is used to establish the model of leader to interpretate the physical mechanism of lightning initiation and development. High-speed video camera observation not only provides direct evidence for the development of bidirectional leader, but also gives details of positive and negative ends. Previous observations show that the negative leader developed in a step-wise manner by relying on the space stem/leader and the corona streamer at the front of the new leader's tip. Positive leaders propagate in a continuous or step-wise manner, but to date, characteristics and mechanisms of the development of positive leaders are still unclear. In recent years, the abrupt elongations of the positive leader (or the positive end of bidirectional leader) are found in the high-speed video observation of the leader. This phenomenon might be closely related to the propagation mechanism of the positive leader and deserves further observation and analysis.Based on synchronization data of high-speed video camera and electric field change of an upward flash at the Canton Tower, the abrupt elongations phenomenon of the positive end before and after the return stroke is analyzed in detail.Results show that the positive end of the second dart leader intermittently extends into the air. There are three abrupt elongations of the second dart leader of the positive end, and the second abrupt elongation is caused by the connection between the positive end and the floating channel in which the tip of the positive end appears. After the second subsequent return stroke, there are two abrupt elongations of the channel tip. The two-dimensional average speed of the three abrupt elongations of the positive end is approximately 2.3×10⁶ m·s^-1, and the average length of the three abrupt elongations is approximately 115 m. After the return stroke, the two-dimensional average speed of the two abrupt elongations of the channel tip is approximately 4.3×10⁶ m·s^-1, and the average length of the two abrupt elongations is approximately 212 m.

Two-dimensional Optical Observation of Striking Distance of Lightning Flashes to Two Buildings in Guangzhou

Qi Qi, Lü Weitao, Wu Bin, Ma Ying, Chen Lüwen, Jiang Ruijiao

2020, 31(2): 156-164. DOI: 10.11898/1001-7313.20200203

[FullText HTML] (1358) [PDF 967KB](42)

Abstract:
Lightning can strike directly on buildings, lightning protection devices or the lateral surface of buildings, endangering buildings on the ground. Effective lightning protection measures can avoid lightning damage to buildings and prevent possible fire, explosion or other hazards. Striking distance is an important reference index in lightning protection design of buildings, which is widely used in various common lightning protection design methods, such as rolling ball method, collecting volume method, etc. With the development of social economy, there are more and more tall-object in modern cities. It becomes more challenging to accurately estimate the striking distance of buildings with different heights and to formulate more effective lightning protection schemes. Up to now, lots of researches on the attachment process of natural lightning are conducted, especially by means of optical observation, which mainly benefits from the intuition of optical data. Although a large number of observations have been made on the lightning attachment process, reports on the striking distance are still rare.Based on optical data of 21 lightning discharges on two steeple buildings, the Canton Tower (600 m, 12 cases) and the Guangsheng International Building (360 m, 9 cases) from 2012 to 2018, and data of return stroke peak current provided by Guangdong Power Grid Lightning Location System, influences of building height and return stroke peak current intensity on the striking distance are analyzed. Results show that the striking distance on higher buildings is longer, and the median lightning strike distance of the Canton Tower is about 2 times of that of the Guangsheng International Building. For buildings with a certain height, the striking distance tends to increase with the peak current increasing. Moreover, the higher the building is, the stronger the peak current of the corresponding return stroke is. The peak current of return stroke on the Canton Tower is obviously stronger than (about 1.7 times) that on the Guangsheng International Building. In the attachment process, the two-diensional average speed ratio of the downward leader and the upward leader is less than 4 at 0.1 ms before the return stroke. The number of cases with a ratio of 0 to 1 is the largest, accounting for about 65% of the total number of cases.

Detection Results of Guangdong-Hongkong-Macao Lightning Location System for Tall-object Lightning

Chen Lüwen, Lü Weitao, Zhang Yijun, Ma Ying, Qi Qi, Wu Bin

2020, 31(2): 165-174. DOI: 10.11898/1001-7313.20200204

[FullText HTML] (1926) [PDF 1010KB](853)

Abstract:
Tall-object lightning is one hot topic in the field of atmospheric electricity. In order to explore the application of Guangdong-Hong Kong-Macao Lightning Location System (GHMLLS) data on tall-object lightning, the performance of GHMLLS is evaluated based on the observation obtained at the Tall-Object Lightning Observatory in Guangzhou (TOLOG) during 2016 to 2017, and a preliminary comparative analysis on characteristics of lightning activities in the tall-object area in Guangzhou are made. Results show that the flash detection and the stroke detection of GHMLLS is about 93% (214/229) and 93% (449/481). The arithmetic mean (median) value of the location error is about 361 m(188 m), 252 m(167 m) and 294 m(173 m) for downward first negative strokes, downward subsequent negative strokes and upward negative strokes respectively. It is found that the cloud-to-ground (CG) lightning with higher grounding points are more likely to be misidentified as intro-cloud (IC) lightning by the GHMLLS. When the grounding point is (is not) lower than 200 m, the identification accuracy of the IC/CG classification is 99% (80%) and 93% (35%), respectively, for downward first negative strokes and the subsequent negative strokes. The accuracy of the IC/CG classification for upward negative strokes is found to be only 17%. According to statistical results of GHMLLS dataset from 2014 to 2018, negative IC records in tall-object area in Guangzhou are found to be more prominently gathered near buildings with height over 300 m than CG records, indicating that most negative IC records are actually tall-object strokes but just misjudged by GHMLLS. Mean shift algorithm is used to identify the lightning density centers of three relatively isolated tall buildings, namely the Canton Tower (600 m), the CITIC Plaza (390 m) and the GF Securities Headquarters (308 m). It's found that the number of strokes within the radius of 200 m in the lightning density center of the Canton Tower is about 5 times as much as that of the CITIC Plaza or the GF Securities Headquarters. It is speculated that most lightning occurred on the Canton Tower belonging to upward type, while those hitting the CITIC Plaza and the GF Securities Headquarters are mainly downward type.

Numerical Simulation of Main Negative Charge Area Parameters for Upward Negative Cloud-to-ground Lightning

Wang Yiru, Tan Yongbo, Zheng Tianxue, Yu Junhao, Li Chunsun, Liu Minzhi

2020, 31(2): 175-184. DOI: 10.11898/1001-7313.20200205

[FullText HTML] (1580) [PDF 2538KB](44)

Abstract:
Different types of lightning initiation condition are the focus of scientific research. The initiation of cloud lightning is competitive with self-sustaining uplink lightning. Based on the classical dipolar charge structure and the existing lightning discharge parameterization schemes, under the background of the classic dipolar charge structure, two-dimensional(2-D) high-resolution lightning discharge simulation experiments are carried out by adjusting the parameters of the main negative charge regions. Cloud charge structures which are beneficial to the onset of self-sustaining upward lightning are discussed. Results show that the part of the lightning at the beginning of the building has almost no bifurcation, and spreads straight toward the vertical direction. As the lightning continues developing, it gradually extends horizontally, and lightning channel branches gradually increase. Branches generally develop from bottom to top. The development of new branches in the front section of the main channel will hinder the development of lightning branches located to a certain extent. Branches at the same height may have a simultaneous developing trend. It can be inferred that this is related to the flat surface potential direction and the potential well in the charge region of the cloud. It is found that the higher the charge area is, the larger the charge density and distribution range the initial-self-sustaining uplink lightning needs. There speculates a threshold of the height of the main negative charge regions for the onset of upward lightning. When the negative charge region is higher than that, with the accumulation of charge in the main negative regions, the lightning will start in clouds instead of upward lightning. Only when the main negative area is lower than the threshold, the steady accumulation of electric charge will cause the flash originating. This may charge far away from the earth, and the cloud charge accumulation effects are greater than the increasing process of the electric field in the cloud for building sophisticated growth effects of electric field. As the main negative charge accumulates, the lightning starting conditions are firstly met in the cloud, initiating flashes in the clouds.

Lightning Activities in a Convection Cell Dominated by Heavy Warm Cloud Precipitation

Liu Ze, Guo Fengxia, Zheng Dong, Zhang Yang, Wu Chong, Yao Wen

2020, 31(2): 185-196. DOI: 10.11898/1001-7313.20200206

[FullText HTML] (1503) [PDF 8849KB](154)

Abstract:
Lightning activity in a convection cell that occurred in Guangzhou of China on 7 May 2017 dominated by heavy warm cloud precipitation and its relationship with the precipitation structure of the cell are disscussed, using three-dimensional lightning location data of the Low_Frequency E-field Dection Array (LFEDA) in the Field Experiment Base on Lightning Sciences, China Meteorological Administration (CMA_FEBLS) and Guangzhou polarimetric radar observations. According to the ground precipitation obtained by radar inversion, the maximum cumulative precipitation from 0000 BT to 0400 BT in the cell dominated by warm cloud precipitation is 261 mm. The cell produces a total of 1250 detected lightning flashes within 4 h, with the ratio of cloud-to-ground flashes being about 24%. Lightning discharges mainly occur in the height range of 4-12 km, corresponding to the isotherm layers between approximately 0℃ and -40℃. The height and isotherm associated with the peak-frequency lightning discharges are about 8.5 km and -19℃, respectively. The heavy rainfall cell represents general tripolar charge structure, i.e., the upper positive charge region, middle negative charge region and lower positive charge region, with the negative charge core being between approximately -8℃ and -15℃ layers. The region featuring lightning discharges and dominated by dry snow account for about 82% of all, while the ratio for the region featuring lightning discharges and dominated by graupel account for about 11%. Most graupel-dominating regions associate with lightning discharges are located between 4 km and 8 km layers. This may be related to the weak convection in the cell dominated by warm cloud precipitation. Total lightning rate show relatively significant correlations with the 30 dBZ radar echo top height and volumes of the regions where radar echoes are greater than 20 dBZ and heights are larger than -20℃ level. The average height of lightning discharges is well related with the 20 dBZ radar echo top height and volumes of regions where radar echoes are greater than 30 dBZ and heights are larger than -20℃ level. Relative prominent corresponding relationship is also found between total flash frequency and maximum precipitation intensity. Meanwhile, the rainfall per flash is in the order of 10⁷ kg/fl.

Observation of the Whole Discharge Process During a Multi-stroke Triggered Lightning by Continuous Interferometer

Zhang Yang, Chen Zefang, Wang Jingxuan, Fan Yanfeng, Zheng Dong, Lü Weitao, Zhang Yijun

2020, 31(2): 197-212. DOI: 10.11898/1001-7313.20200207

[FullText HTML] (1202) [PDF 10471KB](61)

Abstract:
The discharge process is often discussed in the lightning physics. VHF (very high frequency) observations play an important role in studying lightning discharge processes, because they show where air breakdown occurs. Triggered lightning flashes are ideal research objects, because they are generated in a fixed time and place. Regular experiments of triggering lightning have been conducted by Chinese Academy of Meteorological Sciences since 2006, and 189 lightning flashes are triggered successfully. Besides regular observations, such as fast/slow antenna, channel-based current and high-speed video, the continuous interferometer (CINTF) is developed and deployed at the site for triggered lightning after 2016 in order to observe detailed discharge processes. The CINTF can provide high-precision location with time resolution of around 1 μs.A multi-stroke triggered lightning with 8 return strokes is observed by a self-developed lightning continuous interferometer during Guangdong Comprehensive Observation Experiment on Lightning Discharge on 11 June 2019. The whole discharge processes, including precursor current pulse, initial precursor current pulse, initial continuous current and return strokes, last for about 1600 ms. The channel-based current, electric field change and radiation source positions are obtained. CINTF results show that the radiation signal generated from discrete precursor current pulse with a minimum value of about 8 A can be detected. The average transfer charge of initial precursor current pulse is twice as much as that of the precursor current pulse. During the initial continuous current stage after continuously developing of the upward positive leader, the obviously forward channel extension is the main discharge form at the beginning, and then the extension can be accompanied by discontinuously backward propagation, and recoil leader often occurs in the following stage. M-component in the ICC stage is caused by two discharge processes:Forward positive streamer (or leader) or recoil leader generated in the existing channel. Multi recoil leaders occur during the whole M process, which maintain M discharge. The main discharge form in return stroke stage is recoiling discharge. There are multi recoil leaders before return strokes, however, most of them cannot develop to ground, keep as an attempted leader until the last leader-stroke occurs. Adjacent two strokes with a short duration time of 4.5 ms have the same optical channel in optical images, but they are obviously different in development paths, which lead to a very short RS interval.

Characteristics of Medium-low Frequency Magnetic Fields of Initial Continuous Current in Rocket-triggered Lightning

Fan Yanfeng, Lu Gaopeng, Zhang Yang, Lü Weitao, Zheng Dong, Fan Xiangpeng

2020, 31(2): 213-223. DOI: 10.11898/1001-7313.20200208

[FullText HTML] (2319) [PDF 2736KB](1534)

Abstract:
Rocket-triggered lightning experiment conducted in Field Experiment Base on Lightning Sciences, China Meteorological Administration (CMA_FEBLS) provides a good opportunity to study the discharge process and its related electromagnetic effects. During the experiment, two medium-low frequency magnetic antennas are deployed at different distances from the rocket launch site, named close antenna (about 80 m) and far antenna (about 1.9 km), respectively, and magnetic fields are observed with high sensitivity by two antennas. Combined with the synchronous channel-base current and fast electric fields, electromagnetic characteristics of initial continuous current are analyzed. Benefitting from the expansion of the bandwidth of the antenna, magnetic pulse signals can be observed throughout the triggered lightning, including the initial magnetic pulse (IMP), magnetic pulse of the signal quiet period, the magnetic pulse burst (MPB) and the regular magnetic pulse (RMP). IMPs can be divided into two categories (i.e., impulsive and ripple pulses) according to the discernibility of separation between individual pulses. Impulsive pulses are well simulated by the transmission-line model, which suggests that these pulses are generated by leader current pulses propagating downward along the steel wire. Magnetic pulses of the signal quiet period are observed for the first time. The mean pulse width and inter-pulse interval of these pulses are about 1 μs and 14 μs, respectively, which indicates that the propagation of upward leaders during the stage is in the form of small-scale breakdown. The MPB can be observed by both close and far antennas, and the mean inter-pulse interval of the MPB(24.5 μs) is larger than that of the signal quiet period pulse. Furthermore, the channel-base current during the stage of MPB increases to dozens of hundreds of amperes, so it can be concluded that the electric field condition is conductive to the development of the upward leaders. In addition, the magnetic signal recorded at close distance indicates the physical process leading to initial continuous current pulse (ICCP), M-component as well as the direct measurement of current enhancement at the channel base, due to the charge transfer in the ICCP or M-component. Magnetic antennas can also record the regular magnetic pulses (RMPs) that are attributed to the interception of recoil leader with existing lightning channel. Inter-pulse intervals of RMPs are one order smaller than that of MPBs and IMPs, and observations may reflect differences between the positive polarity breakdown and the negative polarity breakdown of leaders.

Relationship Between Current Characteristics of Rocket-triggered Lightning During Different Discharge Stages

Wang Jingxuan, Zhang Yang, Chen Zefang, Sun Xiubin, Fan Yanfeng, Zheng Dong, Fan Xiangpeng, Lü Weitao, Zhang Yijun

2020, 31(2): 224-235. DOI: 10.11898/1001-7313.20200209

[FullText HTML] (1333) [PDF 1703KB](54)

Abstract:
The research on rocket-triggered lightning has made great progress since 1967. Various parts of rocket-triggered lightning are studied, but differences between initial stages of different types of rocket-triggered lightning and the relevance on each part of rocket-triggered lightning is still not well understood. Based on the rocket-triggered lightning current data obtained from Guangdong Comprehensive Observation Experiment on Lightning Discharge in Field Experiment Base on Lightning Sciences, China Meteorological Administration, two types of rocket-triggered lightning are classified considering the existence of return storke(RS)and the length of the initial continuous current(ICC). Differences of rocket-triggered lightning in different categories and the correlation between various parts are analyzed. Firstly, the difference between the rocket-triggered lightning with return strokes and the rocket-triggered lightning without return strokes is obvious. The precursor current pulse(PCP)and initial precursor current pulse(IPCP)average peak current, IPCP transfer charge, ICC average current and total transfer charge and duration of lightning with RSs are greater than those of lightning without RSs. Secondly, longer ICC duration corresponds to higher average peak current, first return charge and first RS peak current, PCP and IPCP peak current, PCP and IPCP average transfer charge of triggered lightning. It is also found that the average peak value of PCP and IPCP has the strongest correlation with ICC duration. The larger the average peak value of PCP and IPCP is, the longer the ICC duration is. It's also found that during the ascent of the rocket, the transfer charge of PCP tends to increase as the altitude of the rocket increases. There are some PCP pulse clusters that cannot produce a continuous upward leader (UPL), and it fails to produce an ICC. The average transfer charge of the PCP cluster is smaller than the average transfer charge of the initial precursor current pulse cluster (IPCP). According to the analysis of data, one of the conditions for the PCP cluster that fails to generate ICC to become IPCP is that the average transfer charge is greater than 25.91 μC.

Ground Potential Rise and Surge Protective Device Damage Caused by Initial Long Continuous Current Process in Triggered Lightning

Chen Shaodong, Zhang Yijun, Yan Xu, Du Sai, Lü Weitao, Zhang Yang

2020, 31(2): 236-246. DOI: 10.11898/1001-7313.20200210

[FullText HTML] (1476) [PDF 1284KB](55)

Abstract:
In the field of grounding on electronic and electrical systems, damage effects of ground potential rise on electronic equipments are of great importance. Based on triggered lightning technology, an observation experiment is carried out on impacts of the ground potential rise in grounding grid upon surge protective device (SPD), especially on damage effects of initial long continuous current process of triggered lightning on SPD. It's found that upon injection of triggered lighting into grounding grid, SPD damage of rated flow rate is likely to occur under combined effects of initial long continuous current process and subsequent return stroke of triggered lightning. When the energy flowing through SPD is accumulated to a given extent, initial long continuous current process alone can also cause SPD damage. The impact on SPD is closely related to different waveforms of initial long continuous current process. When initial long continuous current process superposes ICCP with current of faster rise time and higher amplitude, energy flowing through SPD will increase rapidly, which is the most critical factor causing SPD damage in long continuous current process. The analysis of two cases indicates that, when the duration of initial long continuous current process and average current reaches about 100 ms and 200 A, the magnitude of discharge is 25 C, and energy flowing through SPD is up to about 1000 J, it is apt to cause 20 kA nominal discharge current and even higher SPD damage. Two processes (T0702 and T0726) of SPD are damaged by initial long continuous current of triggered lightning, when the peak value of current flowing through ground wire is 396.5 A and 392.7 A, respectively, the average current of main stage before damage is 23.6 A and 19.7 A, accounting for 10.8% and 6.7% of the average value of trigger lightning current, and the duration of current flowing through SPD above 50 A is 9.2 ms and 6.6 ms, respectively. When SPD is damaged, there is a sudden change in the residual voltage at both ends of SPD, which is obviously different from the disappearance of normal SPD residual voltage.

Ground Potential Rise and Transient Response of the Grounding Grid Based on the Triggered Lightning

Yan Xu, Zhang Yijun, Du Sai, Chen Shaodong, Lü Weitao

2020, 31(2): 247-256. DOI: 10.11898/1001-7313.20200211

[FullText HTML] (1469) [PDF 1480KB](52)

Abstract:
It is extremely dangerous of ground grid potentials which significantly rise when the lightning currents are flowing through the grid. A statistical analysis on 39 return-strokes, 10 M-components and the ground potential rise (GPR) caused by them based on 7 triggered lightnings is carried out. According to the analysis, the geometric mean (GM) of the current peak values of 39 return-strokes is -12.78 kA, and the corresponding GM of GPR reaches -138.97 kV; the GM of the current peak values of the M-components is -0.60 kA, while the corresponding GM of GPR is -7.18 kV. There are distinct sub-peaks in the waveform of the GPR caused by the return-strokes, and the GM of the sub-peaks falls to -90.09 kV within several microseconds, about 64.86% of peak values. During the return stroke stage, the linear correlation coefficient of GPR voltages and the direct lightning current is 0.94, and the linear correlation coefficient of GPR voltages and the gradient is 0.55. It indicates that the GPR in return stroke stage is mainly caused by lightning current discharge in soil and the inductive coupling is relatively weaker. During the M-component stage, the correlation coefficient of peak value of GPR voltages and direct lightning current reaches 0.99, which means the GPR during M-component stage is mostly caused by lightning current discharge in soil. The impulse grounding resistance in the stage of return stroke when lightning current dispersing through grounding grids is 10.87 Ω, and it is 12.02 Ω in the stage of M-component. Both of the impulse grounding resistances are smaller than the DC grounding resistance, and the difference reaches 1.1 times. The minimum half-peak width of the GPR caused by the return-strokes is 0.44 μs, of which the GM is 1.93 μs, only 25.8% of the half-peak width of the corresponding current return-stroke. And the half-peak width of the GPR caused by M-components can be up to 2 microseconds, about 124 times of the GM of the half-peak width of the return-strokes, keeping the surge protective devices (SPD) running long which easily leads to crashing damages.

Vol.31, NO.2, 2020