Study of spacetime and electrical properties of artificially triggered and natural lightning discharge based on VHF and electric field observations

PolyU Library Call No.: [THS] LG51 .H577P BSE 2017 Shenxxiv, 195 pages :color illustrationsIn this study, lightning initiation and propagation processes observed in both a negative rocket-triggered lightning discharge and a negative natural lightning discharge in southern China in 1999 were analyzed and discussed. Observations included a broadband VHF lightning interferometer and a slow antenna, both of which were located 90 m south of the rocket launcher, and a video camera, which was located 1.3 km west of the launcher. First, by combining a picture from the camera with the VHF radiation source data from the interferometer, we obtained 3D evolution of lightning processes in the triggered lightning discharge. The results showed that, while a classically-triggered negative discharge usually starts with a positive leader initiated from the tip of the ascending triggering-wire and moves upward, no such initial upward positive leader (UPL) was observed for the triggered discharge. This was probably due to the low resolution and sensitivity of the measurements. Instead, there was a downward negative leader (preliminary downward leader, PDL) at the preliminary stage of the discharge being observed, followed by an M-component-wise process and two leader/return-stroke processes. The PDL was most likely to be a leader process along the channel trace, possibly built by the undetected UPL, as its speed, which ranged from 3.7×10⁶ m/s to 0.3×10⁶ m/s, was similar to that of a dart leader in the existing literature. The M-component-wise process consists of a slow negative-going change stage (Ma), followed by a fast negative-going change stage (Mb) and a slow positive-going change stage (Mc). The Ma stage was found to be intra-cloud negative breakdowns moving towards the overhead position of the PDL trace. The Mb stage would be considered an M-component (channel brightening) that starts with a K breakdown in cloud (Mb1) moving horizontally towards the overhead position of the previous PDL, followed by an event (Mb2) moving up from ground to cloud along the PDL trace. As Mb2 reached the cloud, more new K breakdowns (Mc) appeared in the cloud around the extremities of the pre-built channels by Ma and Mb. Each of the leaders before the return stroke with many VHF sources located could be divided into three stages; namely L1a, L1b and L1c and L2a, L2b and L2c. The elevations of VHF radiation sources showed a descending trend for L1a, a horizontally expanding trend for L1b and a sharply ascending trend for L1c, while the azimuths showed a slowly ascending trend for L1a and L1b, but a descending trend for L1c. The trends of L2a, L2b and L2c were similar. The leader preceding the first return stroke (L1) started inside the cloud and propagated downward (L1a) to the triggering-wire trace, but with a different channel from that of the PDL. As the leader touched the triggering wire trace (L1b), it appeared to propagate upward (L1c) along the same channel of the PDL. The upward portion of the L1 (L1c) could possibly be interpreted as a branching and reflection behavior of the L1 current when it attached to the triggering-wire and PDL traces, due to the differences in conductivity and potential between the PDL and triggering wire traces and the L1 channel. The speed of the downward portion of L1 (L1a) decreased from 2.32 ×10⁶ m/s to 0.32 ×10⁶ m/s during descent, while that of the upward portion of L1 (L1c) increased from 0.85 ×10⁶ m/s to 2.70 ×10⁶ m/s during ascent. The leader preceding the second return stroke (L2) behaved similarly to L1 but with higher speeds. Second, we proposed a new approach for estimating the line charge density along the channel of a downward negative leader, based on ground measurements of electric field change and leader spatial developments with a broadband interferometer system. In this approach, the two downward negative leaders in the triggered lightning flash were studied. Due to the limitations of the methods used, only the downward part of the first leader (L1a) and the second leader (L2a) were investigated in detail. The results showed that the line charge density of L1a showed a general trend to increase as it went down, with a few large pulses superposing on the general trend. These large pulses were most likely associated with the leader channel turns when comparing with the leader channel image. The line charge density of L2a was generally smaller than that of L1a but with a similar trend and pulses. The range of line charge density for both L1a and L2a was 0.01-1.00 mC/m. Third, we proposed a new approach for retrieving the pre-existing background electric field profile and, hence, the space charge distribution, along a lightning leader channel if the line charge density of the leader channel is given. This approach provides a new way to learn the electrical conditions of the initiation of lightning discharge in a thunderstorm. With this approach, and the line charge density data obtained in above study, the electric field profile and space charge distribution along the leader channels of L1a and L2a were obtained. The background electric field of L1a above 680 m high was almost linear and ranged from -13 kV/m to 13 kV/m. The electric field under 680 m high was second order correlation with height and ranged from -108 kV/m to 150 kV/m with some large bipolar pulses (1112 kV/m to 925 kV/m) superposing on the general trend. The results above are reasonable according to others' observations. The bipolar pulses of the background electric field meant that there were some positive charge groups when the leader propagated to the locations of the pulses.The space charge density of L1a increased as the leader developed. The general range of the space charge density of L1a under 680 m high was -3 nC/m³ to 5 nC/m³, with some unipolar pulses (-31 nC/m³ to 72 nC/m³). The space charge density of L1a above 680 m high was very small and ranged from -0.8 nC/m3 to 0.4 nC/m³. These pulses indicate that there were positive charge groups present. The allocations of the pulses of the space charge density were the same as for the bipolar pulses of the background electric field, the pulses of the negative line charge density, and for the big curves of the L1a channel. Therefore, the complexity of environmental electrical conditions can cause distortion of the leader channel. Similar conclusions can be drawn for the L2a channel, except in the smaller range of the space charge density. The range of the space charge density for L2a was from -0.4 nC/m³ to 0.45 nC/m³, with some large pulses ranging from -8 nC/m³ to 8 nC/m³. Due to the limitations of this approach, the ranges of the background electric field and space charge density may not be accurate, but the trends and the pulses further our understanding of the propagation of lightning discharge. Fourth, using a similar approach as for the triggered lightning discharge, we studied the spatial and electrical properties of natural lightning discharge. As we had no optical images of the natural lightning, its channel was rebuilt by the data of the VHF source radiation from the broadband interferometer with a presumed height of the lightning initiation position in the cloud. Based on the approximate channel that was rebuilt, it was found that this discharge was a downward stepped-leader initiated negative cloud-toground discharge with only one leader/return stroke process. The general speed of the leader when it propagated horizontally in the cloud was 0.4 ×10⁶ m/s to 1 ×10⁶ m/s, and when it propagated vertically was about 3.5 ×10⁴ m/s to 1.5 ×10⁵ m/s. The average speed was 7.0 ×10 ⁴ m/s. As it went downward, the leader branched significantly. Additionally, while the leader moved downward, the VHF sources (breakdown processes) within branches around the leader tip frequently showed backward movements (about 400 m) with a speed of about 2 ×10⁶ m/s, which is much faster than the average leader downward speed of 7.0 ×10⁴ m/s. Electrical properties of the leader above 3000 m were also obtained based on a rough leader channel without branches and backward movements. With the development of the leader channel, the leader brought negative charges downward along the leader channel and the changes of the electric field increased. The estimated line charge density showed a trend to increase as the leader propagated downward, ranging from -0.1 mC/m to -1.1 mC/m. The pulses of the leader charge density show that the leader transported plenty of negative charge when it propagated along the leader channel, which may have caused the curves of the leader channel. Lastly, the findings and their scientific merits are summarized and discussed. Through our study of artificially-triggered lightning discharge and natural lightning discharge, we know that while the cases analyzed here are special, they have some common features. The range and the trend of the electrical and spatial properties show the applicability of the approaches and models in my study. The changes of electric field observed on the ground, the space evolution of the lightning leader channel and the propagation speed of the leader channel were closely related to the spatial initial environment of the lightning discharge. Differences in environment can cause different propagated movement of lightning discharge, depending on the results. This means that a different distribution of space charge or a complex environmental electric field profile can lead to a complex leader channel and movement. A stronger background electric field indicates stronger VHF radiation sources, a faster speed of the leader channel, a more complex lightning channel and more charge in the leader channel. In future research, we need to improve the methods of observation, to obtain more detail of the lightning discharge. With high resolution observations and more sophisticated detail for the lightning discharge, additional spatial evolution characteristics and electrical properties of the discharge, such as the charge density of the branches of the leader channel, can be analyzed. The accuracy of the results also will be improved. Lightning discharge with more complicated initial electrical environments also can be studied in the future, using observation and simulation.Department of Building Services EngineeringPh.D., Department of Building Services Engineering, The Hong Kong Polytechnic University, 2017Doctorat