Based on fast electric field waveforms of the Low‐frequency E‐field Detection Array (LFEDA), we introduce the time reversal technique into lightning three‐dimensional location for the first time and propose a new algorithm for the three‐dimensional location of lightning low‐frequency discharges. Without using complex filtering algorithms to remove higher‐frequency component, this method obtains similar results to the newly reported LFEDA refinement algorithm. The new algorithm can obtain finer, more continuous, and richer positioning results with a minimum of four stations, 5‐dB signal‐to‐noise ratio, and 500‐ns time error compared with the low‐frequency signal time of arrival three‐dimensional positioning method. These results indicate that the new algorithm has the advantages of low requirements on the number of stations, certain anti‐interference ability, and low requirements on time accuracy. The standard deviations in the X and Y directions for return strokes of triggered lightning flashes are both approximately 90 m.

Lightning location provides an important means for the study of lightning discharge process and thunderstorms activity. The fine positioning capability of total lightning based on low-frequency signals has been improved in many aspects, but most of them are based on post waveform processing, and the positioning speed is slow. In this study, artificial intelligence technology is introduced for the first time to lightning positioning, based on low-frequency electric-field detection array (LFEDA). A new method based on deep-learning encoding features matching is also proposed, which provides a means for fast and fine location of total lightning. Compared to other LFEDA positioning methods, the new method greatly improves the matching efficiency, up to more than 50%, thereby considerably improving the positioning speed. Moreover, the new algorithm has greater fine-positioning and anti-interference abilities, and maintains high-quality positioning under low signal-to-noise ratio conditions. The positioning efficiency for return strokes of triggered lightning was 99.17%, and the standard deviation of the positioning accuracy in the X and Y directions was approximately 70 m.

The initiation of a leader is an important lightning discharge process, but how an upward positive leader (UPL) initiates is still not fully understood. The evolution characteristics of a UPL during its initial stage was systematically studied based on directly measured current data of 14 triggered lightning events in 2019. It was found that 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. In each initial process, precursor current pulses (PCPs), PCP clusters and initial precursor current pulse (IPCPs) are typical current waveforms, and the pulse durations and transferred charges of PCPs increase linearly with time. However, in the multiple initial process form, the pulse durations and transferred charges of PCPs will reduce significantly after each previous initial process and then continue to increase in the following initial process. In each initial process, when the pulse duration and transferred charge of a PCP increase to a certain extent, PCP clusters and IPCPs begin to appear. For the emergence of PCP clusters, the average values of the threshold are 3.48 μs and 19.53 μC, respectively. For the occurrence of IPCPs, the corresponding values are 4.69 μs and 27.23 μC, respectively. The average values of pulse durations and transferred charges of IPCPs are larger than those of PCP clusters. Compared with adjacent PCP clusters, IPCPs contain more pulses, with a critical range of 6–7. IPCPs also last longer, and have a critical range of 138–198 μs.

The discharge signal in the initial stage of lightning is weak. The revelation of the discharge mechanism at this stage depends especially on close observation. In this study, a continuous interferometer (CINTF) was used to observe the initial stage of the upward positive leader (UPL) of the triggered lightning in Conghua District, Guangzhou City, Guangdong Province. The positioning error of CINTF for a close-range radiation source was analyzed, and the positioning error calibration method of CINTF for a specific close-range radiation source was studied, which improved the observation accuracy of elevation angle at the initial stage of the UPL of the triggered lightning. With the rise of the rocket, the positioning error in altitude during the initial stage of the UPL increased obviously. Under the layout condition of the Guangzhou field experiment site for lightning research, when the positioning results of the elevation angle of the initial stage of the UPL were 40°, 50°, and 60°, respectively, the calibrated altitude positioning error could be reduced by about 11 m, 14 m, and 20 m, respectively. On the basis of the calibrated observation results, the evolution characteristics of the initial stage of the UPL were studied, and its discharge mechanism was revealed. The precursor current pulse (PCP) was generated by a weak upward positive breakdown and a subsequent strong downward negative breakdown near the rising rocket tip, which was in the form of a single pulse. The precursor current pulse cluster (PCP cluster) and initial precursor current pulse cluster (IPCP) were both signs of self-sustaining development of the UPL. After the PCP cluster, self-sustaining development stopped immediately. The self-sustaining development after IPCP could be short-term or continuous.

No abstract