Lightning waveform plays an important role in lightning observation, location, and lightning disaster investigation. Based on a large amount of lightning waveform data provided by existing real-time very low frequency/low frequency (VLF/LF) lightning waveform acquisition equipment, an automatic and accurate lightning waveform classification method becomes extremely important. With the widespread application of deep learning in image and speech recognition, it becomes possible to use deep learning to classify lightning waveforms. In this study, 50,000 lightning waveform samples were collected. The data was divided into the following categories: positive cloud ground flash, negative cloud ground flash, cloud ground flash with ionosphere reflection signal, positive narrow bipolar event, negative narrow bipolar event, positive pre-breakdown process, negative pre-breakdown process, continuous multi-pulse cloud flash, bipolar pulse, skywave. A multi-layer one-dimensional convolutional neural network (1D-CNN) was designed to automatically extract VLF/LF lightning waveform features and distinguish lightning waveforms. The model achieved an overall accuracy of 99.11% in the lightning dataset and overall accuracy of 97.55% in a thunderstorm process. Considering its excellent performance, this model could be used in lightning sensors to assist in lightning monitoring and positioning.

The Asia-Pacific Lightning Location Network (APLLN) is a lightning location system consisting of a series of very low-frequency signal detection sites. Since 2018, 16 detection sites have been deployed with an average baseline longer than 1000 km. The detection site used a trigger sampling method to record the lightning signal with a duration of 2 ms and calculates the lightning arrival time based on digital filtering and the Hilbert envelope method. APLLN used a time difference location algorithm and improved Levenberg–Marquardt non-linear least squares iterative algorithm to calculate and optimize the lightning location results. The analysis results of a strong thunderstorm process show that the average detection efficiency of APLLN was 55.34% for intracloud (IC) strokes, 63.55% for cloud-to-ground (CG) strokes and 61.83% for all strokes (IC + CG). The average location error of APLLN for this thunderstorm is 5–10 km.

Hai‐Nan Lightning Detection Network (HNLDN) is composed of seven sites that accept very low frequency/low frequency (VLF/LF) signals. A total of seven sites are evenly distributed on Hai‐Nan Island. The baseline distance is between 50 and 150 km, which is about five times longer than in typical short‐baseline VLF/LF imaging networks. Each detection station uses zero‐phase digital filtering to avoid the phase deviation caused by filtering. HNLDN uses an improved peak search method to increase the number of detected of lightning waveforms, applies a One‐dimensional convolutional neural network (1D‐CNN) to identify and classify the waveforms, and to the increase the detection efficiency of intracloud lightning. This article analyzes three typical thunderstorm processes. From inside to outside the network, the detection efficiency of HNLDN relative to the advanced direction‐time lightning detection system (ADTD) is 239%, 181%, and 19%, respectively. It can be seen that the detection efficiency is negatively correlated with the distance from the thunderstorm to HNLDN. The average location error of HNLDN inside the network is 190 m.

Advances in technology have facilitated the development of lightning research and data processing. The electromagnetic pulse signals emitted by lightning (LEMP) can be collected by very low frequency (VLF)/low frequency (LF) instruments in real time. The storage and transmission of the obtained data is a crucial link, and a good compression method can improve the efficiency of this process. In this paper, a lightning convolutional stack autoencoder (LCSAE) model for compressing LEMP data was designed, which converts the data into low-dimensional feature vectors through the encoder part and reconstructs the waveform through the decoder part. Finally, we investigated the compression performance of the LCSAE model for LEMP waveform data under different compression ratios. The results show that the compression performance is positively correlated with the minimum feature of the neural network extraction model. When the compressed minimum feature is 64, the average coefficient of determination $R^2$ of the reconstructed waveform and the original waveform can reach 96.7%. It can effectively solve the problem regarding the compression of LEMP signals collected by the lightning sensor and improve the efficiency of remote data transmission.