William Rison

Optical emissions associated with Terrestrial Gamma ray Flashes (TGFs) have recently become important subjects in space‐based and ground‐based observations as they can help us understand how TGFs are produced during thunderstorms. In this paper, we present the first time‐resolved leader spectra of the optical component associated with a downward TGF. The TGF was observed by the Telescope Array Surface Detector (TASD) simultaneously with other lightning detectors, including a Lightning Mapping Array (LMA), an INTerFerometer (INTF), a Fast Antenna (FA), and a spectroscopic system. The spectroscopic system recorded leader spectra at 29,900 frames per second (33.44 s time resolution), covering a spectral range from 400 to 900 nm, with 2.1 nm per pixel. The recordings of the leader spectra began 11.7 ms before the kA return stroke and at a height of 2.37 km above the ground. These spectra reveal that optical emissions of singly ionized nitrogen and oxygen occur between 167 s before and 267 s after the TGF detection, while optical emissions of neutrals (H I, 656 nm; N I, 744 nm, and O I, 777 nm) occur right at the moment of the detection. The time‐dependent spectra reveal differences in the optical emissions of lightning leaders with and without downward TGFs.

On 11 September 2021, two small thunderstorms developed over the Telescope Array Surface Detector (TASD) that produced an unprecedented number of six downward terrestrial gamma ray flashes (TGFs) within one‐hour timeframe. The TGFs occurred during the initial stage of negative cloud‐to‐ground flashes whose return strokes had increasingly large peak currents up to 223 kA, 147 GeV energy deposit in up to 25 1.2 km‐spaced surface detectors, and intermittent bursts of gamma‐rays with total durations up to 717 s. The analyses are based on observations recorded by the TASD network, complemented by data from a 3D lightning mapping array, broadband VHF interferometer, fast electric field change sensor, high‐speed video camera, and the National Lightning Detection Network. The TGFs of the final two flashes had gamma fluences of and 8, logarithmically bridging the gap between previous TASD and satellite‐based detections. The observations further emphasize the similarity between upward and downward TGF varieties, suggesting a common mechanism for their production.

Interferometric measurements of VHF radio-frequency signals produced by lightning are one of the most effective techniques for studying lightning breakdown processes, so uncertainty and error analyses of interferometric location results have become important topics. Based on the plane wave approximation of lightning RF signal transmission for interferometric location, a geometric model for the systematic error in the interferometric location due to the wavefront being spherical is developed and evaluated for short baseline interferometers, beginning with orthogonal and equilateral triangle baseline layouts. The symmetry of the baseline layout is shown to be helpful for reducing the systematic error caused by the plane wave approximation in interferometric location using a geometric relationship. Furthermore, a baseline layout scheme with the center of an equilateral triangle as the coordinate origin optimizes the systematic error caused by the plane wave approximation.

Based on the very high-frequency (VHF) pulse signal detected by interferometer (INTF) antennas as the reference unit, a lightning location method based on pulse matching and peak extraction is proposed in this paper. The first key step of this method is to optimize the raw VHF data of the INTF by using the ensemble empirical mode decomposition (EEMD) method to control the quality of the original signal by bandpass filtering to preserve only the relatively high-frequency components of 40-80 MHz. Then, through a combination of main and auxiliary windows, the matching of waveforms from different antennas is realized by means of generalized cross-correlation. In a microscale window (11 ns), the pulse signals are further accurately matched, and the arrival time difference is calculated under threshold and similarity constraints. Finally, the two-dimensional coordinates of the matched pulsed radiation sources are obtained using the nonlinear least squares method. Compared with the results of the traditional ’centroid’ approach, the number of radiation sources obtained with the proposed method is greatly increased. For lightning cases with different discharge intensities, the number of located radiation sources can be increased by a factor of 10-20, and for specific short-duration and rapidly changing discharge processes (such as dart leaders and K events), the number of located radiation sources can be increased by a factor of nearly 100. For INTF operation at 20-88 MHz with a sampling rate of 180 MHz, the analytical resolution of the lightning discharge process can be improved to 10 ns. When the radiation sources are fully located, the widths of the lightning channels can be successfully reconstructed from the density of no-size points representing the located radiation sources in space.

Cloud‐to‐ground strokes, narrow bipolar events, and energetic in‐cloud pulses are known classes of high peak‐current lightning processes that occur in thunderstorms. Here, we report one more distinct class of high peak‐current events observed exclusively over mountainous terrain, usually above 2,000 m altitude, in the continental Unites States. These events, which we call mountain‐top energetic pulses (MEPs), are bipolar pulses with negative radiated field polarities. MEPs are generated between the high mountain tops and compact overhead thunderclouds. Evidence supports the hypothesis that MEPs are produced by terrain‐initiated upward positive leaders propagating in high electric fields due to the proximity of the low negative charge regions of the thunderstorms. This scenario further suggests the possibility that MEPs are associated with downward terrestrial gamma‐ray flashes, and their high peak currents imply that they may produce elves.

Since their introduction 22 years ago, lightning mapping arrays (LMA) have played a central role in the investigation of lightning physics. Even in recent years with the proliferation of digital interferometers and the introduction of the LOw Frequency ARray (LOFAR) radio telescope, LMAs still play an important role in lightning science. LMA networks use a simple windowing technique that records the highest pulse in either 80 μs or 10 μs fixed windows in order to apply a time-of-arrival location technique. In this work we develop an LMA-emulator that uses lightning data recorded by LOFAR to simulate an LMA, and we use it to test three new styles of pulse windowing. We show that they produce very similar results as the more traditional LMA windowing, implying that LMA lightning mapping results are relatively independent of windowing technique. In addition, each LMA station has its own GPS-conditioned clock. While the timing accuracy of GPS receivers has improved significantly over the years, they still significantly limit the timing measurements of the LMA. Recently, new time-of-arrival techniques have been introduced that can be used to self-calibrate systematic offsets between different receiving stations. Applying this calibration technique to a set of data with 32 ns uncertainty, observed by the Colorado LMA, improves the timing uncertainty to 19 ns. This technique is not limited to LMAs and could be used to help calibrate future multi-station lightning interferometers.

Since their introduction 22 years ago, lightning mapping arrays (LMA) have played a central role in the investigation of lightning physics. Even in recent years with the proliferation of digital interferometers and the introduction of the LOw Frequency ARray (LOFAR) radio telescope, LMAs still play an important role in lightning science. LMA networks use a simple windowing technique that records the highest pulse in either 80 $\mu$s or 10 $\mu$s fixed windows in order to apply a time-of-arrival location technique. In this work we develop an LMA-emulator that uses lightning data recorded by LOFAR to simulate an LMA, and we use it to test three new styles of pulse windowing. We show that they produce very similar results as the more traditional LMA windowing, implying that LMA lightning mapping results are relatively independent of windowing technique. In addition, each LMA station has its own GPS-conditioned clock. While the timing accuracy of GPS receivers has improved significantly over the years, they still significantly limit the timing measurements of the LMA. Recently, new time-of-arrival techniques have been introduced that can be used to self-calibrate systematic offsets between different receiving stations. Applying this calibration technique to a set of data with 32 ns uncertainty, observed by the Colorado LMA, improves the timing uncertainty to 19 ns. This technique is not limited to LMAs and could be used to help calibrate future multi-station lightning interferometers.

Based on the ensemble empirical mode decomposition (EEMD) method, a DAF method for signal construction is proposed that repeatedly decomposes (D) the signal, amplifies (A) the local signal characteristics, and then filters (F) the signal. This method is used to decompose and reconstruct the electric field waveform (called a sferic) of an energetic in-cloud pulse (EIP) with a 247-kA peak current detected by a fast antenna (FA). Based on synchronous sub-microsecond very high-frequency (VHF, 14–88 MHz) radio interferometer (INTF) observations and observed downward fast positive and upward fast negative breakdowns, which occurred simultaneously with the EIP, the EIP sferic is decomposed by the DAF method in 11 steps into two independent sferics: a smoother filtered EIP sferic and an embedded narrow bipolar-like event (NBE). It is verified that strong VHF radiation is generated by the NBE-like event, rather than being associated with the smooth EIP sferic. The analysis, decomposition, and reconstruction of the correlated signals by the EEMD-based DAF method proposed in this article support the idea that the large-amplitude EIP sferic was generated by relativistic discharge responsible for an accompanying terrestrial gamma-ray flash (TGF) rather than by streamer or leader activity.