The paper reviews recent advances in studies of electric discharges in the stratosphere and mesosphere above thunderstorms, and their effects on the atmosphere. The primary focus is on the sprite discharge occurring in the mesosphere, which is the most commonly observed high altitude discharge by imaging cameras from the ground, but effects on the upper atmosphere by electromagnetic radiation from lightning are also considered. During the past few years, co-ordinated observations over Southern Europe have been made of a wide range of parameters related to sprites and their causative thunderstorms. Observations have been complemented by the modelling of processes ranging from the electric discharge to perturbations of trace gas concentrations in the upper atmosphere. Observations point to significant energy deposition by sprites in the neutral atmosphere as observed by infrasound waves detected at up to 1000 km distance, whereas elves and lightning have been shown significantly to affect ionization and heating of the lower ionosphere/mesosphere. Studies of the thunderstorm systems powering high altitude discharges show the important role of intracloud (IC) lightning in sprite generation as seen by the first simultaneous observations of IC activity, sprite activity and broadband, electromagnetic radiation in the VLF range. Simulations of sprite ignition suggest that, under certain conditions, energetic electrons in the runaway regime are generated in streamer discharges. Such electrons may be the source of X-and Gamma-rays observed in lightning, thunderstorms and the so-called Terrestrial Gamma-ray Flashes (TGFs) observed from space over thunderstorm regions. Model estimates of sprite perturbations to the global atmospheric electric circuit, trace gas concentrations and atmospheric dynamics suggest significant local perturbations, and possibly significant meso-scale effects, but negligible global effects.

[1] Five gigantic jets (GJs) have been recorded with video and photograph cameras on 7 March 2010 above an isolated tropical storm east of Réunion Island. Three of them were produced before the storm reached its coldest cloud top temperature (approximately −81°C), and two others occurred during the cloud extension. Thanks to the close distance of observation (∼50 km), the luminosity within the cloud was recorded, and the events are analyzed in unprecedented detail. The tops of the GJs are estimated between 80 and 90 km. All these GJs are accompanied by long, continuous cloud illumination, and they are preceded and followed by intermittent optical flashes from the cloud, most of time without any cloud-to-ground (CG) flash simultaneously detected, which suggests they originated mainly as intracloud discharges and without any charge transfer to Earth. The CG lightning activity is observed to cease a few tens of seconds before the jets. According to ELF data recorded at Nagycenk, Hungary, the five GJs serve to raise negative charge. Their duration ranges from 333 to 850 ms. The leading jet has the most variable duration (33-167 ms) and propagates faster at higher altitudes. The trailing jet exhibits a continuous decrease of luminosity in different parts of the jet (lower channel, transition zone and, for most events, carrot sprite-like top) and in the cloud, with possible rebrightening. The lower channels (∼20-40 km altitude) produce blue luminosity which decreases with altitude and become more and more diffuse with time. The transition zone (around 40-65 km) consists of bright red, luminous beads slowly going up (∼10 4 m s −1), retracing the initial leading jet channels.Citation: Soula, S., O. van der Velde, J. Montanya, P. Huet, C. Barthe, and J. Bór (2011), Gigantic jets produced by an isolated tropical thunderstorm near Réunion Island,

The salient issues related to lightning protection of long wind-turbine blades are discussed in this paper. We show that the lightning protection of modern wind turbines presents a number of new challenges due to the geometrical, electrical, and mechanical particularities of the turbines. The risk assessment for the lightning-protection-system design is solely based today on downward flashes. We show in this paper that the majority of the strikes to modern turbines are expected to be upward lightning. Neglecting upward flashes, as implicitly done by the International Electrotechnical Commission, might result in an important underestimation of the actual number of strikes to a tall wind turbine. In addition, we show that the rotation of the blades may have a considerable influence on the number of strikes to modern wind turbines as these may be triggering their own lightning. Because wind turbines are tall structures, the lightning currents that are injected by return strokes into the turbines will be affected by reflections at the top, bottom, and junction of the blades with the static base of the turbine. This is of capital importance when calculating the protection of internal circuitry that may be affected by magnetically induced electromotive forces that depend directly on the characteristics of the current in the turbine. The presence of carbon-reinforced plastics (CRP) in the blades introduces a new set of problems to be dealt with in the design of the turbines' lightning protection system. One problem is the mechanical stress resulting from the energy dissipation in CRP laminates due to the circulation of eddy currents. We evaluate in this paper the dissipated energy and propose recommendations as to the number of down conductors and their orientation with respect to the CRP laminates so that the dissipated energy is minimized. It is also emphasized that the high static fields under thunderclouds might have an influence on the moving carbon-fiber parts. This issue needs to be addressed by lightning protection researchers and Manuscript engineers. Representative full-scale blade tests are still complex because lightning currents from an impulse current generator are conditioned to the electrical characteristics of the element under test and return paths. It is therefore desirable to complement laboratory tests with theoretical and computer modeling for the estimation of fields, currents, and voltages within the blades.

New observations with a 3-D Lightning Mapping Array and high-speed video are presented and discussed. The first set of observations shows that under certain thunderstorm conditions, wind turbine blades can produce electric discharges at regular intervals of~3 s in relation to its rotation, over periods of time that range from a few minutes up to hours. This periodic effect has not been observed in static towers indicating that the effect of rotation is playing a critical role. The repeated discharges can occur tens of kilometers away from electrically active thunderstorm areas and may or may not precede a fully developed upward lightning discharge from the turbine. Similar to rockets used for triggering lightning, the fast movement of the blade tip plays an important role on the initiation of the discharge. The movement of the rotor blades allows the tip to "runaway" from the generated corona charge. The second observation is an uncommon upward/downward flash triggered by a wind turbine. In that flash, a negative upward leader was initiated from a wind turbine without preceding lightning activity. The flash produced a negative cloud-to-ground stroke several kilometers from the initiation point. The third observation corresponds to a high-speed video record showing simultaneous upward positive leaders from a group of wind turbines triggered by a preceding intracloud flash. The fact that multiple leaders develop simultaneously indicates a poor shielding effect among them. All these observations provide some special features on the initiation of lightning by nonstatic and complex tall structures.

Lightning flashes develop as a bidirectional tree, with a negative and a positive end propagating simultaneously into cloud charges of the opposite polarity. In recent years, this process has been modeled assuming that the bidirectional channel remains a zero‐net‐charge perfect conductor starting at the electric potential of its initial location. So far, both leader ends were treated identically. We summarize characteristics of bidirectional leader development of various lightning flashes registered by the Ebro 3‐D Lightning Mapping Array at the east coast of Spain, supplemented by high‐speed camera records. In order to follow the horizontal development of positive and negative leaders over time, a time‐distance‐altitude graph was designed, using the flash origin or a cloud‐to‐ground stroke as reference. The examples confirm that negative and positive leaders propagate at characteristic horizontal speeds (105 and 2 · 104 m s−1). The positive leader's low apparent speed corresponds to the phase when recoil processes are active. Very fast negative leaders (up to 8 · 105 m s−1) are detected in association with positive cloud‐to‐ground strokes. Negative leaders respawn repeatedly from the origin or as a retrograde negative leader at the positive branch. Positive leaders remain propagating throughout the flash or until reaching ground. We show that the velocity difference shifts the potential of the leader, increasing the gradient with cloud charge at the positive end while reducing it at the negative end. The positive section lowers its potential through retrograde negative leaders, eventually making it possible to emit a new negative leader into the upper positive potential well.

During the summers of 2003 to 2006 sprites were observed over thunderstorms in France by cameras on mountain tops in Southern France. The observations were part of a larger coordinated effort, the EuroSprite campaigns, with data collected simultaneously from other sources including the French radar network for precipitation structure, Meteosat with images of cloud top temperature and the Météorage network for detection of cloud-to-ground (CG) flash activity. In this paper two storms are analyzed, each producing 27 sprite events. Both storms were identified as Mesoscale Convective Systems (MCS) with a trailing stratiform configuration (ST) and reaching a maximum cloud area of~120,000 km 2 . Most of the sprites were produced while the stratiform area was clearly developed and during periods of substantial increase of rainfall in regions with radar reflectivity between 30 and 40 dBZ. The sprite-producing periods followed a maximum in the CG lightning activity and were characterized by a low CG flash rate with a high proportion of +CG flashes, typically around 50%. All sprites were associated with +CGs except one which was observed after a −CG as detected by the Météorage network. This −CG was estimated to have −800 C km charge moment change. The peak current of spriteproducing + CG (SP + CG) flashes was twice the average value of +CGs and close to 60 kA with little variation between the periods of sprite activity. The SP + CG flashes were further characterized by short time intervals before a subsequent CG flash (median value b 0.5 s) and with clusters of several CG flashes which suggest that SP + CG flashes often are part of multi-CG flash processes. One case of a lightning process associated with a sprite consisted of 7 CG flashes.

Lightning flashes are known to initiate in regions of strong electric fields inside thunderstorms, between layers of positively and negatively charged precipitation particles. For that reason, lightning inception is typically hidden from sight of camera systems used in research. Other technology such as lightning mapping systems based on radio waves can typically detect only some aspects of the lightning initiation process and subsequent development of positive and negative leaders. We report here a serendipitous recording of bidirectional lightning initiation in virgin air under the cloud base at ~11,000 images per second, and the differences in characteristics of opposite polarity leader sections during the earliest stages of the discharge. This case reveals natural lightning initiation, propagation and a return stroke as in negative cloud-to-ground flashes, upon connection to another lightning channel – without any masking by cloud.

A World Meteorological Organization weather and climate extremes committee has judged that the world’s longest reported distance for a single lightning flash occurred with a horizontal distance of 321 km (199.5 mi) over Oklahoma in 2007, while the world’s longest reported duration for a single lightning flash is an event that lasted continuously for 7.74 s over southern France in 2012. In addition, the committee has unanimously recommended amendment of the AMS Glossary of Meteorology definition of lightning discharge as a “series of electrical processes taking place within 1 s” by removing the phrase “within 1 s” and replacing it with “continuously.” Validation of these new world extremes 1) demonstrates the recent and ongoing dramatic augmentations and improvements to regional lightning detection and measurement networks, 2) provides reinforcement regarding the dangers of lightning, and 3) provides new information for lightning engineering concerns.