← News
Science Breakthroughs Science Breakthroughs Knowledge

Lightning’s “fan” may be made of tiny, steadily advancing leaders

Lightning’s “fan” may be made of tiny, steadily advancing leaders
LOFAR Telescope
About 1 M Filament radius
About 100,000 M/S Leader speed
Hundreds Of Meters filament length
100,000 M/S filament growth speed

Some high-altitude lightning filaments travel hundreds of meters without the stop-start motion that scientists usually associate with negative leaders. In a sharply resolved radio image, the filaments advanced at about 100,000 meters per second, stayed remarkably narrow, and kept emitting radio pulses along their tips. The observations suggest that lightning’s familiar fan-shaped bursts may be built from small, leader-like channels—not simply from the fleeting streamers long assumed to fill them.

That distinction matters because the way a lightning channel grows is not just a matter of appearance. It goes to a central puzzle in atmospheric electricity: why does one end of a lightning discharge often advance smoothly, while the negative end tends to jump forward in steps? The new measurements do not solve that puzzle. They do show that the structures inside high-altitude negative leaders are more intricate than the usual picture of a single step followed by a spray of streamers.

The work by M. A. A. Lourens and colleagues uses the LOFAR radio telescope to map a high-altitude negative leader in unusual detail. The team argues that the filaments making up its fans behave like negative leaders in important respects. They are not claiming to have watched hot, glowing channels directly: the evidence comes from radio emissions and the spatial and temporal pattern of those emissions. But that evidence is enough to challenge a long-standing interpretation—and to point toward observations that could settle what these structures really are.

(a)
(a) Source: M. A. A. Lourens, B. M. Hare

The Science

Lightning is a growing electrical discharge. As it travels through air, it forms conducting channels called leaders. A leader has two ends with opposite electrical charges; the negative end is usually described as advancing in a sequence of discrete jumps, or steps, while the positive end can propagate more continuously. The detailed physics behind that contrast remains an open question.

The negative leader’s tip is surrounded by a region of intense electric field, where smaller discharges can form. One common explanation is that the visible or radio-mapped filaments ahead of the leader are streamer coronas. Streamers are narrow, self-propagating electrical discharges: they can extend through air, but unlike leaders they are not generally expected to leave behind a lasting, heated, conducting channel. The distinction is important. A collection of streamers would imply a different way for the overall lightning channel to advance than a set of leader-like filaments would.

Earlier observations often showed high-altitude negative leaders as a series of fan-shaped structures, with each fan interpreted as a step. At lower resolution, that picture is intuitive: a broad, radiating front appears, then another one farther along. Yet the detailed anatomy of the fan remained obscure. The filaments within it were often treated as streamer activity at the leader tip, in part because fine-scale measurements were not available.

Lourens and colleagues revisited a lightning event recorded on August 14, 2020. LOFAR, the Low Frequency Array, uses antennas spread across multiple stations; for this event, 200 antennas across 36 stations in the northern Netherlands observed radio signals in the 30–80 MHz band. Lightning produces many brief pulses in this very-high-frequency (VHF) range, particularly near the tips of negative leaders. By combining the signals from many antennas, researchers can infer where the pulses originated and reconstruct the evolving discharge in three dimensions.

The team used a recently developed imaging method called ATRI-D, which coherently combines radio signals and can reveal weak sources that are hard to distinguish on individual antennas. Its high spatial resolution allowed the researchers to zoom into one part of a leader’s fan. The paper presents the result as the highest-resolution image of high-altitude negative-leader propagation available to the authors.

(b)
(b) Source: M. A. A. Lourens, B. M. Hare

This is not a photograph of the lightning channel’s temperature or electrical current. The instrument maps radio-emitting events, and the inferred width of a filament is limited by the image’s resolution and by interference from many sources firing at nearly the same time. The researchers are careful about this: the measured width is an upper limit, not a precise physical diameter. They infer conductivity from how radio-emitting bursts appear along previously formed tracks and from the way the structures develop, rather than measuring resistance directly.

What They Found

At the broad scale, the leader still looks like a succession of fans. These structures can appear milliseconds apart and can be separated by as much as 300 meters. But when the researchers examined one fan closely, they found a dense, branching network of thin filaments. Different filaments began at different positions and times; the fan was not one uniformly advancing sheet.

Some individual tracks could be followed as they moved forward. The filaments were reported to be about 100 to 500 meters long. Their average speed was roughly 100,000 meters per second. In the imaged example, a filament covered about 250 meters in three milliseconds, corresponding to a speed of about 80,000 meters per second—consistent with the paper’s approximate overall estimate. The team saw no clear broadening as filaments traveled long distances. In the particular imaged structure, their apparent width was below 1.1 meters, though the authors stress that the instrument may not resolve their true width.

Observed filament lengths

The paper reports individual filaments 100–500 meters long; these endpoints show the reported range, not a frequency distribution.

Observed filament lengths
LabelValue
Lower reported length100 m
Upper reported length500 m

The radio emission was not confined to a single explosive start. Short VHF pulses continued to appear near the advancing tips. Stronger bursts marked the beginnings of new filament networks, and the team also identified later bursts along older filament tracks. In several cases, a new branch began from the body of an existing filament rather than only from its furthest point. The authors interpret this pattern as evidence that charge can accumulate at different places along a filament over timescales reaching milliseconds.

That interpretation suggests a channel carrying some current: enough charge moves along the structure for later activity to occur at different points, but not so freely that charge cannot build up locally. The authors describe the filaments as likely having moderate conductivity. That is an inference, not a direct conductivity measurement. It nevertheless gives the filaments a property more characteristic of leader-like channels than of isolated streamers.

The researchers offer several reasons why they do not think each observed filament is simply one negative streamer. In their account, a single streamer traveling at a steady speed would not explain the repeated, strong VHF pulses detected at the moving tip. A streamer explanation would require many streamers acting together, yet the mapped filaments are few and narrow compared with the enormous number of streamers expected near a leader tip. And a bundle of branching streamers is expected to spread outward; the observed filaments, by contrast, remain tightly confined over hundreds of meters.

Figure 2: Deviations of located VHF sources from the average track indicated in Fig. 1(b) as histograms and scatter plots in time. The three histograms show the deviations of VHF source locations from the average location of the filament in easting, northing, and altitude, which are indicated using the vertical axis. A scatter showing the time evolution of these deviations is plotted on top of the histograms with time indicated on the top abscissa. The broadening of the easting and altitude scatter at 954.5 ms is due to the filament branching. The RMS deviation from the average track in each dimension is indicated in the bottom right corner of each panel.
Figure 2: Deviations of located VHF sources from the average track indicated in Fig. 1(b) as histograms and scatter plots in time. The three histograms show the deviations of VHF source locations from the average location of the filament in easting, northing, and altitude, which are indicated using the vertical axis. A scatter showing the time evolution of these deviations is plotted on top of the histograms with time indicated on the top abscissa. The broadening of the easting and altitude scatter at 954.5 ms is due to the filament branching. The RMS deviation from the average track in each dimension is indicated in the bottom right corner of each panel. Source: M. A. A. Lourens, B. M. Hare

The comparison with earlier high-altitude observations also highlights that these events can look different. Previous work cited by the authors reported a fast-propagating discharge at about 10 million meters per second that extended roughly 100 meters before producing a diffuse cloud of emission. In the fan analyzed here, the researchers saw no comparable fast advance: instead, narrow filaments progressed at roughly 100,000 meters per second, with the strongest emission localized to a small region. The paper therefore describes the filaments in this event as a distinct mode, not a universal description of every high-altitude leader.

Speeds in two reported high-altitude leader observations

Contrasts the approximate speed of filaments in this study with the fast discharge described in earlier work cited by the authors.

Speeds in two reported high-altitude leader observations
LabelValue
Filaments in this study100,000 m/s
Fast discharge in prior work10,000,000 m/s

Branching was common, but it did not obviously change the propagation speed. The filaments also tended to bend around earlier structures and avoid crossing through the same volume again. Some later tracks began on older ones, while appearing to steer around parts already occupied. These spatial relationships hint that the discharge’s history matters: previous activity may influence where the next channel can form. The paper documents the pattern, but does not yet identify the mechanism behind it.

The key result is not that the traditional concept of a lightning step has vanished. At the scale of a whole leader, fans still appear as repeated advances. The result is that the fan’s interior is not a simple object. Several separately initiated, sometimes branching filaments can contribute to what a lower-resolution image makes look like a single step.

Why This Changes Things

The conventional account of negative leaders links their stepping behavior to the fact that negative charge repels free electrons. Positive leaders, by contrast, attract electrons and are associated with more continuous growth. How that basic charge difference produces the observed step-by-step behavior is still a major question in lightning research. These new observations complicate the story: at least some negative filaments appear to move continuously, at nearly constant speed, for hundreds of meters.

That finding does not mean the negative and positive ends of a lightning discharge are now understood to behave alike. The authors emphasize that negative filaments emit VHF pulses from their tips, while positive leaders are relatively quiet in this radio band. The new observation is narrower but still striking: a negative structure can show leader-like, self-focused continuous growth even though negative leaders as a whole are commonly characterized by stepping.

The difference between scales may help reconcile the apparent contradiction. If many filaments grow at different times and from different points, their combined activity can produce a fan and make the overall leader look as though it advances in discrete jumps. What looks like one step from farther away may be a bundle of processes unfolding within it. That is a reminder that scientific categories built from coarse images can conceal the behavior of their components.

The authors propose that a feedback between heating and conductivity might help keep a filament narrow. Streamer activity at the tip could heat the center of the advancing region; once the channel becomes sufficiently conductive, current through it may add further heating. That could concentrate energy near the channel’s center rather than allowing the structure to fan outward. This is a possible explanation, not a mechanism demonstrated by the radio measurements. The researchers also note that space leaders—small, hot channels sometimes invoked to explain steps—could still play a role in heating the filaments, even if the observed propagation does not fit the standard stepping account neatly.

LOFAR observation: receiver frequency band

The radio telescope recorded lightning emissions in the VHF band stated in the paper.

LOFAR observation: receiver frequency band
LabelValue
Lower frequency30 MHz
Upper frequency80 MHz

The implications reach beyond one unusual cloud-to-air discharge, but their range is not yet known. The event was observed at high altitude, where the air is thinner. The paper leaves open whether negative leaders lower in the atmosphere behave in the same way, perhaps with the structures scaled differently, or whether the high-altitude fans depend on local features such as pockets of electrical charge. The answer matters for deciding whether this is a general feature of negative lightning or a special mode found in particular atmospheric conditions.

There is also an interpretive consequence for optical observations. Dim, extended light near a leader tip has often been attributed to streamer coronas. The authors point out that leader channels can have a broad range of temperatures, and that the light a plasma emits can change dramatically across that range. A leader-like filament could therefore be faint enough to resemble a streamer in a photograph. That possibility does not prove that earlier optical interpretations were wrong, but it does mean that brightness alone may not settle the question.

What’s Next

The most immediate need is to observe the same structures in multiple ways at once. LOFAR maps radio-emitting activity with remarkable spatial detail, but it does not directly reveal channel temperature or show whether a filament is optically bright. High-speed optical imaging synchronized with radio measurements could connect the radio pulses to the physical channels and test whether the mapped filaments are the smoothly moving negative channels seen in some video observations.

The team also plans to compare filaments across different environments and investigate whether lower-altitude negative leaders develop similar structures. A more detailed analysis of radio polarization could help reveal the orientation of the emitting dipoles, although the authors say interpretation is complex and their preliminary look at one filament is not yet conclusive. Improved models will need to explain both the persistent narrowness of the tracks and how multiple networks start at different locations along earlier channels.

The study has important limits. It analyzes a selected region of a single high-altitude event, and the apparent filament width is constrained by the resolution of the observation. The researchers infer moderate conductivity from patterns in the radio sources; they do not measure channel current or temperature directly. Nor do they yet explain why the filaments organize into repeated fans, or how the fan-scale pattern relates to the familiar steps of negative leaders.

Even so, the observation changes what scientists should look for. A lightning fan need not be a single advancing front, and its narrow strands need not be disposable streamers. Some may be conducting channels that grow smoothly, branch, and leave a history that shapes what happens next. Better joint radio and optical observations can now test that possibility—bringing researchers closer to understanding how a discharge becomes a lightning bolt, one filament at a time.