Lightning is one of nature’s most reliable fire starters. Across the contiguous United States alone, lightning sparks roughly 16 percent of all wildfires, and those fires account for more than half the total area burned because they tend to ignite in remote, dry landscapes where suppression is slow to arrive. The connection between a bolt from the sky and a burning forest is not as straightforward as it looks, though. Not every flash has the electrical properties needed to kindle fuel, and even the ones that do depend on a precise alignment of weather, terrain, and vegetation moisture to turn a spark into a spreading fire.
Why Some Flashes Start Fires and Others Don’t
A typical lightning bolt lasts only a fraction of a second, and that brief jolt alone rarely sets vegetation alight. What matters most is whether the electrical discharge includes what researchers call a continuing current, a lower-amplitude flow of electricity that persists after the initial bright return stroke. Laboratory experiments and field observations consistently show that continuing currents lasting more than a few tens of milliseconds are the critical ingredient for ignition. A flash with a continuing current essentially holds a sustained arc of energy against the fuel, giving it time to heat plant material past its ignition point.
Only a small fraction of all lightning flashes carry these prolonged currents. Across North America during summer, flashes with long continuing currents make up less than about ten percent of all lightning. Yet when researchers in the western United States matched lightning detection data to known wildfire ignition points, they found that these current-carrying flashes were responsible for a disproportionately large share of fire starts.1Nature Communications. Variation of lightning-ignited wildfire patterns under climate change A separate analysis confirmed the pattern: flashes with continuing currents longer than ten milliseconds had a higher probability of ignition than ordinary cloud-to-ground lightning, and one video-recorded fire-igniting flash showed a continuing current lasting about 400 milliseconds paired with an optical signal that glowed for two to four seconds.2Journal of Geophysical Research: Atmospheres. On the Role of Continuing Currents in Lightning‐Induced Fire Ignition
Positive Versus Negative Strokes
Lightning comes in two polarities. Most cloud-to-ground strokes carry negative charge from the cloud to the ground, while a smaller percentage carry positive charge. Positive strokes have long been suspected of being more effective fire starters because they tend to carry higher peak currents and are more likely to include those fire-critical continuing currents.
The evidence is mostly consistent with that idea, though not perfectly so. A study of lightning-caused forest fires in Austria found that positive strokes were significantly more likely to ignite a fire than negative ones.3Agricultural and Forest Meteorology. Characteristics of lightnings igniting forest fires in Austria Research in the Alps similarly found a significantly higher proportion of positive strokes among fire-igniting lightning compared to ordinary background lightning.4Agricultural and Forest Meteorology. Lightning-caused fires in the Alps: Identifying the igniting strokes But a Finnish study reached a more nuanced conclusion: positive strokes were not more likely to start fires than negative ones on a per-stroke basis, and adding more strokes to a single flash actually decreased the ignition probability of any individual stroke.5Agricultural and Forest Meteorology. Lightning that ignites forest fires in Finland The discrepancy may reflect differences in landscape, fuel type, and climate between Finland’s boreal forests and the Alpine and Austrian settings. In short, polarity matters, but the local environment can shift the odds enough to complicate the simple “positive equals worse” story.
Dry Lightning and the Weather Conditions That Matter
The phrase “dry lightning” describes thunderstorms whose rain evaporates before reaching the ground, a scenario that terrifies wildland firefighters because the lightning arrives without any accompanying moisture to wet down fuels. These events are responsible for some of the largest fire outbreaks on record, including the August 2020 siege in California where thousands of dry lightning strikes in a matter of days ignited hundreds of fires simultaneously.
Researchers studying central and northern California found that dry lightning days are characterized by a specific atmospheric fingerprint: enhanced moisture and instability in the middle atmosphere combined with drier-than-normal conditions near the surface and elevated surface temperatures.6Environmental Research: Climate. Meteorological and geographical factors associated with dry lightning in central and northern California That combination is exactly what you would design if you wanted to maximize fire risk. The midlevel moisture feeds the thunderstorm’s convection, but the dry, hot layer near the ground evaporates any rain and leaves fuels parched. The result is lightning delivered to a landscape primed to burn.
Even when rain does reach the ground, the amount matters. A light sprinkle from a passing thunderstorm cell may not be enough to offset the ignition potential of a single well-placed stroke, especially if the rain falls on already-dry fuel beds in summer.
Fuel Moisture Is the Gatekeeper
Lightning can supply the heat, but whether that heat produces a fire depends overwhelmingly on the state of the vegetation it hits. Live fuel moisture content, the amount of water in living plants, acts as a threshold switch. A study using causal inference methods across California showed that when live fuel moisture dropped below a critical flammability threshold, the likelihood of a lightning strike turning into a fire was about 1.8 times higher statewide. In shrublands, where the vegetation is particularly fire-prone, that factor jumped to 2.5 times higher.7Geophysical Research Letters. Dry Live Fuels Increase the Likelihood of Lightning‐Caused Fires
This is why lightning-caused fires cluster so heavily in late summer and early fall in western North America. By that point, vegetation has been baking in summer heat for months, fine fuels like grass and needle litter are thoroughly dried out, and even live shrubs and tree canopies have drawn down their moisture reserves. A lightning flash that would fizzle harmlessly in June can set off a major fire in September given the same terrain.
Holdover Fires and Delayed Ignition
One of the more counterintuitive aspects of lightning fire is that a strike can smolder underground or in deep organic soil layers for days or even weeks before producing visible flames. These are called holdover fires, and they create enormous headaches for fire managers trying to figure out what started a blaze.
The process works like this: a lightning stroke ignites organic material in the litter or duff layer of the forest floor, but the rain and cooler conditions that typically accompany a thunderstorm prevent the fire from immediately flaring up. Instead, the ignition enters a smoldering phase, slowly burning through subsurface organic material at low temperatures and without visible flame. The fire can persist in this state for days, surviving rainfall and even cool nights. When conditions eventually shift, with drier weather, lower humidity, or wind picking up, the smoldering combustion transitions to flaming combustion and a visible wildfire emerges seemingly out of nowhere.8Copernicus Publications (Earth System Science Data). A global database on holdover time of lightning-ignited wildfires
Holdover periods commonly range from a day or two up to about two weeks, though rare cases have been documented at even longer intervals. The phenomenon means a fire’s official discovery date can be far removed from the actual lightning strike that caused it, complicating both forensic investigation and the statistical models researchers use to link lightning detection data to fire origins.
Which Trees Are Most Likely to Be Struck
Not all trees in a forest face equal lightning risk. A modeling study that combined electrical theory with field data found that the likelihood of a direct strike increases with larger exposed crown area and with relative canopy position. Trees that rise above their neighbors, the emergent and upper-canopy trees, are struck far more often than those tucked beneath the canopy. Once lightning hits, secondary damage from the electrical current spreading through the ground or jumping between trees depends on the tree’s diameter and how close its neighbors stand.9Journal of Ecology. A mechanistic and empirically supported lightning risk model for forest trees
This has practical implications for fire behavior. The tallest trees in a stand, often old-growth specimens with thick bark and deep root systems, are the most common strike targets. If a strike ignites a fire in the canopy of one of these trees, it can drop burning material to the forest floor and start a surface fire from above. In some ecosystems, the same tall trees that are most strike-prone are also the most fire-resistant, having evolved thick bark and self-pruning lower branches precisely because they occupy landscapes shaped by millennia of lightning fire.
Lightning Fire Versus Human-Caused Fire
Despite the dramatic imagery of lightning-sparked blazes, people are the dominant source of wildfire ignitions in the United States. Over a 21-year study period, human-caused fires accounted for roughly 84 percent of all wildfires and about 44 percent of total area burned. The human-caused fire season stretched three times longer than the lightning fire season and added an average of 40,000 extra wildfires per year. Human-ignited fires were dominant across more than 5.1 million square kilometers of the country, while lightning-caused fires dominated in only about 0.7 million square kilometers, mostly in the sparsely populated mountainous West.10PubMed Central. Human-started wildfires expand the fire niche across the United States
The reason lightning fires still account for such a large share of area burned, despite being fewer in number, is geography. Lightning fires start in remote wildlands where fuel loads are heavy and suppression response is slower. Human fires tend to start closer to roads, structures, and communities, where they are detected and attacked faster but also where they do more immediate damage to property. The distinction matters for land management: reducing human-caused ignitions requires behavioral interventions and regulation, while managing lightning fire risk centers on fuel treatment, prescribed burning, and landscape-level planning.
When Fire Creates Its Own Lightning
In one of nature’s more dramatic feedback loops, large wildfires can generate their own thunderstorms and lightning. When a fire grows powerful enough, the intense heat creates a strong updraft that lofts smoke, ash, and moisture high into the atmosphere, forming towering pyrocumulonimbus clouds. These fire-generated storms behave much like regular thunderstorms and can produce lightning, which in turn can ignite new fires downwind of the original blaze.11Journal of Geophysical Research: Atmospheres. Impact of Smoke Aerosol Loading on Lightning Characteristics of Pyrocumulonimbus Compared With Other High‐Based Thunderstorms
Pyrocumulonimbus events are relatively rare, but they tend to happen during exactly the conditions, extreme heat, low humidity, and massive fire energy output, that also make new ignitions most dangerous. They can also inject smoke particles high into the stratosphere, affecting air quality thousands of kilometers from the fire. As fires grow larger and more intense under a warming climate, researchers expect these fire-generated lightning events to become more frequent, adding another complication to an already challenging fire management picture.
Climate Change Is Shifting the Odds
The relationship between lightning and fire is not static. Warmer temperatures and shifting weather patterns are changing both how often lightning occurs in fire-prone areas and how receptive the landscape is to ignition when it does. Global modeling work has found that even over time spans shorter than a decade, climate changes have steadily increased the global risk of lightning-ignited wildfires, with extreme fires becoming more frequent.12PubMed Central. Global lightning-ignited wildfires prediction and climate change projections based on explainable machine learning models
In the western United States specifically, projections using climate models under a moderate warming scenario show widespread increases in cloud-to-ground lightning days by mid-century. The northern parts of the region could see four to twelve additional lightning days per year compared to the recent historical period. And those lightning days are projected to coincide more frequently with dangerous fire weather, meaning not just more lightning, but more lightning arriving when conditions are ripe for ignition.13Earth’s Future. Projections of Lightning‐Ignited Wildfire Risk in the Western United States The combination of drier fuels, hotter surface temperatures, and more frequent thunderstorms in already fire-prone seasons creates a compounding risk that fire agencies are actively trying to plan for.
Tracking Lightning to Find Fire Origins
Fire investigators and managers rely on ground-based lightning detection networks to match wildfire ignition points to specific lightning strokes. In practice, this turns out to be harder than it sounds. A comparison of two operational detection networks used by fire managers in the western United States found that when looking for strokes within about 1.6 kilometers of 4,408 known lightning-fire origins, the two networks agreed on a detection near only 55 to 65 percent of fires. Using at least one of the two networks pushed that number up to 65 to 79 percent, but neither network detected any lightning near roughly a quarter to a third of all known lightning-caused fires.14Natural Hazards. Comparing ground-based lightning detection networks near wildfire points-of-origin
The detection gaps come from several sources. Lightning detection networks can miss low-energy flashes or flashes in areas with poor sensor coverage. Holdover fires shift the relevant time window, making it harder to identify which storm produced the igniting stroke. And location accuracy of both the lightning stroke and the fire’s origin point introduces uncertainty. For fire managers, the takeaway is that relying on a single detection network is not enough, and the absence of a detected stroke near a fire does not rule out lightning as the cause.
What Lightning Does to the Ground Itself
Beyond starting fires in vegetation, lightning can melt soil and rock at the point of impact, creating glass tubes called fulgurites. These natural artifacts form when the extreme heat of the lightning channel, which can exceed 30,000 degrees Celsius, fuses silica-rich soil into a hollow, branching glass structure. Experimental work replicating lightning conditions in the lab showed that fulgurites only formed when the discharge included a continuing current, the same electrical property required for fire ignition. The experiments used currents ranging from 220 to 350 amperes sustained for 100 to 500 milliseconds, and the resulting fulgurites showed decreasing porosity with longer current duration as more of the surrounding material melted.15PubMed Central. Experimental generation of fulgurite under realistic lightning discharge conditions
Fulgurites are more than curiosities. Geologists use them as physical records of past lightning activity, and their chemistry can provide information about the composition and temperature of the soil at the time of the strike. In sandy desert soils, fulgurites can extend a meter or more underground, tracing the path the electrical current followed as it dissipated into the earth.
Lightning Fire as an Ecological Force
For many ecosystems, lightning fire is not a catastrophe but a necessary part of the landscape’s life cycle. The longleaf pine forests that once covered vast stretches of the southeastern United States evolved under a regime of frequent, low-intensity surface fires ignited by lightning and by Indigenous burning practices. These fires occurred every two to eight years across the species’ range and were the dominant force shaping the vegetation, keeping understory competition in check and maintaining the open, parklike structure that longleaf pine depends on.16Elsevier (ScienceDirect). Lightning, fire and longleaf pine: Using natural disturbance to guide management
The connection between lightning and fire goes back much further than any modern ecosystem. Charcoal, preserved as fusain in ancient coal deposits and sediments, appears widely in rocks from the Carboniferous period, over 300 million years ago. Researchers studying these deposits have concluded that lightning and volcanic activity were the primary ignition sources for fires in those ancient forests.17Elsevier (ScienceDirect). The nature and influence of fire in Carboniferous ecosystems Fire, in other words, has been part of Earth’s terrestrial ecology for as long as there have been land plants to burn. The idea that fire is inherently destructive or unnatural is a modern misconception rooted in decades of fire-suppression policy. For many plant communities, the absence of fire, not its presence, is the ecological emergency.
Why Some Lightning Storms Cause Fire Outbreaks and Others Pass Quietly
Given everything above, the natural question is why a thunderstorm can roll through a bone-dry forest and leave no fires while a seemingly similar storm a week later sets the landscape ablaze. The answer is that fire ignition from lightning requires a stack of conditions to align simultaneously: the flash needs a sufficiently long continuing current; the struck surface needs to be dry enough to ignite; the immediate weather after the strike needs to be dry enough to sustain combustion rather than extinguishing it; and wind or terrain needs to be favorable enough for the fire to spread beyond a single smoldering point.
Each of those factors is independently variable. A storm might produce thousands of cloud-to-ground flashes, but if fewer than ten percent carry continuing currents, only a fraction of the storm’s total lightning has real ignition potential. Of those, only the strokes that contact receptive fuel, dry grass, needle litter, punky wood, or organic soil, have a chance of producing even a smoldering ignition. And of the smoldering ignitions that do occur, many extinguish on their own before conditions favor a transition to flaming combustion. The result is that the conversion rate from lightning flash to established wildfire is extremely low, on the order of a few percent or less of all cloud-to-ground strokes. But given that Earth receives millions of lightning strikes per day, even that small percentage translates to thousands of fires every year worldwide.