Read our COVID-19 research and news.

Time-lapse photo of a plane in a lightning storm

Gene Blevins/REUTERS

Where will lightning strike next? A new model could help pilots avoid dangerous storms

Weather models can forecast severe storms pretty well, but predictions of lightning have remained elusive. Now, researchers have created global simulations of lightning that more accurately capture when and where the strikes will occur—which could help people trying to avoid them, such as airline pilots.

Lightning generally requires two ingredients. First, it needs warm, rising air, or convection, to create thunderclouds. It also needs the thunderclouds to contain icy pellets known as graupel. Colliding pellets transfer electric charge, creating an electric field. A lightning bolt forms when that field gets big enough.

Weather and climate models, which divide the atmosphere into grid boxes of a certain size, have struggled to simulate lightning because their spatial resolution is too coarse, typically 100 kilometers or so. The processes that give rise to convective thunderclouds and graupel happen at too small a scale for computers to simulate them globally in any reasonable amount of time. To make daily forecasts, weather models have to instead rely on “parameterizations” for things like convection—ad hoc rules of thumb that can be performed quickly.

To avoid some of these assumptions, Paul Field, a cloud scientist at the Met Office, the United Kingdom’s national weather service with headquarters in Essex, and his colleagues simulated 5 years’ worth of lightning in a global model that could resolve details as small as 10 kilometers. That allowed them to simulate the convective cloud formation processes explicitly, although they still had to make assumptions about the formation of graupel.

Even so, the team’s model accurately pinpointed lightning hot spots in South America, Africa, and southeast Asia that experience nearly 100 lightning flashes per square kilometer per year. The model also correctly captured how lightning typically occurs in the afternoon, around 3 p.m. local time. That timing makes sense because the ground has warmed by then, and warmer air has had time to move upward and form clouds, Field says.

The simulations also reproduced some real-world peculiarities of lightning. For instance, the new model accurately showed how lightning over Lake Victoria in Africa occurs late in the day. This effect is due to the lake’s water heating more slowly than the surrounding land, resulting in delayed upward movement of warmer air, Field says. The model also reproduced the daily eastward movement of lightning over the Great Plains of the United States, a pattern caused by prevailing winds, the team reports in the Journal of Geophysical Research.

These new lighting maps could potentially provide a better estimate of lightning threats to aircraft, the researchers suggest. Scientists could use this model to generate a map of weather hazards for aviation, says Field, because existing maps are “pretty crude.”

The work could also be used to predict how lightning patterns might shift in different climate change scenarios, changes that will impact Earth’s atmosphere, says Declan Finney, an atmospheric scientist at the University of Leeds in the United Kingdom who was not involved in the research.