Lightning is one of the most destructive forces of nature, as in 2020 when it sparked the massive California Lightning Complex fires, but it remains hard to predict. A new study led by the ����Ӱ�Ӵ�ý shows that machine learning — computer algorithms that improve themselves without direct programming by humans — can be used to improve lightning forecasts.

Better lightning forecasts could help to prepare for potential wildfires, improve safety warnings for lightning and create more accurate long-range climate models.

“The best subjects for machine learning are things that we don’t fully understand. And what is something in the atmospheric sciences field that remains poorly understood? Lightning,” said , a UW associate professor of atmospheric sciences. “To our knowledge, our work is the first to demonstrate that machine learning algorithms can work for lightning.”

The new technique combines weather forecasts with a machine learning equation based on analyses of past lightning events. The hybrid method, Dec. 13 at the American Geophysical Union’s fall meeting, can forecast lightning over the southeastern U.S. two days earlier than the leading existing technique.

“This demonstrates that forecasts of severe weather systems, such as thunderstorms, can be improved by using methods based on machine learning,” said , who did the work for his UW doctorate in atmospheric sciences. “It encourages the exploration of machine learning methods for other types of severe weather forecasts, such as tornadoes or hailstorms.”

Researchers trained the system with lightning data from 2010 to 2016, letting the computer discover relationships between weather variables and lightning bolts. Then they tested the technique on weather from 2017 to 2019, comparing the AI-supported technique and an existing physics-based method, using actual lightning observations to evaluate both.

The new method was able to forecast lightning with the same skill about two days earlier than the leading technique in places, like the southeastern U.S., that get a lot of lightning. Because the method was trained on the entire U.S., its performance wasn’t as accurate for places where lightning is less common.

The approach used for comparison was a recently developed technique to forecast lightning based on the amount of precipitation and the ascent speed of storm clouds. That method has projected and a continued .

“The existing method just multiplies two variables. That comes from a human’s idea, it’s simple. But it’s not necessarily the best way to use these two variables to predict lightning,” Kim said.

The machine learning was trained on lightning observations from the , a collaborative based at the UW that has tracked global lightning since 2008.

“Machine learning requires a lot of data — that’s one of the necessary conditions for a machine learning algorithm to do some valuable things,” Kim said. “Five years ago, this would not have been possible because we did not have enough data, even from WWLLN.”

Commercial networks of instruments to monitor lightning now exist in the U.S., and newer geostationary satellites can monitor one area continuously from space, supplying the precise lightning data to make more machine learning possible.

“The key factors are the amount and the quality of the data, which are exactly what WWLLN can provide us,” Cheng said. “As machine learning techniques advance, having an accurate and reliable lightning observation dataset will be increasingly important.”

The researchers hope to improve their method using more data sources, more weather variables and more sophisticated techniques. They would like to improve predictions of particular situations like dry lightning, or lightning without rainfall, since these are especially dangerous for wildfires.

Researchers believe their method could also be applied to longer-range projections. Longer-range trends are important partly because lightning affects air chemistry, so predicting lightning leads to better climate models.

“In atmospheric sciences, as in other sciences, some people are still skeptical about the use of machine learning algorithms — because as scientists, we don’t trust something we don’t understand,” Kim said. “I was one of the skeptics, but after seeing the results in this and other studies, I am convinced.”

Other collaborators are and at the UW, and Yoo-Geun Ham and Jeong-Hwan Kim at Chonnam National University in South Korea.

For more information, contact Kim at daehyun@uw.edu or Cheng at wycheng@uw.edu. Cheng will this research online at 12:45 p.m. Central Time (10:45 a.m. Pacific, 1:45 p.m. Eastern) on Monday, Dec. 13.

Lightning strikes in the Arctic tripled from 2010 to 2020, a finding ����Ӱ�Ӵ�ý researchers attribute to rising temperatures due to human-caused climate change. The results, researchers say, suggest Arctic residents in northern Russia, Canada, Europe and Alaska need to prepare for the danger of more frequent lightning strikes.

The , published March 22 in Geophysical Research Letters, used data from the UW-based to map lightning strikes across the globe from 2010 to 2020. WWLLN sensors detect the short burst of radio waves emitted during a lightning strike.

The new study found the number of lightning strikes above 65 degrees north latitude during the summer months tripled from 2010 to 2020 as compared to the total number of lightning strikes over the entire globe during the same period.

“With long periods of ice-free ocean and increasing shipping in the Arctic, you’re going to have the same problem you have at lower latitudes — when there’s a lot of people and they don’t know about the lightning threat and it becomes a problem,” said lead author , a UW professor emeritus of Earth and space sciences.

Holzworth and his colleagues analyzed the frequency of Arctic lightning strikes occurring during the summer months of June, July and August from 2010 to 2020. They found the percentage of lightning strikes occurring in the Arctic tripled from 0.2% of global lightning strikes in 2010 to 0.6% in 2020. The actual number of lightning strikes above 65 degrees north increased from about 18,000 in 2010 to over 150,000 in 2020.

During the same time period, Arctic temperatures increased from 0.65 to 0.95 degrees Celsius above pre-industrial times. Holzworth and his colleagues attribute the increased lightning strikes to these rising temperatures, as warmer summers mean more chances for intense thunderstorms to develop and create lightning.

Lightning in the Arctic is historically rare, as it usually isn’t warm enough to generate the right thunderstorm conditions during which lightning occurs. But researchers have recently noticed more strikes occurring in the northernmost latitudes and they even reported several lightning strikes near the north pole in August 2019. Lightning strikes that do occur in the Arctic tend to happen in the summer when thunderstorms are most likely to form.

The Arctic is warming faster than any other region on Earth, and the study authors found the uptick in lightning strikes matched rising temperatures in the region over the past decade. Arctic temperatures increased by 0.3 degrees Celsius from 2010 to 2020; that warming has created more favorable conditions for intense summer thunderstorms that produce lightning, according to the authors.

Arctic sea ice is declining by about 13% every decade, . Less ice means more ocean will be available for shipping through the Arctic, especially in the summer months. Countries like Russia, China, Canada and the United States are already preparing to use the Arctic Ocean as a viable shipping route in the future.

The new study suggests shipping vessels throughout the Arctic could be more vulnerable to lightning strikes, in addition to those who call the Arctic home.

Co-authors are Michael McCarthy, Abram Jacobson, Craig Rodger and Todd Anderson at the UW; and James Brundell at the University of Otago in New Zealand.

For more information, contact Holzworth at bobholz@uw.edu. This was adapted from a from the AGU. An interactive embeddable graphic is available .

The lightning season in the Southeastern U.S. is almost finished for this year, but the peak season for the most powerful strokes of lightning won’t begin until November, according to a newly published global survey of these rare events.

A ����Ӱ�Ӵ�ý maps the location and timing of “superbolts” — bolts that release electrical energy of more than 1 million joules, or a thousand times more energy than the average lightning bolt, in the in which lightning is most active. Results show that superbolts tend to hit the Earth in a fundamentally different pattern from regular lightning, for reasons that are not yet fully understood.

The study was published Sept. 9 in the , a journal of the American Geophysical Union.

“It’s very unexpected and unusual where and when the very big strokes occur,” said lead author , a UW professor of Earth and space sciences who has been tracking lightning for almost two decades.

Holzworth manages the , a UW-managed research consortium that operates about 100 lightning detection stations around the world, from Antarctica to northern Finland. By seeing precisely when lightning reaches three or more different stations, the network can compare the readings to determine a lightning bolt’s size and location.

The network has operated since the early 2000s. For the new study, the researchers looked at 2 billion lightning strokes recorded between 2010 and 2018. Some 8,000 events — one in 250,000 strokes, or less than a thousandth of a percent — were confirmed superbolts.

“Until the last couple of years, we didn’t have enough data to do this kind of study,” Holzworth said.

The authors compared their network’s data against lightning observations from the Maryland-based company Earth Networks and from the New Zealand MetService.

The new paper shows that superbolts are most common in the Mediterranean Sea, the northeast Atlantic and over the Andes, with lesser hotspots east of Japan, in the tropical oceans and off the tip of South Africa. Unlike regular lightning, the superbolts tend to strike over water.

“Ninety percent of lightning strikes occur over land,” Holzworth said. “But superbolts happen mostly over the water going right up to the coast. In fact, in the northeast Atlantic Ocean you can see Spain and England’s coasts nicely outlined in the maps of superbolt distribution.”

“The average stroke energy over water is greater than the average stroke energy over land — we knew that,” Holzworth said. “But that’s for the typical energy levels. We were not expecting this dramatic difference.”

The time of year for superbolts also doesn’t follow the rules for typical lightning. Regular lightning hits in the summertime — the three major so-called “lightning chimneys” for regular bolts coincide with summer thunderstorms over the Americas, sub-Saharan Africa and Southeast Asia. But superbolts, which are more common in the Northern Hemisphere, strike both hemispheres between the months of November and February.

The reason for the pattern is still mysterious. Some years have many more superbolts than others: late 2013 was an all-time high, and late 2014 was the next highest, with other years having far fewer events.

“We think it could be related to sunspots or cosmic rays, but we’re leaving that as stimulation for future research,” Holzworth said. “For now, we are showing that this previously unknown pattern exists.”

Co-authors are research associate professor and senior research scientist at the UW; and and at the University of Otago in New Zealand. The research was funded by the UW.

For more information, contact Holzworth at bobholz@uw.edu or 206-685-7410.

A mapping lightning around the globe finds lightning strokes occur nearly twice as often directly above heavily-trafficked shipping lanes in the Indian Ocean and the South China Sea than they do in areas of the ocean adjacent to shipping lanes that have similar climates.

The difference in lightning activity can’t be explained by changes in the weather, according to the study’s authors, who conclude that aerosol particles emitted in ship exhaust are changing how storm clouds form over the ocean.

The study published Sept. 7 in Geophysical Research Letters is the first to show ship exhaust can alter thunderstorm intensity. The researchers conclude that particles from ship exhaust make cloud droplets smaller, lifting them higher in the atmosphere. This creates more ice particles and leads to more lightning.

The results provide some of the first evidence that humans are changing cloud formation on a nearly continual basis, rather than after a specific incident like a wildfire, according to the authors. Cloud formation can affect rainfall patterns and alter climate by changing how much sunlight clouds reflect to space.

“It’s one of the clearest examples of how humans are actually changing the intensity of storm processes on Earth through the emission of particulates from combustion,” said lead author , a UW professor of atmospheric sciences.

All combustion engines emit exhaust, which contains microscopic particles of soot and compounds of nitrogen and sulfur. These particles, known as aerosols, form the smog and haze typical of large cities. They also act as cloud condensation nuclei – the seeds on which clouds form. Water vapor condenses around aerosols in the atmosphere, creating droplets that make up clouds.

Cargo ships crossing oceans emit exhaust continuously and scientists can use ship exhaust to better understand how aerosols affect cloud formation.

Co-author , a former UW postdoctoral researcher who is now an atmospheric scientist at NASA Marshall Space Flight Center in Huntsville, Alabama, was analyzing data from the , a UW-based network of sensors that locates lightning strokes all over the globe, when she noticed a nearly straight line of lightning strokes across the Indian Ocean.

Virts and her colleagues compared the lightning location data to maps of ships’ exhaust plumes from a global database of ship emissions. Looking at the locations of 1.5 billion lightning strokes from 2005 to 2016, the team found nearly twice as many lightning strokes on average over major routes ships take across the northern Indian Ocean, through the Strait of Malacca and into the South China Sea, compared to adjacent areas of the ocean that have similar climates.

“All we had to do was make a map of where the lightning was enhanced and a map of where the ships are traveling and it was pretty obvious just from the co-location of both of those that the ships were somehow involved in enhancing lightning,” Thornton said.

Water molecules need aerosols to condense into clouds. Where the atmosphere has few aerosol particles – over the ocean, for instance – water molecules have fewer particles to condense around, so cloud droplets are large.

When more aerosols are added to the air, like from ship exhaust, water molecules have more particles to collect around. More cloud droplets form, but they are smaller. Being lighter, these smaller droplets travel higher into the atmosphere and more of them reach the freezing line, creating more ice, which creates more lightning. Storm clouds become electrified when ice particles collide with each other and with unfrozen droplets in the cloud. Lightning is the atmosphere’s way of neutralizing that built-up electric charge.

Ships burn dirtier fuels in the open ocean away from port, spewing more aerosols and creating even more lightning, Thornton said.

“It is the first time we have, literally, a smoking gun, showing over pristine ocean areas that the lightning amount is more than doubling,” said Daniel Rosenfeld, an atmospheric scientist at the Hebrew University of Jerusalem who was not connected to the study. “The study shows, highly unambiguously, the relationship between anthropogenic emissions – in this case, from diesel engines – on deep convective clouds.”

Other co-authors are , a UW professor of Earth and space sciences who directs lightning network, and , a research meteorologist at the UW’s Joint Institute for the Study of the Atmosphere and Ocean.

###

For more information, contact Thornton at joelt@uw.edu or 206-962-1430.

This was originally posted as a by the American Geophysical Union.