So-called “” now seem almost commonplace as floods, storms and fires continue to set new standards for largest, strongest and most destructive. But to categorize weather as a true 100-year event, there must be just a 1% chance of it occurring in any given year. The trouble is that researchers don’t always know whether the weather aligns with the current climate or defies the odds.

Traditional weather forecasting models run on energy-hogging supercomputers that are typically housed at large research institutions. In the past five years, artificial intelligence has emerged as a powerful tool for cheaper, faster forecasting, but most AI-powered models can only accurately forecast 10 days into the future. ��Still, longer-range forecasts are critical for climate science — and helping people prepare for seasons to come.

In a in AGU Advances, 91̽�� researchers used AI to simulate the Earth’s current climate and interannual variability for up to 1,000 years. The model runs on a single processor and takes just 12 hours to generate a forecast. On a state-of-the-art supercomputer, the same simulation would take approximately 90 days.

“We are developing a tool that examines the variability in our current climate to help answer this lingering question: Is a given event the kind of thing that happens naturally, or not?” said , a 91̽��professor of atmospheric and climate science.

Durran was one of the first to introduce AI into weather forecasting more than five years ago when he and former 91̽��graduate student partnered with Microsoft Research. Durran also holds a joint position as a researcher with California-based Nvidia.

“To train an AI model, you have to give it tons of data,” Durran said. “But if you break up the available historical data by season, you don’t get very many chunks.”

The most accurate global datasets for the daily weather go back to roughly 1979. Although there are plenty of days between then and now that can be used to train a daily weather forecast model, the same period contains fewer seasons.��This lack of historical data was perceived as a barrier to using AI for seasonal forecasting.

Counterintuitively, the Durran group’s latest contribution to forecasting, Deep Learning Earth SYstem Model, or DLESyM��, was trained for one-day forecasts, but still learned how to capture seasonal variability.

The model combines two neural networks: one representing the atmosphere and the other, the ocean. While traditional Earth-system models often join atmospheric and oceanic forecasts, researchers had yet to incorporate this approach into models powered by AI alone.

“We were the first to apply this framework to AI and we found out that it worked really well,” said lead author , a 91̽��graduate student in atmospheric and climate science. “We’re presenting this as a model that defies a lot of the present assumptions surrounding AI in climate science.”

Because the temperature of the sea surface changes slower than the air temperature, the oceanic model updates its predictions every four days, while the atmospheric model updates every 12 hours. Cresswell-Clay is currently working on adding a land-surface model to DLESyM.

“Our design opens the door for adding other components of the Earth system in the future,” he said, especially components that have been difficult to model in the past, such as the relationship between soil, plants and the atmosphere. Instead of researchers coming up with an equation to represent this complex relationship, AI learns directly from the data.

The researchers showcased the model’s performance by comparing its forecasts of past events to those generated by the four leading models from the sixth phase of the Coupled Model Intercomparison Project, or CMIP6, all of which run on supercomputers. Climate predictions of future climate from these models were key resources used in the last report from the .

DLESyM simulated tropical cyclones and the seasonal cycle of the Indian summer monsoon better than the CMIP6 models. In mid-latitudes, DLESyM captured the month-to-month and interannual variability of weather patterns at least as well as the CMIP6 models.

For example, the model captured atmospheric “blocking” events just as well as the leading physics-based models. Blocking refers to the formation of atmospheric ridges that keep regions hot and dry, and others cold or wet, by deflecting incoming weather systems. “A lot of the existing climate models actually don’t do a very good job capturing this pattern,” Cresswell-Clay said. “The quality of our results validates our model and improves our trust in its future projections.”

Neither the CMIP6 models nor DLESyM are 100% accurate, but the fact that the AI-based approach was competitive while using so much less power is significant.

“Not only does the model have a much lower carbon footprint, but anyone can download it from our website and run complex experiments, even if they don’t have supercomputer access,” Durran said. “This puts the technology within reach for many other researchers.”

Other authors include , a visiting 91̽��doctoral student in atmospheric and climate science; a 91̽��doctoral student in atmospheric and climate science; , a 91̽��doctoral student in atmospheric and climate science; Raúl A. Moreno, a doctoral student in atmospheric and climate science and , a postdoctoral researcher in neuro-cognitive modeling at the University of Tübingen in Germany.

This work was funded by the U.S. Office of Naval Research, the U.S. Department of Defense, the University of Chinese Academy of Sciences, the National Science Foundation of China, Deutscher Akademischer Austauschdienst, International Max Planck Research School for Intelligent Systems, Deutsche Forschungsgemeinschaft, U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research and the NVIDIA Applied Research Accelerator Program.

For more information, contact Nathaniel Cresswell-Clay at nacc@atmos.washington.edu or Dale Durran at drdee@uw.edu

91̽�� atmospheric scientist Lynn McMurdie has led campaigns to measure rain and snowfall in places ranging from Washington’s Olympic Peninsula to Argentina to the Eastern U.S. Now she’s among the leaders of a field campaign in Colorado to better understand and forecast snowfall in the mountains of the Western U.S.

A scientific expedition this coming winter in Colorado’s Yampa Valley will improve forecasts of snowfall and estimates of how climate change will impact snowpack and water availability in mountainous regions of the West.

, a research professor of atmospheric sciences at the UW, is one of the principal investigators on the effort, with a $4.8 million grant from the National Science Foundation and led by the University of Michigan. Other participating institutions include the University of Wisconsin, the University of Utah, Colorado State University and Stony Brook University.

The Snow Sensitivity to Clouds in a Mountain Environment experiment, or S2noCliME, will use several radars and snow-sampling instruments to measure the size and shape of snowflakes and aerosol particles. The resulting data will help estimate water availability in the Yampa Valley in northwest Colorado. This area on the northwest side of the Rocky Mountains feeds the Yampa River, the largest free-flowing tributary of the Colorado River. Like Western Washington, it relies on melting snow for its summer water supplies, and faces drought and wildfire risk when these snow reservoirs are lower than normal.

“We hope that our data will ultimately improve winter storm forecasts and tell Western cities when to expect a drought because of insufficient snowpack,” said lead investigator��, an assistant professor at the University of Michigan.

The team will deploy instruments in the Steamboat Springs region of Colorado’s Park Range, a section of the Rockies that extends from southern Wyoming to northwestern Colorado and is poorly covered by the National Weather Service radar network.

Today’s models of snowfall often struggle in these types of mountainous areas.

“In the West, our forecasting models and satellite estimates of precipitation really underpredict snowfall and often don’t get the distribution right,” McMurdie said. “The western U.S. depends on snowpack accumulated during the winter for summer water supply for agriculture, fisheries and municipal water sources. Accurate snowfall prediction and understanding the underlying processes producing mountain snow are critical in order to provide guidance on water availability.”

Collecting data over an entire winter will provide the statistical power necessary to more accurately predict snowpack after winter storms and over longer timescales. Data will also be shared directly with the , a working group that brings together scientists and local water managers.

By combining snow-sampling instruments with radars that indirectly study snow, the researchers will overcome a major challenge of using radar: It’s hard to connect the reflected radar signal with the size, shape and number of snowflakes, which determines the amount of water in the snow.

The team’s radars at will use multiple frequencies for detecting snowflakes of different sizes. At the same time, the team will deploy a portable in Hayden, Colorado. This radar’s view toward and the surrounding 62 miles will be combined with the National Weather Service’s radars to see how the strength of the storms change as they move from the west toward the Rockies.

The results will be compared with those from earlier projects in the Pacific Northwest, such as the UW-led project.

“We expect to have some similarities, such as how upstream conditions affect the snow-producing cloud processes during this campaign, as well as some differences since the Colorado region is much drier and more interior than the Olympic Mountains or the Cascades,” McMurdie said.

, an affiliate 91̽��faculty member and former 91̽��postdoctoral researcher who is now a faculty member at the University of Wisconsin-Madison is another of the project’s lead investigators. , research associate professor of atmospheric sciences at the UW, a 91̽��postdoctoral researcher and one or more 91̽��graduate students will also participate in the campaign.

The first instruments will arrive at Mt. Werner this summer, and the team will begin collecting measurements in December.

For more information, contact McMurdie at lynnm@uw.edu. This article was adapted from a University of Michigan .

“For me it was an inconvenience, but for some people it could be life-threatening,” said , now an assistant professor in the 91̽��’s Department of Environmental and Occupational Health Sciences. “If you had an uncle that had an electric heart pump, basically, his heart wouldn’t work without power. You could use a backup battery for eight hours, but after that, if you don’t have access to electricity, you have to go to the emergency room. This is a really dangerous situation.”

Years later, Casey has answers. April 29 in the journal Nature Communications analyzed three years of power outages across the U.S., finding that Americans already bearing the brunt of climate change and health inequities are clustered in four regions — Louisiana, Arkansas, central Alabama and northern Michigan — and that they are most at risk of impact by a lengthy blackout.

The findings could help shape the future of local energy infrastructure, especially as climate change intensifies and the American power grid continues to age. Last year’s Inflation Reduction Act included billions of dollars to revamp energy systems, and Casey hopes federal agencies will consult the newly published findings to target energy upgrades.��

The study is the first county-level analysis of power outages, which the federal government reports only at the state level. That poses a problem for researchers: a federally reported outage in Washington state could occur in Seattle, Spokane, or somewhere in between, making it difficult to understand specifically which population is affected.��

Casey and her team found that between 2018 and 2020, more than 231,000 power outages lasting more than an hour occurred nationwide. Of those, 17,484 stretched at least eight hours —��a duration widely viewed as medically relevant.

Most counties that experienced an electrical outage had at least one event lasting more than eight hours. These counties were most concentrated in the South, Northeast and Appalachia.

Next, researchers looked at how power outages overlapped with severe weather. They wanted to know which weather events are most likely to cause an outage, and which parts of the U.S. are most often hit with a blackout-causing storm.

They found that heavy precipitation in a given area makes a power outage five times more likely. Tropical cyclones, storms with high winds that originate over tropical oceans, make a power outage 14 times more likely. And a tropical cyclone with heavy precipitation on a hot day — like the hurricanes that each fall hit the Gulf Coast? They make power outages 52 times more likely.

“We look at weather reports and decide whether or not to bring an umbrella or stay home,” Casey said. “But thinking about being prepared for an outage when one of these events is rolling through is a new element to consider.”

Then came questions of equity. Incorporating a combination of socioeconomic and medical factors, Casey’s team identified communities that would likely be especially vulnerable during a long power outage. Using that data, the researchers were able to identify communities that experienced both high social vulnerability and frequent power outages.����

A map of those counties shows a bright cluster in Louisiana and Arkansas, with more clusters in central Alabama and northern Michigan. In those places especially, the country’s inevitable change in energy infrastructure provides the greatest opportunity to improve public health.

“Any time we can identify another factor that we can intervene on to get closer to health equity, it’s exciting,” Casey said. “I think we’re going to see tremendous change, especially in the way our energy systems are set up, in the next couple decades. It’s this huge opportunity to get equity into every conversation and talk about what we’re going to do to make two decades from now look different from where we are.”��

This study began while Casey was a professor in Columbia University’s Mailman School of Public Health. Other authors are Vivian Do (first author), Heather McBrien, Nina Flores, Alexander Northrop and Jeffrey Schlegelmilch at Columbia University and Mathew Kiang at Stanford University. The research was funded by the National Institute on Aging and the National Institute of Environmental Health Sciences. ��

For more information, contact Casey at jacasey@uw.edu.��

A warming climate is causing a decline in sea ice in the Arctic Ocean, where loss of sea ice has important ecological, economic and climate impacts. On top of this long-term shift due to climate change are weather events that affect the sea ice from week to week.

The strongest Arctic cyclone ever observed poleward of 70 degrees north latitude struck in January 2022 northeast of Greenland. A new analysis led by the 91̽�� shows that while weather forecasts accurately predicted the storm, ice models seriously underestimated its impact on the region’s sea ice.

The , published in October in the Journal of Geophysical Research–Atmospheres, suggests that existing models underestimate the impact of big waves on ice floes in the Arctic Ocean.

“The loss of sea ice in six days was the biggest change we could find in the historical observations since 1979, and the area of ice lost was 30% greater than the previous record,” said lead author , a research assistant professor of atmospheric sciences at the UW. “The ice models did predict some loss, but only about half of what we saw in the real world.”

Accurate sea ice forecasts are important safety tools for Northern communities, mariners and others operating in Arctic waters. The accuracy of forecasts in the Arctic Ocean also has broader effects.

“The skill of a weather forecast in the Arctic affects the skill of weather forecasts in other places,” Blanchard-Wrigglesworth said.

The January 2022 cyclone had the lowest pressure center estimated since satellite records began in 1979 above 70 degrees north. It was an extreme version of a typical winter storm. Climate change doesn’t appear responsible for the cyclone: The researchers didn’t find a trend in the strength of intense Arctic cyclones since 1979, and sea ice area was close to the historical normal for that region before the storm hit.

Waves travel through sea ice in the Arctic Ocean, as seen from a ship in October 2015. Credit: Ed Blanchard-Wrigglesworth/91̽��

During the storm, record winds howled over the Arctic Ocean. The waves grew to 8 meters (26 feet) tall in open water and remained surprisingly strong as they traveled through the sea ice. The ice heaved 2 meters (6 feet) up and down near the edge of the pack, and NASA’s ICESat-2 satellite shows that the waves reached as far as 100 kilometers (60 miles) toward the center of the ice pack.

Six days after the storm struck, the sea ice had thinned significantly in the affected waters north of Norway and Russia, in places losing more than half a meter (about 1.5 feet) of thickness.

“It was a monster storm, and the sea ice got pummeled. And the sea ice models didn’t predict that loss, which suggests there are ways we could improve the model physics,” said second author , a research assistant professor at the University of Alaska Fairbanks. She begins a research position at the 91̽��Applied Physics Laboratory in the new year.

The new analysis shows that the atmospheric heat from the storm had a small effect, meaning some other mechanism was to blame for the ice loss. Possibilities, Blanchard-Wrigglesworth suggests, include sea ice that was thinner before the storm hit than models had estimated; that the storm’s waves broke up ice floes more forcefully than models predicted as they penetrated deep into the ice pack; or that waves churned up deeper, warmer water and brought it into contact with the sea ice, melting the ice from below.

The unexpected ice loss, despite an accurate storm forecast, suggests that this is an area where models could improve. The researchers hope to monitor future storms to pinpoint exactly what led to the dramatic sea ice loss, potentially by placing sensors in the path of a future approaching storm.

While this storm doesn’t appear to be linked to climate change, the increase of open water as sea ice melts is allowing for larger waves that are eroding Arctic coastlines. Those waves, researchers said, could also affect the remaining sea ice pack.

“Going into the future, this is something to keep in mind, that these extreme events might produce these episodes of huge sea ice loss,” Blanchard-Wrigglesworth said.

Other co-authors are at NASA, at NASA and the University of Maryland and at the University of Auckland and Brown University. The research was funded by NASA, the U.S. Navy’s Office of Naval Research and Schmidt Futures.

For more information, contact Blanchard-Wrigglesworth at ed@atmos.uw.edu or Webster at mwebster3@alaska.edu.

Grants: NASA: 80NSSC20K0922, 80NSSC20K0959, ONR-DRI: N00014-21-1-2490

Forecasters are predicting a “” this year. This will be the third winter in a row that the Pacific Ocean has been in a La Niña cycle, something that’s happened only twice before in records going back to 1950.

New research led by the 91̽�� offers a possible explanation. The , recently published in Geophysical Research Letters, suggests that climate change is, in the short term, favoring La Niñas.

“The Pacific Ocean naturally cycles between El Niño and La Niña conditions, but our work suggests that climate change could currently be weighing the dice toward La Niña,” said lead author , a 91̽��research scientist in atmospheric sciences. “At some point, we expect anthropogenic, or human-caused, influences to reverse these trends and give El Niño the upper hand.”

Scientists hope to predict the direction of these longer-term El Niño-like or La Niña-like climate trends in order to protect human life and property.

“This is an important question over the next century for regions that are strongly influenced by El Niño, which includes western North America, South America, East and Southeast Asia and Australia,” Wills said.

El Niño and La Niña events have , affecting patterns of rainfall, flooding and drought around the Pacific Rim. A winter tends to be cooler and wetter in the Pacific Northwest and hotter and drier in the U.S. Southwest. Other worldwide effects include drier conditions in East Africa, and rainier weather in Australia, Indonesia, Malaysia and the Philippines.

Knowing what to expect in the future helps communities prepare for potential weather in the coming season and in years to come.

Global warming is widely expected to favor El Niños. The reason is that the cold, deep water rising to the sea surface off South America will meet warmer air. Anyone who’s sweated knows that evaporation has a cooling effect, so the chillier ocean off South America, which has less evaporation, will warm faster than the warmer ocean off Asia. This decreases the temperature difference across the tropical Pacific and lightens the surface winds blowing toward Indonesia, the same as occurs during El Niño. Past climate records confirm that the climate was more El Niño-like during warmer periods.

But while Earth’s atmosphere has warmed in recent decades, the new study shows a surprising trend in the tropical ocean. The authors looked at temperatures at the surface of the ocean recorded by ship-based measurements and ocean buoys from 1979 to 2020. The Pacific Ocean off South America has actually cooled slightly, along with ocean regions farther south. Meanwhile, the western Pacific Ocean and nearby eastern Indian Ocean have warmed more than elsewhere. Neither phenomenon can be explained by the natural cycles simulated by climate models. This suggests that some process missing in current models could be responsible.

The upshot of these changes on either side of the tropical Pacific is that the temperature difference between the eastern and western Pacific has grown, surface winds blowing toward Indonesia have strengthened, and people are experiencing conditions typical of La Niña winters. The study focuses on temperature patterns at the ocean’s surface. Thirty years of data is too short to study the frequency of El Niño and La Niña events.

“The climate models are still getting reasonable answers for the average warming, but there’s something about the regional variation, the spatial pattern of warming in the tropical oceans, that is off,” Wills said.

The researchers aren’t sure why this pattern is happening. Their current work is exploring tropical climate processes and possible links to the ocean around Antarctica. Once they know what’s responsible, they may be able to predict when it will eventually switch to favor El Niños.

“If it turns out to be natural long-term cycles, maybe we can expect it to switch in the next five to 10 years, but if it is a long-term trend due to some processes that are not well represented in the climate models, then it would be longer. Some mechanisms have a switch that would happen over the next few decades, but others could be a century or longer,” Wills said.

The study was conducted before this year’s potential triple La Niña was announced. But Wills is cautious about declaring victory.

“These year-to-year changes are very unpredictable and it’s important not to get too hung up on any individual year — it doesn’t add a lot of statistical weight,” Wills said. “But I think it’s something that we should watch for in the next few years.”

Co-authors of the study are and at the UW; , a postdoctoral researcher at the Lamont-Doherty Earth Observatory who did the work as part of her 91̽��doctoral research; and at the University of Illinois at Urbana-Champaign. The study was funded by the National Science Foundation, the National Oceanic and Atmospheric Administration and the Alfred P. Sloan Foundation.

For more information, contact Wills at rcwills@uw.edu. Note: Wills is currently based in Colorado.

, a 91̽��professor of atmospheric sciences, was traveling to a conference in Boston as Hurricane Ian approached the Gulf of Mexico. During breaks at the conference, she provided her thoughts on the closely watched catastrophic storm system that made landfall in Florida on Sept. 28.

Q: What are your thoughts on Hurricane Ian? How does it compare to other storms?

Shuyi Chen: Each hurricane is unique in many ways. However, Hurricane Ian has some similarity to Hurricane Irma in 2017 in terms of landfall location on the southwest coast of Florida. Both hurricanes caused large storm surges — about 10 feet of water — near Naples and Fort Myers with strong onshore winds, while the water got “sucked” out of Tampa Bay by offshore winds associated with the hurricane. Ian is a larger storm than Irma, in terms of size of strong wind and rain areas.

Q: What are your thoughts on this year’s Atlantic hurricane season, with Hurricane Fiona and now Hurricane Ian both expected to have severe impacts?

SC: The overall Atlantic hurricane season this year has been slower, with less named storms and generally weaker and shorter-lived storms than the recent average, especially compared with the last two busy years in 2020 and 2021. However, hurricane impacts on society are not necessarily correlated the number of storms. It depends on individual storms and where they occur. Land-falling hurricanes, like Fiona and Ian, have very high impacts, including extreme wind, rain, storm surge and flooding. In fact, one of the most costly events, Hurricane Andrew in 1992, occurred during a year when we had one of the lowest number of storms.

Q: You are working with the National Oceanic and Atmospheric Administration to use Saildrone observations. What do you hope to learn from this new technology?

SC: Saildrones have transformed the way we observe hurricanes. Because of the extreme wind, rain, and ocean waves in hurricanes, in the past, we have not be able to get data inside the hurricane at the surface where hurricanes get energy to fuel themselves. Saildrones were able to collect valuable data in Hurricane Sam for the first time in 2021, and repeated its success in Hurricane Fiona in 2022. During Hurricane Ian, Saildrones and other technologies, including one developed at the UW, are collecting data in the hurricane environment, which will help us compare the data both in and outside of the storm. This will help develop new coupled models and evaluate model results.

Q: As a researcher, what aspect of this hurricane will you be looking at most closely?

SC: My group’s current research focuses on better understanding, observing and predicting hurricane impacts, especially coastal and inland flooding when storms make landfall. We work to develop next-generation coupled models that combine atmosphere, ocean and waves for hurricane prediction. We also collect observations in the atmosphere, ocean, and at the air-sea interface for large ocean waves.

Q: Any other thoughts?

SC: The complexity of forecasting hurricane impacts, such as flooding caused by a combination of rain, storm surge, river runoff and the built environment in urban and coastal cities, requires new interdisciplinary research involving scientists, engineers and decision-makers. We need to advance multidisciplinary research, technologies and training of new generation of scientists to take on this grand challenge.

For more information, contact Chen at shuyic@uw.edu. ��

Jacob Davis, a 91̽��doctoral student in civil and environmental engineering, (right) releases a wave-monitoring sensor from a U.S. Navy aircraft on Sept. 26 off the coast of Florida. Data from this instrument developed at the 91̽��Applied Physics Laboratory will be combined with other observations to try and improve hurricane forecasts around the world.

Researchers dropped technology developed at the 91̽�� off the coast of Florida on Monday to measure ocean waves in the path of Hurricane Ian. The test is one part of a broad effort to improve forecasts for these fast-moving and deadly systems.

The team, including , a 91̽��doctoral student in civil and environmental engineering, and members of the U.S. Navy’s VXS-1 Squadron deployed the devices in the path of Hurricane Ian, before the hurricane made landfall. The five instruments developed at the 91̽��are now sending back data that can be viewed on this .

The UW-built sensors are known as the , or SWIFTs. For this project, the team used a smaller version, known as microSWIFTs. The sensors can drift with the waves to gather detailed measurements of waves and currents at the ocean’s surface. Past deployments used the sensors to study waves in the changing Arctic Ocean and near potential sites for marine turbines.

The current effort in the path of Hurricane Ian aims to understand how the extreme low-pressure storm system affects the ocean and, ultimately, coastal areas.

“The goal is to understand the details of wave generation in hurricanes, which are unique in how fast they move and how strong the winds are. This causes rapid wave evolution that’s not well described by current forecast models,” said Jim Thomson, an oceanographer at the 91̽��Applied Physics Laboratory and a 91̽��professor of civil and environmental engineering. “The end goal is to improve the forecasts for when and where waves will impact the coasts, including storm surge.”

Researchers emphasize that the deployment is part of a . The microSWIFT observations at the ocean’s surface will be combined with other observations, including technologies deployed on the same flight by Scripps Institution of Oceanography and Sofar Ocean Technologies.

This work was done with the National Oceanographic Partnership Program’s Hurricane Coastal Impacts program with supporting flights by the U.S. Navy’s Scientific Development Squadron. The research was funded by the National Oceanographic Partnership Program and managed by the U.S. Navy’s Office of Naval Research.

For more information contact Davis at davisjr@uw.edu or Thomson at jthomson@apl.washington.edu. (Note: Thomson is on a ship in the Arctic and his responses will be delayed)

Photo and video available .

Conditions in the tropical ocean affect weather patterns worldwide. The most well-known examples are El Niño or La Niña events, but scientists believe other key elements of the tropical climate remain undiscovered.

In a recently published in Geophysical Research Letters, scientists from the 91̽�� and NOAA’s Pacific Marine Environmental Laboratory use remotely-piloted sailboats to gather data on , or pockets of cooler air that form below tropical storm clouds.

“Atmospheric cold pools are cold air masses that flow outward beneath intense thunderstorms and alter the surrounding environment,” said lead author , a postdoctoral researcher at the Cooperative Institute for Climate, Ocean and Ecosystem Studies. “They are a key source of variability in surface temperature, wind and moisture over the ocean.”

The paper is one of the first tropical Pacific studies to rely on data from Saildrones, wind-propelled sailing drones with a tall, hard wing and solar-powered scientific instruments. Co-authors on the NOAA-funded study are at CICOES and at NOAA.

Atmospheric cold pools produce dramatic changes in air temperature and wind speed near the surface of the tropical ocean. The pockets of cooler air form when rain evaporates below thunderstorm clouds. These relatively dense air masses, ranging between 6 to 125 miles (10 to 200 kilometers) across, lead to downdrafts that, upon hitting the ocean surface, produce temperature fronts and strong winds that affect their surroundings. How this affects the larger atmospheric circulation is unclear.

“Results from previous studies suggest that cold pools are important for triggering and organizing storm activity over tropical ocean regions,” Wills said.

To understand the possible role of cold pools in larger tropical climate cycles, scientists need detailed measurements of these events, but it is hard to witness an event as it happens. The new study used uncrewed surface vehicles, or USVs, to observe the phenomena.

Over three multi-month missions between 2017 and 2019, 10 USVs covered over 85,000 miles (137,000 kilometers) and made measurements of more than 300 cold pool events, defined as temperature drops of at least 1.5 degrees Celsius in 10 minutes. In one case, a fleet of four vehicles separated by several miles captured the minute-by-minute evolution of an event and revealed how the cold pool propagated across the region.

“This technology is exciting as it allows us to collect observations over hard-to-reach, under-sampled ocean regions for extended periods of time,” Wills said.

The paper includes observations of air temperature, wind speed, humidity, air pressure, sea surface temperature and ocean salinity during cold pool events. The authors use the data to better describe these phenomena, including how much and how quickly air temperatures drops, how long it takes the wind to reach peak speeds, and how sea surface temperature changes nearby. Results can be used to evaluate mathematical models of tropical convection and explore more questions, like how the gusts created by the temperature difference affect the transfer of heat between the air and ocean.

For more information, contact Wills at smwills@uw.edu. Parts of this post were adapted from an in AGU Eos.

Lightning strikes in the Arctic tripled from 2010 to 2020, a finding 91̽�� 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 91̽��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 United States experiences more tornadoes than any other country, with a season that peaks in spring or summer depending on the region. Tornadoes are often deadly, especially in places where buildings can’t withstand high winds.

Accurate advanced warnings can save lives. A study from the 91̽�� and the National Oceanic and Atmospheric Administration describes a new way to rate and possibly improve tornado warnings. It finds that nighttime twisters, summer tornadoes and smaller events remain the biggest challenges for the forecasting community.

“This new method lets us measure how forecast skill is improving, decreasing or staying the same in different situations,” said , a 91̽��assistant professor of atmospheric sciences. “The tornado forecasting community needs to know what we’re doing best at, and where we can focus training and research in the future.”

She is lead author of the published online in December in the Bulletin of the American Meteorological Society.

Though the southern and central U.S. see the most tornadoes, can experience twisters. ��Scientific understanding of tornadoes is biased toward populated places, Anderson-Frey said, where people are more likely to observe and report the events.

“As population density increases in different areas, including outside the U.S., I think we’re getting more of an idea of the range of environments in which tornadoes can actually form,” Anderson-Frey said.

The paper develops a new method to rate the skill of a tornado warning based on the difficulty of the environment. It then evaluates thousands of tornadoes and associated warnings over the continental United States between 2003 and 2017.

The NOAA-funded study finds that nighttime tornadoes have a lower probability of detection and a higher false-alarm rate than the environmental conditions would suggest. Summertime tornadoes, occurring in June, July or August, also are more likely to evade warning.

The nighttime events may be harder to forecast because the absence of daytime warming makes the conditions less favorable, and because there were fewer eyewitness reports, Anderson-Frey said. Summer events may be more difficult because summer has more relatively weak tornadoes that occur in marginal environments, meaning on the edge of conditions that produce a tornado.

Larger events — those rated 2 or above on the — actually generated better warnings than expected for the conditions. The results can inform how research, training or observational technology could improve future tornado warnings.

“The forecasting community is not just looking at the big, photogenic situations that will crop up in the Great Plains. We’re looking at tornadoes in regions where vulnerability is high, including in regions that don’t normally get tornadoes, where by definition the vulnerability is high,” Anderson-Frey said.

“There’s a real effort in the forecasting research community to bring in the human element — being able to identify where we can do the most good.”

Although tornado forecasts and warnings are improving overall, so are some types of risk. Populations are growing and moving into new, remote environments. Mobile or manufactured homes without anchored foundations are less able to withstand high winds.

“What really excites me about this work is the opportunity to look at performance by how difficult the warning situation was,” said co-author at NOAA’s National Severe Storm Laboratory in Norman, Oklahoma. “We have the chance to measure improvement through the years taking into account that some situations and years may be harder or easier forecasts.”

Anderson-Frey She moved from Oklahoma to join the 91̽��faculty in 2019. In related research, she is now analyzing data for past tornadoes to determine the environmental conditions that can produce events in unexpected places, like the that struck Port Orchard in Washington state.

“I’m working on applying a machine-learning technique to studying what prototypical tornadic environments look like in different parts of the United States,” she said.

For more information, contact Anderson-Frey at akaf@uw.edu.

NOAA grant: #NA16OAR432011