EOS

Wastewater from agricultural runoff and human waste contains nutrients such as nitrogen and phosphorus, which can pollute natural waters and cause harmful algal blooms. These nutrients may persist at low levels even after wastewater has been biologically and chemically treated.

In southern Florida, many wastewater treatment facilities inject the treated water into wells 60–120 feet deep. As the water percolates underground, it is diluted by rainwater and groundwater. Chemical interactions with Key Largo Limestone—the rock that makes up the subsurface in the upper and middle Keys—and with microbes living in the groundwater filter out residual nutrients, such as the nitrogen-bearing compounds nitrate and nitrite.

“The idea is that the microbial nitrogen cycle will transform nitrate and nitrite to ammonium and nitrogen gas, which is then released to the atmosphere, before the effluent reaches coastal waters,” explained Miquela Ingalls, a sedimentary geochemist at Pennsylvania State University (Penn State). Similarly, phosphorus is removed from wastewater via a chemical reaction that attaches it onto the limestone bedrock.

But a new study, coauthored by Ingalls, shows that shallow injection doesn’t entirely eliminate nitrogen, raising the possibility that the nutrient may be contaminating coastal ecosystems.

The study’s authors analyzed water from monitoring wells near a wastewater injection site in the Florida Keys. They found that nitrogen and phosphorus levels decreased as water moved away from the site, but were still detectable at a depth of 6 meters (20 feet) and a distance of 350 meters (~1,150 feet) from the injection site, close to the Florida Bay shoreline.

In an earlier study targeting phosphorus, also in the Florida Keys, Ingalls and other researchers at Penn State concluded that up to 10% of initially present soluble reactive phosphorus (the form of phosphorus that can be directly taken up by plants) remained in injected water and was ultimately discharged into the ocean.

The new study aimed to determine how effectively shallow injection eliminates nitrogen.

Some Nitrogen Persists

The study focused on a wastewater injection site in Marathon, located on Vaca Key. The city of Marathon currently pumps treated wastewater between 18 and 27 meters (59–89 feet) into the underlying Key Largo Limestone. In 2021 and 2022, Penn State scientists installed nine monitoring wells 3–27 meters (10–89 feet) deep near the injection site. The researchers then measured levels of nitrate, dissolved nitrogen, and other chemicals at the monitoring wells from 2021 to 2023 and compared them with the levels found in the injected wastewater.

Nitrate levels were elevated at a monitoring well 350 meters (~1,150 feet) from the injection site and close to the Florida Bay.

At most wells, nitrate was completely eliminated from wastewater 2 weeks after injection. However, nitrate levels remained elevated 3–6 meters (10–20 feet) deep in a monitoring well 350 meters (~1,150 feet) north of the injection site and close to the Florida Bay.

This monitoring well is farthest from the injection site and is probably in the path of injected wastewater, according to the previous study, which found that wastewater injected in Marathon mostly travels north and east. As it travels, the water rises toward the surface. This may explain why nitrate is elevated at the most distant well: Close to the injection site, wastewater remains at its initial depth, but after traveling hundreds of meters, it has risen far enough to contaminate shallow groundwater. The contaminated well’s location on the Florida Bay shoreline suggests that along with phosphorus, some wastewater-derived nitrogen may be washing out to sea.

The authors attributed the imperfect filtering of nutrients to the unique setting of the Florida Keys. For wastewater injection, timing is important: The longer wastewater remains underground, the more time there is for microbial and chemical processes to filter out contaminants. But the Keys are mostly composed of small, narrow islands, so the injected wastewater doesn’t travel very far before reaching the ocean.

Groundwater in the Keys also mixes with seawater, making an especially salty and dense mixture. The injected wastewater has a lower density, causing it to buoy up toward the surface and limit the time spent underground.

Nitrogen Pollution Harms Coastal Ecosystems

The discharge of nitrogen into the Florida Bay may have consequences for marine life. The researchers found total nitrogen concentrations of 18 micromoles per kilogram just offshore of the contaminated monitoring well, surpassing the local threshold of 16.1 micromoles per kilogram defined by the Florida Department of Environmental Protection.

Such pollution from wastewater and other human activities is harmful to wildlife, said Brian Lapointe, a marine scientist at Florida Atlantic University who was not involved in the research, because “increased nutrient concentrations support not only algal blooms and microbial pathogens, including coral diseases, but also myriad water quality problems.”

“Nutrient pollution from shallow injection wells has been a major local pollution source driving eutrophication in coastal waters of the Florida Keys for decades.”

“Nutrient pollution from shallow injection wells has been a major local pollution source driving eutrophication in coastal waters of the Florida Keys for decades,” Lapointe said. But both Lapointe and Ingalls highlighted that nutrient pollution in certain parts of the Florida Keys has decreased in recent years, largely due to advances in wastewater treatment. “The effluent being injected into the subsurface starts with a lower concentration of nutrients, so there is less to remediate by biological and chemical processes within the Key Largo Limestone,” Ingalls said.

Nonetheless, Lapointe recommended eliminating shallow injection of wastewater as a way to reduce nutrient pollution. The city of Marathon is set to phase out shallow injection following a 2022 lawsuit filed by the environmental group Friends of the Lower Keys (FOLKs). Instead, the city will transition to deep well injection, which is used in other parts of the Florida Keys. Deep wells inject wastewater more than 2,000 feet underground, lowering the chances that wastewater will rise to the surface before microbes and chemical reactions can filter out contaminants.

—Caroline Hasler (@carbonbasedcary), Science Writer

Citation: Hasler, C. (2025), Shallow injection imperfectly filters Florida wastewater, Eos, 106, https://doi.org/10.1029/2025EO250357. Published on 26 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Editors’ Highlights are summaries of recent papers by AGU’s journal editors. Source: Journal of Geophysical Research: Solid Earth

Seismologists have long detected unusual structures deep beneath continents at mid-lithospheric depths (80–120 kilometers), but their cause has remained uncertain.

In a new study, Zhang et al. [2025] use state-of-the-art computer simulations that combine first-principles (or fundamental assumption) calculations with machine learning to discover a new form of calcium carbonate, an important carbon-bearing mineral in Earth’s deep interior. This newly identified phase undergoes remarkable elastic softening under mid-lithospheric conditions, greatly reducing seismic wave speeds. Even trace amounts of such carbonate could explain the puzzling seismic signals and anomalous electrical properties observed beneath ancient continental regions.

These findings suggest that carbonates play a far more important role in shaping continental structure than previously recognized. Moreover, the results demonstrate that advanced computational methods can uncover unexpected aspects of the deep carbon cycle and the long-term stability of continental roots.

Citation: Zhang, P., Man, L., Yuan, L., Wu, X., & Zhang, J. (2025). Ultra-low-velocity disordered CaCO3 may explain mid-lithospheric discontinuities. Journal of Geophysical Research: Solid Earth, 130, e2025JB031906. https://doi.org/10.1029/2025JB031906

—Jun Tsuchiya, Editor, JGR: Solid Earth

Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

body {background-color: #D2D1D5;} Research & Developments is a blog for brief updates that provide context for the flurry of news that impacts science and scientists today.

Unchecked greenhouse gas emissions could cause the world’s income to fall by nearly a quarter within the century, projects a new study published in PLOS Climate.

“Climate change reduces income in all countries, hot and cold, rich and poor alike.”

“Climate change reduces income in all countries, hot and cold, rich and poor alike,” the study’s authors wrote in a press release.

Gross domestic product, or GDP, is the total value of goods and services produced in a given year. A country’s per-capita GDP is a measure of the average income of a person living there. 

In the new study, researchers turned to information about the previously studied link between rising temperatures and GDP, along with possible scenarios of future warming, called Shared Socioeconomic Pathways. They used this data to simulate more precisely how rising temperatures might impact global GDP over time.

Their projections show warming could cause widespread economic losses as higher temperatures and climate variability impact the activities of industries including agriculture and manufacturing. But reducing greenhouse gas emissions could soften the effect, the authors write. 

In a moderate emissions scenario (SSP2-4.5), global GDP decreased by about 2.5% by 2100. In a high-emissions scenario with minimal adaptation (SSP3-7.0), the projections showed global per-capita GDP dropping by up to 11%. And in a more extreme emissions scenario (SSP5-8.5) without any climate mitigation or adaptation, the researchers projected per-capita income losses of up to 24%.

Under this extreme scenario, the United States would lose nearly a third of its per-capita GDP by 2100. 

The estimated changes in global income compared to the 1960-2014 warming trend, as well as a world without climate change, varied depending on the warming scenario that the researchers used. Credit: Mohaddes and Raissi, 2025, https://doi.org/10.1371/journal.pclm. 0000621

The results are in line with previous research from 2024, which indicated a likely decrease in global income by 11% to 29%, depending on future emissions scenarios. However, the new study finds that if the world’s governments abide by Paris Agreement goals—that is, limiting temperature increases to 0.01°C (0.02°F) per year—global GDP could slightly increase by 0.25% by 2100.

 
Related

•  Read the paper: “Rising temperatures, melting incomes: Country-specific macroeconomic effects of climate scenarios”
•  Climate Change Is Likely to Slash Global Income
• Get Involved: AGU Science Policy Action Center
 

In both the new study and previous research, low-income and hotter countries fared the worst: In the new study’s projections, countries located in hot climates and classified as low-income faced income losses between 30% and 60% of the global average. 

The publication comes on the heels of a speech at the United Nations General Assembly in which President Trump called climate change the “greatest con job ever perpetrated on the world.” That denialism won’t preserve global income, according to the authors: “Urgent action is needed to address climate change and protect economies from further income losses,” they wrote.

—Grace van Deelen (@gvd.bsky.social), Staff Writer

These updates are made possible through information from the scientific community. Do you have a story about science or scientists? Send us a tip at eos@agu.org. Text © 2025. AGU. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Editors’ Highlights are summaries of recent papers by AGU’s journal editors. Source: Community Science

This is an authorized translation of an Eos article. 本文是Eos文章的授权翻译。

在中国，通过大规模耕作实现农业现代化的努力，将许多参与了全国大部分粮食生产的小农户推到了边缘。一个颇具前景的解决方案是“循环农业”，它通过鼓励大规模和小规模农业经营之间的合作，重点关注可持续性、生产力和农村经济发展。

在Community Science期刊关于可持续农业跨学科合作的特刊中，Li 和 Nielsen[2025]研究了中国西南部一个将柚子种植与养猪相结合的循环农业项目。作者对小农户、政府官员、金融机构员工以及其他各类利益相关者进行了35次访谈，了解了该模式面临的各种利益和风险。

他们的研究结果表明，地方政府在搭建合作平台方面发挥着关键作用，而农业合作社则是商务管理的核心。该研究还揭示了政府参与往往出于政治动机，小农户在决策过程中可能失去自主权和公平代表权等挑战。作者建议，要使循环农业真正惠及每个人，小农户需要有话语权和权力来决定自己的未来，而不能让自己的利益受到侵犯。

Citation: Li, H., & Nielsen, J. Ø. (2025). Smallholders, capital, and circular agriculture—The case of combined pomelo and pig farming in southwest China. Community Science, 4, e2025CSJ000127. https://doi.org/10.1029/2025CSJ000127

—Claire Beveridge, Editor, Community Science

Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

A stalagmite in a Yucatán cave has provided new insight into the role drought may have played in Maya sociopolitical shifts more than 1,000 years ago. A recent analysis of a rainfall proxy in the Maya lowlands revealed that several episodes of severe, prolonged drought occurred during the Maya Terminal Classic Period (roughly 800–1000 CE), a time when large urban centers experienced major sociopolitical shifts.

The researchers suggest that just as climate change acts as a threat multiplier today, drought may have amplified existing troubles in Maya political centers like Chichén Itzá and Uxmal and added climate stress to societies already under pressure.

“These climate events would have affected each individual site in a very specific way depending on the resilience of that site at that time,” said lead researcher Daniel James, who studies paleoenvironmental reconstruction at University College London. “Hopefully the precision of this record allows that [analysis] to be done at individual sites…then we can really start to build up a picture of what I am certain will be a wide variety of societal responses to climate change across this time and across the region.”

Extended Droughts During Wet Seasons

During the Maya Terminal Classic Period, several Maya city-states in the southern lowlands (in modern-day Mexico, Belize, and Guatemala) experienced sociopolitical upheaval, site abandonment, and depopulation. Political and cultural centers shifted northward. Although the societal changes are clear in the archaeological record, there is still widespread debate about the potential drivers of these shifts as well as why some city-states survived while others did not.

This map of the Maya lowlands in Yucatán marks sites of prior paleoclimate studies with white squares, with this study’s site, Grutas Tzabnah, marked with an X. White circles denote Northern Maya Lowland sites, and stars denote sites of interest to this study. The land is shaded on the basis of its elevation in meters above sea level (m asl). Blue contours outline modeled mean annual total rainfalls from 1979 to 2022 in millimeters per year. Credit: James et al., 2025, https://doi.org/10.1126/sciadv.adw7661, CC BY 4.0

Drought comes up often in these debates as a potential destabilizer: Insufficient or unpredictable rainfall can lead to food instabilities, trade disruptions, disease, and even military conflicts. But previous paleoclimate studies failed to precisely pin down the timings and durations of droughts in the Maya lowlands during the Terminal Classic Period, James said.

James and his colleagues trekked to a cave called Grutas Tzabnah, in the state of Yucatán, Mexico, located near several large Classic Maya sites, including Chichén Itzá and Uxmal. This cave has been sought out before for paleoclimate studies of the region because of its accessibility and well-preserved cave formations. What’s more, Grutas Tzabnah is also a relatively shallow cave, which means that water does not take long to drip into the cave from ground level.

The researchers chose a stalagmite that has been growing for thousands of years and shows distinct annual growth layers. This particular stalagmite grew fast in the layers that dated back to the Maya Terminal Classic Period, James said, so the team was able to collect 10–20 data points within each annual layer to determine subannual, seasonal rainfall.

Researchers Daniel James (left), Ola Kwiecien (center), and David Hodell (right) install a drip water autosampler in Grutas Tzabnah to analyze seasonal changes in drip chemistry. Credit: Sebastian Breitenbach, 2022

“You can see wet seasons and dry seasons in our record, whereas previous records from the same cave are looking at annual average rainfall,” James said. “Wet season rainfall is what determines the success or failure of agriculture, as opposed to annual average.”

They measured the ages of the layers using uranium-thorium radiometric dating and rainfall quantity using a stable oxygen isotope ratio, δ18O, within calcite. Stalagmite samples that recorded a lower δ18O indicate more rainfall, while higher δ18O indicates less rainfall. The team calibrated their paleoclimate calculations with modern rainwater and cave drip measurements over a few years to ensure that they could convert the stalagmite’s δ18O measurements to rainfall.

From 871 to 1021, the stalagmite recorded eight extreme droughts during wet seasons, each lasting at least 3 years. A 4-year drought that started in 894 was interrupted by a single wet year and was followed by another 5 years of wet-season drought. A few decades later, the region had experienced 13 consecutive years of wet-season drought (929–942), longer than any multiyear drought in local historical records. This research was published in Science Advances in August.

“The chronology makes this one of the most detailed paleoclimate records available for understanding human-climate interactions during the Maya collapse period.”

“This new study represents a significant advancement in our understanding of Terminal Classic drought patterns, primarily due to its exceptional temporal resolution and robust age control with uncertainties of just a few years,” said Sophie Warken, who studies speleothems and climate variability at Universität Heidelberg in Germany and was not involved with this research.

“This high-resolution approach enables the authors to examine the timing and duration of individual drought episodes very precisely, which previous studies could only identify as broad periods of drying,” Warken added. “The chronology makes this one of the most detailed paleoclimate records available for understanding human-climate interactions during the Maya collapse period.”

One Piece of the Puzzle

While this rainfall record is a big step forward, Warken said that she would like to see it verified using additional proxies like trace elements, as well as a longer modern calibration period. She would also like to see this record extended to before and after the Terminal Classic Period to gauge whether those droughts were truly exceptional for the region.

“Such expanded paleoclimate networks could also provide crucial baselines for assessing recent and future climate changes in this vulnerable region,” she added.

Despite the fact that the extended droughts coincide with major societal shifts, James cautioned that this does not mean that drought caused these changes or were even the most important factor.

“I would love for this data to be used to pick apart individual stories from individual sites of resilience and survival, as well as the stories of disintegration of systems and abandonment and loss of population.”

“Other hardships like famine, disease, and internal violence could have been caused by drought or indeed could have been ongoing beforehand and made the society more susceptible to and less prepared for climate hardship,” James said.

Importantly, archaeological evidence suggests that two Maya cities near this cave, Chichén Itzá and the regional capital of Uxmal, did not decline at the same rate. (Uxmal declined much more rapidly.) Understanding the pressures that the two cities experienced, including drought, will be key to creating a holistic picture of how the cities functioned during the Terminal Classic Period.

“While climate stress likely played an important role in the Terminal Classic transformations,” Warken said, “the Maya’s response to drought was probably mediated by existing social, political, and economic vulnerabilities that varied between different centers and regions.”

“It could be how well were they ruled, how rigid or flexible was their political system, how good was their water management at the time,” James said.

“I would love for this data to be used to pick apart individual stories from individual sites of resilience and survival, as well as the stories of disintegration of systems and abandonment and loss of population,” he added.

—Kimberly M. S. Cartier (@astrokimcartier.bsky.social), Staff Writer

Citation: Cartier, K. M. S. (2025), Major droughts coincided with Classic Maya collapse, Eos, 106, https://doi.org/10.1029/2025EO250361. Published on 24 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

By the end of the 19th century, most of the estimated 30 million to 60 million bison roaming North America’s Great Plains had been eradicated, forever changing the landscapes that they lumbered across. Today, only about 400,000 bison remain in North America, and 95% of those animals are raised and managed for meat production. The remaining bison live in carefully controlled conservation settings, such as Yellowstone National Park.

Though existing research suggests that grazing bison can diversify habitats and drive nutrient cycling and productivity, studying how they shape larger, ecosystem-scale landscapes has been difficult. A new study conducted in Yellowstone does just that, showing how bison grazing boosts nitrogen cycling and improves plant health to create resilient grasslands.

Yellowstone offers a unique national laboratory for the study because the 5,000 bison who call the park home range freely along a 50-mile (80-kilometer) migration route. Their migration is a reminder of a time when millions of these massive herbivores roamed the continent.

Grazing Helps the Grass Grow

Researchers measured the plant productivity of 16 Yellowstone sites between 2015 and 2022. These sites spanned across lawn-like valley bottoms, which were grazed throughout grass-growing seasons; dry hillsides, which were grazed in spring, fall, and winter (depending on snowpack); and high-elevation wet areas, which were grazed from early summer until frosts arrived.

“Bison are true grazers. They put their heads down and just do their thing.”

Though other species subsist on these grasses, cameras and GPS data from tracking collars showed bison were the primary grazing animals in these locations. “Bison are true grazers. They put their heads down and just do their thing, whereas elk and mule deer and pronghorn are a little more selective in what they eat,” said study author Bill Hamilton, an ecologist at Washington and Lee University.

To compare areas of vegetation that were and were not grazed by bison, researchers constructed a series of movable “exclosures” that kept bison out of 9-square-meter areas. Some of the exclosures remained fixed in place from April, when snowpack melts and new plants emerge, to October, the time when grasses stop growing. Moveable exclosures, shifted every 30 days, followed the bison along their migration path.

By comparing fenced and unfenced areas, researchers were able to estimate how much biomass had been eaten. The missing biomass was added back into the productivity total to give a fuller picture of plant production. “It would be like if each time you mow your lawn and put it into a bag and then you weighed that pile at the end of the summer, you would know how much your grass you actually grew,” said Hamilton.

Having thousands of large animals grazing at once may seem like it would inhibit plant growth, but the study showed the opposite effects. As bison graze, they speed up the nitrogen cycle in the plants and soil around them. Urine is one concentrated source of nitrogen, and as it enters the soil, enzymes and microbes break it down into ammonia, which is then converted to ammonium. Some plants absorb the ammonium directly, and microbes further convert some of the ammonium into nitrate that is absorbed by other plants.

Populations of ammonium-oxidizing bacteria were concentrated more highly in heavily grazed areas, leading to more available nitrogen for plant growth. Researchers determined nitrogen levels by burying reverse osmosis membranes in the ground and measuring the amount of ammonium and nitrate collected on the membrane after 30 days.

The large amounts of bison urine and feces entering the soil weren’t the only cause of increased nitrogen, Hamilton explained. “We found differences in the amount of nitrogen in the leaf tissue above ground. But where’s it coming from? It’s coming from the nitrogen cycle that’s being stimulated by the grazing.”

Mutually Beneficial

When a bison chomps off the top of a patch of grass, the plants quickly release carbon into the soil through their roots to regrow their lost tissue. The carbon stimulates the microbes belowground to accelerate decomposition and nutrient cycling. “In 7 days, there’s more nitrogen in the leaf tissue of that [grazed] plant than a comparable ungrazed plant,” said Hamilton. “The key is that plants are not passive in this process.”

“Animals, through their natural behavior of congregating and eating and hanging out, are changing the way plant communities grow and respond.”

Increased nitrogen in the grasses means more nutritious food for Yellowstone’s migrating bison herds and resident ruminants: The rise in nitrogen translated to a 156% crude protein increase in lawn-like grasslands, a 155% increase in high-elevation habitats, and a 119% increase in dry areas.

“What this study tells me is that what’s happening in Yellowstone is very special,” said Tyler Kartzinel, an ecologist at Brown University who was not part of the study. “Animals, through their natural behavior of congregating and eating and hanging out, are changing the way plant communities grow and respond. And that has cascading effects on nutrient cycling and the functioning of the entire ecosystem.”

Yellowstone and Beyond

Yellowstone’s free-roaming bison herds provide a glimpse at the past, but they also might offer insight into how to preserve and improve grasslands in the future, the study authors suggest. When large groups of herbivores move freely across landscapes—like wildebeest in the Serengeti and caribou in Alaska—they can influence the broader ecosystem in a positive way.

Restoring bison in large-scale settings allows them to fulfill not only significant ecological roles but cultural roles as well. “That’s important for Indigenous tribes that have spiritual and cultural connections to these animals,” said Hamilton.

Besides Yellowstone, few places exist in North America to study how large grazing animals affect the land that sustains them. The nonprofit American Prairie aims eventually to connect 3.2 million acres (about 1.3 million hectares) of grasslands and host thousands of free-roaming bison in Montana—offering another venue where bison could migrate and interact with plants and soil on a large scale. “This study says it could work,” said Hamilton.

—Rebecca Owen (@beccapox.bsky.social), Science Writer

Citation: Owen, R. (2025), Free-roaming bison graze life into grasslands, Eos, 106, https://doi.org/10.1029/2025EO250355. Published on 24 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

For decades, as Jill Kenny drove around her hometown of Tāmaki Makaurau Auckland, Aotearoa New Zealand, the now-retired geologist and geomorphologist wondered about the shape of the landscape. Auckland, the country’s largest urban area, is located on a volcano-studded isthmus on the North Island. Obvious volcanic mounds aside, why does the city have the shape it does?

Kenny noticed similar flat, eroded surfaces at different heights above sea level and one day in 2004 had an aha moment: Maybe these surfaces were at different elevations because they had been offset by faults. In other places, scientists can trace the echo of fault lines in the landscape or through seismic data, but Auckland has few earthquakes and is covered over by a patchwork of small lava flows and ash deposits, as well as by a modern concrete jungle.

Were there faults hidden beneath the city, and could any of them still be active?

A Passion Project

For years, Kenny trawled through paper and digital records detailing data culled from boreholes drilled through the lava and into Auckland’s underlying sedimentary rock. The boreholes were part of geotechnical investigations conducted for construction purposes and resulted in a total of 2,000 logs.

In those records was evidence of Miocene epoch layers of sandstone and mudstone called the Waitematā Group. The top of this layer, which eroded between 15 million and 5 million years ago, is the “only potential marker horizon that can be followed with any certainty across the Auckland region,” wrote Kenny and University of Auckland volcanologist Jan Lindsay in a 2012 paper describing the borehole research. That study identified a number of previously hidden faults under Auckland, as indicated by adjacent boreholes showing the surface suddenly raised or lowered along a consistent line.

In the years since Kenny and Lindsay’s 2012 paper, Auckland borehole data have proliferated. Logs that were once closely guarded by companies and organizations have been brought together in the New Zealand Geotechnical Database.

“Every little extra piece of information adds to the overall jigsaw puzzle,” said Lindsay.

Now, in a new study published in the New Zealand Journal of Geology and Geophysics, Kenny, Lindsay, and colleagues have used information from 8,200 boreholes in addition to new remote sensing and geophysical data to provide a revised model of post-Miocene faulting in the Auckland region. The model includes a new geospatial database, fault maps, and a suggested standard methodology for classifying obscured urban faults. The research identifies 10 likely and 25 possible faults in the region and erases some nonfaults that have been incorrectly propagated through the literature for decades.

Jill Kenny and colleagues trawled through data from 8,000 urban boreholes to identify 10 likely and 25 possible faults under Auckland. Credit: Kenny et al., 2025, https://doi.org/10.1080/00288306.2025.2519722, CC BY-NC-ND 4.0

“This paper is really a passion project,” said Lindsay. “It was a long, painstaking process, but we think we’ve ended up with a really robust catalog of buried faults in Auckland with a range of different confidences attached to them.”

Finding Urban Faults

While earthquakes are among the deadliest threats to urban areas, cities are tricky places to find faults, said Nicolas Harrichausen, who studies crustal deformation at the University of Alaska Anchorage but was not involved in the New Zealand research. Surface offsets can be just a few meters across, and when a shopping center, road, high-rise, or house is built, evidence of the fault is erased.

Another problem is access. “Say you do find something that is interesting—you can’t go dig a trench in somebody’s yard or dig up a street in the name of finding a potential fault,” said Harrichausen, who recently discovered an active fault in the city of Victoria on Vancouver Island, Canada.

Finally, seismic surveying is one of the most common fault-finding methods, and cities are just seismically noisy places. Vibrations from construction and highways consistently interfere with seismic signals. Some cities restrict or even ban seismic surveys for ecological reasons.

However, cities do have some advantages when compared to places like the extremely remote Alaskan fault he’s currently investigating, Harrichausen said: People are already digging lots of holes.

The new Auckland borehole database is a great starting point for research into the geomorphology of the region, Harrichausen said, but because of the ancient age of the marker horizon, “for earthquake hazard, it’s just kind of a baseline.” The United States considers a fault active if it’s ruptured in the past 12,000 years. In New Zealand, it’s 125,000 years. So far, the faults Kenny and Lindsay have identified can be constrained only to the past 5 million or so years.

“It may be that none of [the suggested Auckland faults] are active,” Lindsay acknowledged. “But we need to better understand our faults and how active they are in order to work out what our actual seismic risk is.”

The study has revealed several compelling candidates for further investigation, she said, including one fault near the peninsula suburb of Bucklands Beach, where offset evidence from the boreholes lines up with a visible scarp through a golf course. “If I were studying faults in Auckland, that’s where I would start,” said Harrichausen.

Where Magma Meets Fault

Even if none of the Auckland faults turn out to be active, for Lindsay, a volcanologist who leads the transdisciplinary Determining Volcanic Risk in Auckland project, the fault maps and database are also the starting point for another line of inquiry with significant implications for the people of the city.

“We need to know as much as we can about the rocks beneath our feet.”

The new maps show that the region’s volcanoes consistently erupted near the faults and associated structures, but not right along them. “Volcanologists are not in agreement as to whether magma prefers to move along faults or whether it prefers to move between faults,” said Lindsay. There seems to be some relationship between the structural fabric of the region and its volcanism, she and her coauthors wrote.

“In order to better understand the active volcanoes in Auckland, we need to know what the subsurface looks like,” Lindsay said. “We need to know the structure, we need to know where the faults are—we need to know as much as we can about the rocks beneath our feet.”

—Kate Evans (@kategevans.bsky.social), Science Writer

Citation: Evans, K. (2025), “Passion project” reveals Auckland’s hidden urban faults, Eos, 106, https://doi.org/10.1029/2025EO250354. Published on 23 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Kelp forests are underwater jungles and some of Earth’s most productive ecosystems, absorbing carbon, providing refuge for a myriad of marine life, and buffering vulnerable shoreline communities and infrastructure. But kelp ecosystems are under increasing climate stress and have been whittled down by overgrazing urchins as key food webs have collapsed.

“Our results suggest that kelp canopy can be a useful indicator of ecosystem resilience within MPAs under climate stress.”

New research led by a team from the University of California, Los Angeles (UCLA) and published in the Journal of Applied Ecology examines the effects of marine protected areas (MPAs) on giant and bull kelp forests off the coast of California. When comparing kelp in protected and unprotected waters, researchers found only modest differences in the surface layer of fronds. Following climate disturbances like marine heat waves, however, kelp within MPAs proved far more resilient, especially in Southern California.

“Our results suggest that kelp canopy can be a useful indicator of ecosystem resilience within MPAs under climate stress,” said Emelly Ortiz-Villa, a graduate student at UCLA and lead author of the study.

California as a Bellwether

California’s kelp forest ecosystems are threatened by factors such as marine heat waves and imbalanced food webs. As sea surface temperatures get hotter, researchers say, California’s experience may be a bellwether for temperate ecosystems globally.

During a catastrophic marine heat wave that struck the California coast between 2014 and 2016, Northern California lost more than 90% of its kelp canopy, which wreaked havoc on marine food webs as well as coastal economies that rely on tourism and fishing.

Compounding the threat posed by marine heat waves is the purple sea urchin, an animal that can devour kelp faster than the seaweed can reproduce. Predators of the sea urchin, including sea stars and sea otters, face pressures, including disease and habitat loss. When urchins outnumber their predators, once lush and verdant kelp forests can become spindly outcrops nicknamed “urchin barrens.”

In the future, said Ortiz-Villa, “research should examine how multiple stressors interact to influence kelp forest recovery, so we can better pinpoint where and when MPAs are most effective at enhancing resilience.”

“What We Mean by Protected”

Marine managers have long sought tools to buy time for kelp forests to recover.

MPAs are one such tool. Many MPAs limit or ban extractive activities, including fishing, but until now, their effectiveness for kelp conservation remained understudied. Using 4 decades of Landsat imagery of the California coast, researchers compared 54 kelp forests in MPAs to those with similar environmental features in unprotected waters.

Following climate disturbances like the 2014–2016 marine heat wave, researchers found an 8.5% increase in kelp coverage in fishing-restricted MPAs. In these protected areas, healthy populations of predator species like California sheepshead and spiny lobsters helped control sea urchin populations that might otherwise have overwhelmed compromised kelp forests.

There’s “a growing body of evidence that we need to be more targeted in terms of what we mean by protected.”

Previous research established the value of MPAs for preserving biodiversity, but this study is among the first to document their advantage to kelp forests. “It’s important to demonstrate that there is an additional benefit of MPAs, and they can be an extra part of the toolbox for protecting kelp forests,” said Aaron Eger, director of the Kelp Forest Alliance and a postdoctoral fellow at the University of New South Wales in Australia who was not involved in the study.

While MPAs tend to follow guidelines established by the International Union for Conservation of Nature, their levels of protection and management can differ greatly. Reserves range from strictly protected “no-take” areas, where all extractive activities are prohibited, to “multiple-use” zones that accommodate fishing and industrial operations. Some MPAs even allow controversial bottom-trawling practices.

The differing standards are “part of a growing body of evidence that we need to be more targeted in terms of what we mean by protected,” said Eger.

—Amelia Macapia (@ameliamacapia), Science Writer

Citation: Macapia, A. (2025), Marine protected areas show promise for kelp forest recovery, Eos, 106, https://doi.org/10.1029/2025EO250350. Published on 23 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Editors’ Vox is a blog from AGU’s Publications Department.

Ice sheets can have a large impact on the surrounding land and sea where they form, which can create many different landforms. One such landform is called a “glacial forebulge,” which is a long hill that forms in front of an ice sheet. These glacial forebulges are important to understand since they influence sea level and earthquake activity in North America and Europe.

A new article in Reviews of Geophysics digs deeper into the characterization of glacial forebulge dynamics. Here, we asked the authors to give an overview of glacial forebulges, how scientists study them, and what questions remain.

How would you define a glacial forebulge to a non-specialist?

Figure 1: Conceptual model of a glacial forebulge. The glacial forebulge is the upheaval of the lithosphere in front of the ice sheet. This model is based on Stewart et al. [2000], Grollimund and Zoback [2003], Keiding et al. [2015], and R. Steffen et al. [2021]. Lithostatic strength distribution based on Bürgmann and Dresen [2008]. Credit: Brandes et al. [2025], Figure 1

A glacial forebulge is an upheaval of the Earth’s surface that is formed in front of an ice sheet and runs parallel to the ice margin. These form because the ice mass bends the lithosphere and deforms the Earth’s surface. Below the ice load, a depression is formed, whereas outside the ice sheet, there is a gentle upward bending of the surface that forms the forebulge. This bending is supported by Earth’s lithosphere and mantle material that flows away from the area pushed down by the ice sheet.

For the large ice sheets that covered northern Europe and North America during the last glacial period, state-of-the-art numerical models show that the crest, or the highest point, of the forebulge has a maximum height of approximately 50 meters and 90 meters in these areas, respectively. Additionally, the crest can be one or two hundred kilometers away from the ice margin’s maximum extent around 20,000 years ago.

Why is it important to understand glacial forebulges?

The formation, geometry, and motion of the glacial forebulge affects sea level, river flow paths, and earthquake activity in North America and Europe. Though the large ice sheet of the last glaciation is already gone, remnants of the forebulge still exist. With the ice load gone, the area that was below the ice is uplifting and, consequently, the area of the forebulge now subsides. This process is slow and still ongoing, because the lithosphere cannot move independently from the underlying mantle. The mantle has a high viscosity and flows slowly. The subsidence in the forebulge area is on the order of a few millimeters per year. At coastlines, this subsidence can result in a rising sea level. The subsidence of the forebulge area also changes the stress state of the lithosphere, which could lead to a reactivation of pre-existing faults that can cause earthquakes.

Figure 2: Stresses in the lithosphere related to the glacial forebulge. The ice sheet loads the lithosphere and causes a depression below the ice sheet and an upheaval outside the ice sheet. The bending causes stresses in the lithosphere indicated by the arrows. Arrows that point towards each other indicate compressional stresses (material is pushed together), whereas arrows pointing in opposing directions indicate extensional stresses (material is pulled apart). These stresses change, when the ice melts. The figure is based on Stein et al. [1989], Stewart et al. [2000], Grollimund and Zoback [2003], and Keiding et al. [2015]. Credit: Brandes et al. [2025], Figure 14

How do scientists observe and measure glacial forebulges?

In the early years of forebulge research, the existence of a glacial forebulge was predicted by theoretical considerations and calculations assuming that the Earth’s surface before the onset of glaciation was “more-or-less flat,” and the deformation was due to the loading and unloading of a single ice sheet. These models were used to interpret the observed relative sea level data in the near field of the ice centers in Northern Europe and North America. In particular, these correspond to a transition zone between a zone showing pure land emergence and a zone showing pure land submergence in the pattern of global sea level change. These zones are carefully explained in our study.

Figure 3: Visualization of the different zones of relative sea level, predicted by state-of-the-art numerical models that simulate glacial isostatic adjustment processes like surface uplift/subsidence, changes in the rotation of the Earth, changes in the gravitational attraction due to the ice mass loss etc. Melting of the large ice sheets in both the northern and southern hemispheres after the Last Glacial Maximum led to changing sea levels around the world. However, changes are not uniform, and the different zones are visualized in different colors here, with zones II to V being affected by the dynamics of the glacial forebulge. Credit: Brandes et al. [2025], Figure 13

How have these methods evolved over time?

Today, high-quality sea level data are used in the study of glacial forebulges and the numerical models used to simulate Earth deformation and relative sea level change are more refined and realistic. For example, the change in ice-thickness history and the migration of ice margins for all the globally important ice sheets and their interactions on Earth deformation are included. Earth material properties can now change both radially and laterally. Mantle flow-law can be linear, nonlinear, and composite.

In the computation of sea level change, effects such as time-dependent coastlines (land-sea boundaries that change geographically over time) and rotational feedback (the concept that the redistribution of Earth’s surface and internal mass can change its rotational or spin motion and thus sea levels) are included. Besides sea level data, other measurements from modern geodetic techniques are also used. For example, forebulge decay is inferred from the Global Navigation Satellite System (GNSS).

What are some of the major controlling factors of glacial forebulge behavior/evolution?

The most important controlling factors for the geometry and behavior of a glacial forebulge are the structure and material properties of the lithosphere and the viscosity, as well as the flow pattern, in the underlying mantle (including the asthenosphere). A thick and rigid lithosphere leads to a lower forebulge crest that is located at a greater distance from the ice margin. If the lithosphere is thinner and less rigid, the forebulge crest is higher and lies closer to the ice margin. In addition, the forebulge can migrate through time. The flow behavior of the material in the mantle determines how the forebulge migrates when the ice sheet decays. When the ice melts, the forebulge can either retreat with the ice margin, move away from the ice margin, or decay in place.

Figure 4: Schematic view showing the controlling factors for forebulge evolution. The height and width of the forebulge depends on properties of the lithosphere. A thick lithosphere makes the forebulge lower and further away from the glacier. If the lithosphere is thinner, the forebulge is higher and closer to the glacier. The highest part of the forebulge can migrate through time. The type of material flow in the mantle controls how the forebulge moves when the ice melts. This figure is based on the work of Wu and Peltier [1983], Wu [1993], Kaufmann et al. [1997], O’Keefe and Wu [1998]. Credit: Brandes et al. [2025], Figure 3

What are some of the recent advances in our understanding of glacial forebulges?

GNSS data have allowed researchers to map the position of the now decaying forebulge with a higher accuracy. GNSS data for North America and Europe show that there is a belt of subsidence parallel to the former ice margin. This subsiding area is the decaying forebulge.

In North America, it can be found in the southern parts of the central Canadian provinces and the northern United States, making a north-east turn just south of the Great Lakes, then following the northeastern US east coast states into the Maritimes. In Central Europe, there is a subsidence zone that runs from the Netherlands across northern Germany into Poland and southern Lithuania.

This knowledge about today’s forebulge behavior can help to better predict future sea level positions along the US east coast and the coasts of northern Central Europe, the Baltic countries and southern Scandinavia.

What are the major unsolved or unresolved questions and where are additional research, data, or modeling efforts needed?

Because of its small magnitude today, the forebulge in Europe is difficult to observe. Moreover, there are overlapping processes from sediment compaction and large-scale tectonics that hamper a clear determination with geodetic techniques in northern Central Europe. In northeastern Europe, the forebulge cannot be traced yet because of insufficient geodetic data.

To solve these questions, more high-quality data and improved modeling of glacial isostatic adjustment are needed. An integrated study should interpret all the observations and modeling results related to the glacial forebulge (including, for example, height and direction of bulge movement, uplift pattern, relative sea level changes, stress changes and seismicity, etc.) to get a more coherent view of the glacial forebulge dynamics. Precise terrestrial GNSS measurements and satellite data should be utilized to enhance the understanding of forebulge evolution and help to distinguish its subsidence signal from the subsidence caused by groundwater extraction or natural gas recovery.

—Christian Brandes (brandes@geowi.uni-hannover.de, 0000-0003-2908-9259), Leibniz Universität Hannover, Germany; Holger Steffen (0000-0001-6682-6209), Lantmäteriet, Sweden; Rebekka Steffen (0000-0003-4739-066X), Lantmäteriet, Sweden; Tanghua Li (0000-0003-0501-0155), Nanyang Technological University, Singapore; and Patrick Wu (0000-0001-5812-4928), University of Calgary, Canada

Citation: Brandes, C., H. Steffen, R. Steffen, T. Li, and P. Wu (2025), How glacial forebulges shape the seas and shake the earth, Eos, 106, https://doi.org/10.1029/2025EO255030. Published on 23 September 2025. This article does not represent the opinion of AGU, Eos, or any of its affiliates. It is solely the opinion of the author(s). Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Throughout Earth’s history, so-called volcanic winters have radically altered Earth’s climate. In these events, gases expelled by powerful eruptions form aerosols that reflect the Sun’s radiation and prevent it from warming the planet.

But eruptions’ effects on Earth systems don’t stop with temperature. Large eruptions can have diverse, down-the-line impacts such as altered rainfall, damaged crops, and, according to a new study, disrupted seasonal flood patterns.

The work “builds a bridge between climate modeling work on volcanic eruptions and the potential impacts on people and societies.”

“Usually, when we think about volcanoes, we think of them through the lens of changes in temperature,” said Gabriele Villarini, a hydroclimatologist at Princeton University and a coauthor of the new study. “The question I had was, ‘How about volcanoes and their impact on flooding at the global scale?’”

Villarini and his colleagues simulated the effects of three major volcanic eruptions in Earth’s past. Their results, published in Nature Geoscience, showed an asymmetric pattern: Major eruptions in the tropics of one hemisphere appeared to coincide with substantial increases in seasonal flooding in the opposite hemisphere. The findings could guide disaster response efforts and offer insight into the possible effects of geoengineering as well.

The work “builds a bridge between climate modeling work on volcanic eruptions and the potential impacts on people and societies,” said Matthew Toohey, a climate scientist at the University of Saskatchewan who was not involved in the new study. 

Opposing Hemispheres

The research team used previous simulations of Earth’s climate system to obtain precipitation and temperature data for 5 years after three highly explosive eruptions: Guatemala’s Santa María in 1902, Indonesia’s Mount Agung in 1963, and the Philippines’ Mount Pinatubo in 1991. The scientists also simulated hypothetical control worlds where those eruptions never happened.

Then, the researchers used those data in a statistical model to see how flooding patterns might respond. The model reproduced streamflow conditions after the three eruptions and in the hypothetical cases where the eruptions did not occur.

When volcanic plumes were confined to one hemisphere, the scientists found, peak stream gauge readings increased in the opposite one. (Such readings have long been an indicator of seasonal flooding.)

Mount Agung’s 1963 plume stayed in the Southern Hemisphere. In the year after the eruption, about 50% of stream gauges in tropical Southern Hemisphere basins showed a decrease in peak readings when compared to the noneruption scenario, according to the model. In the Northern Hemisphere, however, about 40% of stream gauges showed an increase in peak flow in the year after the eruption. 

The effects of the eruption of the Santa María volcano in 1902 showed a similar pattern: In the 2 years after the eruption, simulated stream gauges in the Northern Hemisphere (where the aerosols were concentrated) had decreased flows, while those in the Southern Hemisphere experienced an abrupt increase.

The eruption plume from Mount Pinatubo, however, was more evenly distributed across the Northern and Southern Hemispheres, and its effects were distinct. The team’s simulations showed that in the 3 years after the eruption, Pinatubo’s plume decreased flooding in tropical regions but increased stream gauge readings in arid areas of each hemisphere. 

Eruptions on the Equator

The research team didn’t directly identify the underlying reasons for their results, but Villarini said it’s likely the volcanic emissions and flood patterns are linked via the Intertropical Convergence Zone (ITCZ), a band of strengthened precipitation where Earth’s trade winds meet. 

The band of thunderstorms seen here in the area around northern South America marks part of the Intertropical Convergence Zone (ITCZ). The ITCZ is an area of enhanced moisture circling the globe north of the equator. Credit: NASA/GSFC, Public Domain

Gases from volcanic eruptions, especially sulfur dioxide, oxidize to form tiny particles that scatter sunlight, cooling Earth’s surface and creating a temperature differential that pushes the ITCZ away from the hemisphere containing the plume. This shift in the ITCZ likely pushes moisture-laden air into the opposite hemisphere, contributing to increased flooding, Villarini said. 

“If we can make useful predictions about changes in rainfall and changes in streamflow, that can have a real-world impact.”

Toohey said the results are a step toward being able to predict the potential for unusual flooding or drought across broad areas after a volcanic eruption. “It’s important that we keep working in order to understand these processes better, to be able to make better predictions on a finer scale,” he said. 

“If we can make useful predictions about changes in rainfall and changes in streamflow, that can have a real-world impact,” Toohey said.

Villarini said understanding the long-term, secondary impacts of volcanic eruptions also has implications for potential geoengineering efforts. Volcanic eruptions scatter aerosols in much the same way as efforts to cool Earth’s atmosphere via aerosols would. Possible changes to flood patterns would need to be considered by any aerosol engineering efforts, he said. 

—Grace van Deelen (@gvd.bsky.social), Staff Writer

Citation: van Deelen, G. (2025), Volcanic eruptions in one hemisphere linked to floods in the opposite one, Eos, 106, https://doi.org/10.1029/2025EO250349. Published on 22 September 2025. Text © 2025. AGU. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

On Greenland’s coast, glaciers meet the sea in narrow fjords that have been carved over hundreds of thousands of years. Ice cliffs tower hundreds of meters high.

At a glacier’s terminus, where those cliffs crash into the waters of the Atlantic, small (bus-sized) chunks of ice slough off all the time. Occasionally, a stadium-sized iceberg plunks into the water.

All this glacial calving impacts sea level rise and global climate, but there’s a lot that researchers don’t yet know about how calving happens. Now, scientists have gotten a detailed look at the whole process using a fiber-optic cable on the seafloor 500 meters from a glacier’s calving front. The findings were published last month in Nature.

Maneuvering Through the Mélange

Physical processes at the calving front control a glacier’s stability, said Dominik Gräff, a glaciologist at the University of Washington in Seattle who led the new work.

“We don’t have much idea what’s actually going on below the water.”

But gaining access to a glacier’s front can be difficult, and remote sensing methods are able to visualize only the tiny fraction of the ice mass that isn’t submerged. “We don’t have much idea what’s actually going on below the water,” Gräff said.

“It’s always impressive for people to get any observations near the glacier front,” agreed David Sutherland, a physical oceanographer at the University of Oregon in Eugene who did not contribute to the new paper. Researchers working at the front, he explained, risk losing expensive equipment and have to navigate the mélange, a closely packed mix of sea ice and icebergs.

This was the first time fiber-optic sensing was deployed at a calving front. Unlike other methods, such as remote sensing and the use of submerged seismometers, fiber-optic sensing can capture myriad events across a range of times. “It can just sense everything,” Sutherland said.

Gräff and his team dropped a 10-kilometer (6.2-mile) cable on the ocean bottom across the fjord of the Eqalorutsit Kangilliit Sermiat (EKaS) glacier in South Greenland. The maneuver was somewhat tricky. “If you go too slow, the ice mélange that you push open with your vessel [will close] quickly,” Gräff said. “And that prevents your cable from sinking down.”

Julia Schmale, an assistant professor at École Polytechnique Fédérale de Lausanne (left), and Manuela Köpfli, a University of Washington graduate student in Earth and space science, unspool fiber-optic cable from a large drum on the R/V Adolf Jensen, deploying it to the fjord bottom to record data. Credit: Dominik Gräff/University of Washington

Once the cable was in place, researchers were able to collect a wealth of data.

Waves, Wakes, and Cracking

Laser light pulsing through the fiber-optic cable allowed it to function like an entire network of sensors snaking across the fjord.

Acoustic vibrations associated with calving, for instance, stretched and compressed the cable and changed backscattered light signals. Measuring these changes is the basis for distributed acoustic sensing, or DAS.

In addition to measuring acoustics, fiber optics also allowed researchers to measure how light signals change because of temperature, a technique called distributed temperature sensing, or DTS. DAS and DTS allowed researchers to capture calving events that lasted mere milliseconds.

During the 3-week experiment at EKaS, the glass fiber captured 56,000 iceberg detachments.

(1) Initial cracking at EKaS was detected through an acoustic signature traveling through fjord waters. (2) Fractures eventually led to iceberg detachments that emitted seafloor-water interface waves. (3) Detachments caused calving-induced tsunamis at the water surface that caused changes in pressure along the fiber-optic cable. (4) Calving-induced internal gravity waves traveled between layers of fjord water with different temperatures and salinities. (5) Calved-off icebergs drifted away from the glacier terminus, dragging internal wave wakes behind them, agitating the stratified fjord waters and cooling the seafloor. (6) The internal wave wakes caused seafloor currents that generated vibrations in the cable through vortex shedding. (7) Finally, icebergs disintegrated by fracturing, again detected by fiber-optic sensing of acoustic signals. Credit: Gräff et al., 2025, https://doi.org/10.1038/s41586-025-09347-7, CC BY 4.0

That volume of observations meant researchers could trace the calving process from start to finish. It began as cracks formed in glacial ice. Sounds associated with the cracking traveled through the fjord and were picked up by the cable. Then icebergs detached from the glacier, creating underwater waves that traveled between the ice and the sediment below. Iceberg detachments also caused small, local tsunamis that could be identified by pressure changes on the cable at the bottom of the fjord.

In addition to tsunamis and surface waves, the fiber-optic cable was also able to detect internal gravity waves, which travel at the interface between an iceberg’s upper, cold layer of fresh water and the warmer layer of salty seawater below. The EKaS icebergs created wakes as they drifted from the glacier, dragging internal gravity waves behind them and causing circulation in the water. Researchers measured the resulting temperature changes using DTS.

Finally, the fiber-optic cable captured the sounds of icebergs disintegrating. These signals were similar to the initial sound of cracking in the glacier but instead came from the fjord.

Wealth of Data

“There are very few seismological datasets where, within such a short amount of time, you record so many different phenomena.”

“There are very few seismological datasets where, within such a short amount of time, you record so many different phenomena,” said Andreas Fichtner, a seismologist at ETH Zürich in Switzerland who was not part of the work but collaborates with one of the study’s authors. It takes detective work to decode all those signals and assign them to physical processes, he said. “It’s pretty remarkable.”

Gräff and the other researchers hope their rich datasets can improve glacial calving models, which often underestimate the melt that occurs below the surface. Sutherland said it’s not yet clear how to incorporate details from the study into such models, however. Researchers will need to connect the observed processes and the amount of ice lost to factors they can easily measure or estimate, such as ocean temperature and ice thickness, he explained. And they’ll need to study the calving process of different glaciers. EKaS sits on bedrock where it meets the sea, for instance, while other glaciers have a floating terminus.

Still, having a huge set of observations along with information about ocean conditions, which the researchers collected using a suite of other tools, “is pretty powerful,” Sutherland said. “Maybe we can start using this dataset to try to make predictions of when icebergs are going to calve.”

—Carolyn Wilke, Science Writer

Citation: Wilke, C. (2025), A fiber-optic cable eavesdrops on a calving glacier, Eos, 106, https://doi.org/10.1029/2025EO250351. Published on 22 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Extremely heavy rainfall associated with super typhoon Ragasa could cause the Matai-an landslide dam to overtop in the next two days.

In East Asia, super typhoon Ragasa is moving westwards between Taiwan and the Philippines. At the time of writing, Earth Cut TV has a live feed from the Batanes Islands, almost in the path of the eye (although there is a good chance that data connectivity will be lost in the storm):-

This is an exceptional storm, bringing heavy rainfall and strong winds to a wide area.

The storm has the potential to bring extremely heavy rainfall to southern and eastern Taiwan. There is huge uncertainty as to the magnitude, but reports indicate that the Central Meteorological Administration has estimated that precipitation totals as high as 800 mm could be seen in the mountain areas of Hualien County.

As I have highlighted previously, there is a large valley-blocking landslide at Matai’an in Hualien County, with a large volume of water steadily accumulating. The image below, released by the Hualien Branch of the Forestry and Conservation Department, shows the level of the lake relative to the landslide dam:-

A recent photograph of the Matai’an landslide dam in Taiwan. Image from the Hualien Branch of the Forestry and Conservation Department.

As the image above shows, the freeboard is now quite low.

In consequence, the Forestry and Conservation Administration of the Ministry of Agriculture issued a red alert at 7 a.m. this morning (22 September), mandating the evacuation of vulnerable households downstream. It is estimated that this affects around 1,800 homes.

Should overtopping occur, it is anticipated that 24 September would be the most likely date, so we will need to watch with interest. The Central Weather Administration maintains an exceptional set of web resources recording accumulated precipitation in Taiwan.

Overtopping is not inevitable in the next few days, but that will almost certainly occur in the next few weeks. It is going to be fascinating to see what happens.

Return to The Landslide Blog homepage Text © 2023. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Craters formed by asteroid impacts are ubiquitous on rocky bodies, and our planet is no exception. Researchers believe they’ve pinpointed yet another impact crater on Earth, this one submerged beneath the North Sea. The structure, known as Silverpit Crater, was discovered roughly 2 decades ago, but its provenance has long been debated. With new subsurface imaging and rock samples, the team concluded that an impact produced Silverpit Crater roughly 45 million years ago. These results were published in Nature Communications.

“People simply didn’t believe it was an impact crater.”

Uisdean Nicholson, a geologist now at Heriot-Watt University in Edinburgh, Scotland, remembered the controversy that swirled around Silverpit Crater back in the late aughts. Was the circular feature lurking beneath the waters of the North Sea under roughly 700 meters of sediments caused by an asteroid impact, or something more plebeian like volcanism or subsidence?

Nicholson, a graduate student at the time, remembered the spirited discussion that ensued among scholars attending a Geological Society of London meeting in 2009. “It was a classic, old-school debate,” he said. The vote came out strongly in favor of a nonimpact origin.

“People simply didn’t believe it was an impact crater,” Nicholson said. It looked as though Silverpit Crater wasn’t destined to join the rarefied group of 200 or so confirmed impact structures on Earth.

Begging for Data

As Nicholson focused his research on other impact structures such as Nadir Crater, he kept thinking about Silverpit. One dataset in particular piqued his interest: a survey of the North Sea seafloor sediments collected in 2022. Those data, amassed on behalf of the Northern Endurance Partnership, a venture to explore carbon capture storage under the North Sea, afforded a close-up look at the 3-kilometer crater and its environs. Previous datasets had also imaged a similar area, but they were of lower resolution and did not cover the entire structure.

A colleague alerted Nicholson about the Northern Endurance Partnership data, and the researchers worked for several months to negotiate access to some of the proprietary observations. “I begged,” Nicholson said.

The researchers were ultimately successful in their quest, and the team pored over high-resolution seismic reflection data revealing faults and buried layers of sediments around Silverpit Crater. “The new data gives a far sharper set of images,” Nicholson said.

The fact that Silverpit Crater is so inaccessible is actually important scientifically, said Matthew S. Huber, a planetary scientist at the Planetary Science Institute in Tucson, Ariz., who was not involved in the research. “Because this crater formed in water and it was buried by sediments in the water immediately after it formed, the whole thing is preserved.”

Faults and Holes

The Northern Endurance Partnership data revealed faults consistent with rock being compacted to varying degrees, as would be expected in an impact. The observations also spotlighted several roughly 10-meter-deep and 250-meter-wide troughs near the rim of the crater. Such scarps could be features eroded by water rushing back into the crater after the impact, the team surmised.

In addition, Nicholson and his colleagues noticed a few pits located beyond the crater rim that were tens of meters deep and wide. “We see all these holes, essentially, around the crater for at least a crater diameter,” Nicholson said. The team thinks that such features are secondary craters, that is, structures formed by material lofted outward from the initial impact.

Secondary craters tend to be rare on Earth because they’re often rapidly erased by erosion after an impact. “We think this is the first really robust terrestrial evidence for secondary cratering,” Nicholson said.

Atomic Wrenching

In 1985, the company British Gas drilled an oil and gas well just a few kilometers northwest of Silverpit Crater. As part of the drilling process, debris excavated from the borehole was pumped to the surface, and some of it was retained for analysis. Nicholson and his colleagues obtained some of those sediments. On the basis of the appearance of tiny marine fossils in rocks from the same depth as Silverpit Crater, the team deduced that the feature formed roughly 43–46 million years ago.

Two mineral grains the team analyzed—one quartz and one feldspar, each roughly the diameter of a human hair—exhibited curious microscopic features. Both grains contained so-called planar deformation features, which are atomic rearrangements of the crystalline structure, Nicholson said. Such wrenching on an atomic scale is indicative of the extreme pressures associated with shock waves.

“This could wind up being a controversial paper within the impact community.”

A celestial object such as an asteroid or comet slamming into a rocky body can readily generate such pressures, but not much else can, Nicholson said. “It’s very difficult to form that any other way.”

The discovery of those shocked grains was a dead giveaway that Silverpit Crater formed from an impact, Nicholson and his colleagues proposed.

These results are convincing, Huber said, but a skeptic might rightfully have some questions. For instance, couldn’t the shocked grains have simply washed into the North Sea from another impact event? “They’ve only found one grain of quartz and one grain of feldspar,” Huber said. “This could wind up being a controversial paper within the impact community.”

—Katherine Kornei (@KatherineKornei), Science Writer

Citation: Kornei, K. (2025), Submerged crater near Europe tied to an impact, Eos, 106, https://doi.org/10.1029/2025EO250275. Published on 20 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Every year, soils across Thailand are baked hard and bone-dry for months under the unrelenting tropical Sun. But once the long, hot buildup to the monsoon season comes to a head and the rains arrive—typically in May or June—the landscape transforms in a matter of weeks into a patchwork of verdant wetlands.

With this transformation comes a flurry of activity, because many of these wetlands are, in fact, rice paddies ready to be planted. Workers wade through the fields, scattering seed or planting seedlings one by one in the butter soft soil. Months later, as long as these plants have had adequate water and nutrients, they’ll be harvested for their all-important grains.

Rice paddies in Thailand and elsewhere are vital economically and for food security, with rice being the third-most-grown cereal commodity globally. They are also part of the worldwide system of wetlands. From rivers and lakes to marshlands and intertidal flats, wetlands are important ecological and geochemical systems because of the ecosystem services they provide: biodiversity, natural pollution remediation, carbon sequestration, and protection against storm surges, to name a few.

Worachart Wisawapipat of Kasetsart University checks the condition of the soil at a rice paddy experiment site in northeastern Thailand. Credit: Kurt Barmettler

Rice paddies are also of special interest to geochemists, because the regular seasonal pattern of flooding and drainage in many paddy fields makes them ideal natural laboratories in which to study soil biogeochemical processes. Indeed, in several locations in Thailand in 2021, scientists from the Swiss Federal Institute of Technology (ETH Zurich) and Kasetsart University in Bangkok joined the workers in the fields as the growing season began. The scientists were there to plant not rice, though, but iron minerals, with the purpose of testing a new method for investigating how cycles of wetting and drying contribute to these minerals’ reactivity in soil.

A Linchpin of Soil Functioning

Iron is a linchpin of chemical cycling in the environment.

Iron is usually among the most abundant elements in soils, and it is a linchpin of chemical cycling in the environment. It’s thus of critical interest to farmers concerned with the availability of nutrients to their plants, engineers determining risks posed by toxic elements in soil, and land managers wanting to understand soil carbon storage. For several reasons, iron more often than not significantly controls how soil nutrients and toxic elements behave, how carbon is stored, and how other chemical processes play out in soils.

One reason is that iron is reactive in a variety of environmental conditions. In contact with air, oxidized forms of iron—including the yellow-brown to orange oxyhydroxide minerals commonly recognized as rust—are most stable and abundant. When a soil is flooded and the flow of oxygen through it is restricted, some of these minerals may dissolve. This dissolution occurs because some microorganisms can use iron minerals as an alternative to oxygen in the reduction and oxidation (redox) reactions they rely on for energy. In other words, some microorganisms essentially breathe redox-active iron (see video below), reducing oxidized iron and converting it to other forms.

Iron is also very versatile and combines with many other chemical elements to form a wide variety of minerals such as oxyhydroxides, carbonates, phosphates, sulfides, hydroxysulfates, and others that can host trace elements in their structures. What’s more, iron mineral particles tend to be very small and have large surface areas that allow them to bind other compounds. This property also makes iron mineral particles susceptible to rapid change under evolving environmental conditions.

Following Mineral Transformations in Field Soils

Studying the behavior of iron in soils using measurement techniques that are not specific to iron is challenging. The relatively small signatures of iron minerals are often overshadowed by those of the much more abundant silicate minerals that make up the bulk of soil mass.

With techniques that do selectively detect iron in soil, such as synchrotron X-ray absorption spectroscopy, researchers can follow changes in the composition of a whole soil sample, but not changes in a single target mineral. And if the complexity of the surrounding soil is stripped away to perform simplified laboratory experiments focusing on iron minerals, the mineral transformations observed do not always reflect what happens in natural soils.

Katrin Schiedung of ETH Zurich takes soil samples in a flooded rice paddy in northeastern Thailand. Credit: Ruben Kretzschmar

Iron minerals enriched in iron-57 and having precisely controlled properties can be synthesized in the laboratory, then mixed into soils to undergo reactions similar to those of natural iron minerals.

To overcome the challenge of tracing iron mineral transformations in soil, we developed a new approach using a stable isotope of iron to label synthetic minerals. Iron-57 occurs naturally, making up about 2.1% of the iron in soils and exhibiting the same chemical behavior as other naturally occurring iron isotopes (iron-54, 5.8%; iron-56, 91.8%; iron-58, 0.3%). Iron minerals enriched in iron-57 and having precisely controlled properties can be synthesized in the laboratory, then mixed into soils to undergo reactions similar to those of natural iron minerals. Even if the experimental enrichment of iron in the soil is small, the iron-57 is predominantly present in the synthetic minerals, allowing us to focus specifically on what happens to those minerals.

We chose iron-57 as a tracer in our experiments because of the ability to analyze it using Mössbauer spectroscopy. This technique, based on Rudolf Mössbauer’s fortuitous (and 1961 Nobel Prize–winning) discovery of recoilless nuclear resonance fluorescence, is sensitive to the redox state and chemical environment around iron-57 atoms in a sample. Crucially, all other isotopes of iron are invisible using this technique.

Mössbauer spectroscopy has been widely used in the Earth sciences, including for mineralogical analyses of soil samples and Martian rocks. In our application, adding iron-57 minerals into soils and tracing them with Mössbauer spectroscopy allows us to follow otherwise hidden mineral transformations (Figure 1).

Fig. 1. In the newly developed method, iron-57-labeled (57Fe-labeled) synthetic minerals are used to enrich a natural soil sample, which is placed in a porous mesh bag so it can still interact chemically with the surrounding soil. Following the experiment, Mössbauer spectroscopy, which irradiates the sample with gamma rays of a frequency that interacts only with iron-57, is used to measure the transformation products of the labeled minerals.

After adding portions of synthetic jarosite (a potassium-iron hydroxysulfate) or ferrihydrite or lepidocrocite (both iron oxyhydroxides) to small plots in the Thai rice paddies early in the growing season, we went back several times until the end of the season 4 months later to collect soil samples for analysis.

The mineral transformations we observed with Mössbauer spectroscopy were dominated by the dissolution of the added minerals and the release of reduced iron into soil pore water. Proximity to bacteria in the soil promoted mineral dissolution, and some of the released iron either remained dissolved or was trapped on soil particles. New minerals that formed—including green rust, a highly reactive hydroxide mineral that is usually difficult to detect in the environment—tended to be nanocrystalline in size and often contained both reduced and oxidized forms of iron.

Joëlle Kubeneck of ETH Zurich removes a sediment core from the intertidal wetlands of Germany’s Wadden Sea. Credit: Ruben Kretzschmar

Such results can shed light on biogeochemical cycling that affects ecosystem processes. Dissolution of iron minerals might lead to releases of associated pollutants, nutrients, or carbon compounds, for example. On the other hand, we observed that many of the reaction products in soil are nanocrystalline minerals, which have large reactive surface areas that might adsorb dissolved compounds such as metals.

We have also applied this new approach to understand a range of iron mineral transformation processes in soil and sediment environments other than rice paddies. In sediments along Germany’s north coast, for example, we observed the in situ formation of vivianite, a reduced iron phosphate mineral, in a matter of weeks. Phosphorus in vivianite has limited bioavailability, so formation of the mineral can control the availability of phosphorus in the environment and potentially reduce the risk of eutrophication. In addition, we have used the method to study oxidation reactions of reduced iron minerals, sulfidization of vivianite and lepidocrocite leading to the formation of iron sulfide minerals like greigite, and interactions between iron and organic matter during redox cycles.

An Array of Applications

Iron minerals are ubiquitous in natural environments and are used in many engineering applications. We anticipate that iron-57 labeling of minerals coupled with Mössbauer spectroscopy, although applied only to soils so far, could help to answer questions about transformations of these minerals in other domains of the Earth sciences and beyond.

Using iron-57 tracers could contribute to studies on the origins of iron mineral assemblages in sedimentary deposits on Earth or other astronomical bodies such as Mars.

Using iron-57 tracers could, for example, contribute to studies on geological processes, including weathering or metamorphism, or on the origins of iron mineral assemblages in paleosols and other sedimentary deposits on Earth or other astronomical bodies such as Mars.

Experiments with iron-57-labeled minerals could also help to understand redox-driven iron mineral transformation processes at work in applied geoscience technologies. In pollution management, for example, permeable reactive barriers containing iron are a tool for mitigating the spread of contaminants in groundwater. Another example involves geological deposits of redox-active iron minerals that may be used to store or produce hydrogen as a clean energy source.

Outside the Earth sciences, potential applications of iron-57-labeled synthetic minerals exist in fields as diverse as corrosion science, construction engineering, and experimental archaeology. The formation of rust on iron-bearing objects is the outcome of many interrelated chemical processes. Iron-57 tracers may help to follow and unravel those corrosion processes. They could also probe effects of different metal alloy compositions and exposure environments in tests of the longevity of steel infrastructure or of conservation methods for historical artifacts.

For now, our findings from the synthetic iron minerals we’ve “planted” in rice paddy soils are shaping understanding of the chemistry of periodically flooded soils, revealing that just like plants, the life cycles of iron minerals depend on the composition of and conditions in the soil. With continuing research in soils and with new applications focused on other natural and engineered environments, scientists can gain needed insights into how iron in its many forms affects vital issues from soil health and pollution transport to carbon storage and energy production.

Acknowledgments

This research was carried out as part of the European Research Council–funded IRMIDYN (Iron Mineral Dynamics in Soils and Sediments) project at ETH Zurich. The research was led by Ruben Kretzschmar with team members Laurel ThomasArrigo, Katherine Rothwell, Luiza Notini, Katrin Schiedung (published as Katrin Schulz), Joëlle Kubeneck, Andrew Grigg, Pierre Lefebvre, Sara Martinengo, and Giulia Fantappiè, while each was affiliated with the Institute of Biogeochemistry and Pollutant Dynamics at ETH Zurich, in Switzerland. We acknowledge important contributions to the research made by Kurt Barmettler (ETH Zurich) and Worachart Wisawapipat (Kasetsart University).

Author Information

Andrew R. C. Grigg (andrew.grigg@usys.ethz.ch), Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, Zurich, Switzerland; Katrin Schiedung, Thünen Institute, Braunschweig, Germany; Joëlle Kubeneck, TNO–Geological Survey of the Netherlands, Utrecht; also at Radboud University, Nijmegen, Netherlands; and Ruben Kretzschmar, Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, Zurich, Switzerland

Citation: Grigg, A. R. C., K. Schiedung, J. Kubeneck, and R. Kretzschmar (2025), Tracing iron’s invisible transformations just beneath our feet, Eos, 106, https://doi.org/10.1029/2025EO250347. Published on 19 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Two billion years ago, a massive star exploded. When its light reached Earth in 2021, it joined more than 20,000 recorded supernova candidates observed that year.

This one was unique, however. It exhibited features not shared by any other known stellar explosion: The star had shed almost all of its outer layers before it died, exposing a core rich in silicon and sulfur.

SN 2021yfj, as this supernova is labeled, revealed never before seen details about stellar interiors. Precisely how a star could die this way is a marvelous puzzle that may help researchers learn about the deaths of the most massive stars and how they spread new elements through the cosmos.

“By studying supernovae, we can develop ideas of how stars form, evolve, and die,” said Steve Schulze of Northwestern University in Illinois, who led the observations and analysis of SN 2021yfj, published in Nature.

Onions Have Layers, Stars Have Layers

All stars, including the Sun, spend the majority of their life cycles fusing hydrogen into helium in their cores, which are surrounded by a hydrogen plasma envelope. When this fuel is exhausted, the core contracts and begins fusing helium into carbon and oxygen. For the most massive stars (those at least 8 times the mass of the Sun) this process continues to fuse heavier and heavier elements in the core.

The result is an onion-like (or ogre-like) character of aging high-mass stars: a hydrogen envelope surrounding a shell of helium, around a shell of carbon and oxygen. Layers of neon, magnesium, silicon, sulfur, and higher-mass nuclei form deeper toward the core. Eventually the energy produced by fusion is no longer able to maintain the core’s integrity, at which point the star explodes as a supernova.

This is the story theory tells, at least; these shells aren’t visible to telescopes. Astronomers must deduce the makeup of interiors from the spectrum of light stars emit when they explode.

“The spectra of supernovae are their fingerprints.”

“The spectra of supernovae are their fingerprints,” said Maryam Modjaz, an astronomer at the University of Virginia who was not involved in the study. She called this type of research stellar forensics. “We see the explosion of the star and we work backwards.”

However, very massive stars shed a great deal of their outer envelopes long before they explode, as astronomers observed during the dramatic dimming of Betelgeuse in 2019. Some extremely massive specimens known as Wolf-Rayet (pronounced “rah-YAY”) stars expel their envelopes much earlier in their lifetimes. The most commonly observed type of Wolf-Rayet stars consist of a hydrogen nebula swathing the extremely hot layers surrounding the core, which is dominated by emissions from helium and carbon.

I Like That Spectrum—That Is a Nice Spectrum

SN 2021yfj took this early-shedding process further than any other star yet observed: It had shed not only its hydrogen envelope, but also its helium, carbon, and oxygen shells. The spectrum Schulze and his colleagues measured exhibited emissions from ionized silicon and sulfur, indicating the progenitor had ejected that layer before exploding.

“The progenitor star [of supernova 2021yfj] had essentially lost almost all of its shells,” Schulze said. Astronomers have observed other stars that have been stripped of their outer layers, but never to this extreme.

“We have spectra from hundreds of thousands of supernovae, [and] this is the first time we’ve seen deep into the guts of a dying star.”

Though that provided strong evidence supporting the onion model for high-mass stars, it also was surprising: Nobody expected a star to shed that much material before going supernova.

“We have spectra from hundreds of thousands of supernovae, [and] this is the first time we’ve seen deep into the guts of a dying star,” Modjaz said. “This must be a very unique, very uncommon explosion and therefore [uncommon] progenitor.”

“The properties of supernova 2021yfj are so extreme that it’s challenging to find a model that can describe all of the observations,” Schulze said.

“Our leading hypothesis or leading idea,” he continued, “is that it was a very massive star, around 60 times more mass than the Sun.” Because it had already shed its hydrogen and helium by the time it was observed, though, the star was probably even more massive when it was born. He cautioned against trying to make too many guesses based on the data so far. “Exactly how massive it was will require very detailed simulations [and] the development of models that don’t exist yet.”

Schulze noted that even with hundreds of thousands of identified supernovae, astronomers have yet to see all the possible types. To make things more challenging, supernovae—bright as they are—are single points of light that fade over a matter of days, so spotting them often involves a degree of luck. Extreme stripped-core supernovae are so rare, just by probability astronomers will find tens of thousands more typical explosions before the next example turns up in their data.

However, that task is less daunting than it sounds.

In addition to the Zwicky Transient Facility where astronomers detected SN 2021yfj, Schulze and Modjaz both hailed the Vera C. Rubin Observatory, which is expected to detect a thousand supernovae each night of operation when it comes online in 2026. Though that telescope isn’t being built for spectroscopy, its ability to scan huge swaths of the sky at once will let astronomers identify explosions to analyze in more detail quickly, bringing us closer to an understanding of how massive stars live and die.

—Matthew R. Francis (@BowlerHatScience.org), Science Writer

Citation: Francis, M. R. (2025), This star stripped off its layers long before exploding, Eos, 106, https://doi.org/10.1029/2025EO250340. Published on 19 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Local reports suggested that over 1,500 people died in this event, and a high death toll was reported by some international agencies. However, examination of satellite imagery casts significant doubt on this interpretation.

It has been widely reported that on 31 August 2025, a devastating landslide occurred at Tarasin (there are various spellings of this place) in the Marrah Mountains in Central Darfur, Sudan. Initial reports indicated that 1,000 people had been killed, making this the most deadly landslide of 2025 to date, whilst subsequent reports elevated this number to over 1,500. The reports were given credence by organisations such as Save the Children, who reported that 373 bodies had been recovered. There was only one reported survivor of the disaster.

However, it should also be noted that this very high total was not supported by government reports – noting of course that Sudan has extensive civil conflict, and that this area is not controlled by the government. BBC Verify examined the event too using Maxar satellite data, but could not identify a village that had been destroyed. The Washington Post reported that the United Nations subsequently reduced their estimated total loss of life to “scores”. On Bluesky, Dan Shugar pointed out that the reporting of 1,000 fatalities does not really stack up. The HydrologyNL Bluesky account has also posted some interesting analysis of this event.

The location of this landslide event is [13.01697, 24.38774]. This is a Planet image of the site draped onto the Google Earth DEM:-

Satellite image of the aftermath of the 31 August 2025 landslides in Sudan. Planet image draped onto the Google Earth DEM. Image copyright Planet, used with permission. Image dated 6 September 2025.

The marker highlights the rear scarp of the largest failure, although there are several landslides in the image. These landslides correlate with images from the site posted by news organisations (there are some images of unrelated landslides too, plus some AI slop). It appears that there have been multiple shallow landslides than have transitioned into channelised debris flows.

And this is a Google Earth view, with imagery from 2023 (with some cloud), showing the same area:-

The site of the 31 August 2025 landslides in Sudan. Google Earth image from 2023.

Some local reports of these landslides suggest that 1,500 houses were destroyed, based on information from the Sudan Liberation Movement. However, it is also notable that none of the published photography shows destruction on this scale. And the Google Earth images do not show any large settlements in the path of either the landslides themselves or the channelised debris flows. There is a cluster of houses at the foot of the main failure that has been destroyed:-

A cluster of buildings subsequently destroyed by the 31 August 2025 landslides in Sudan. Google Earth image from 2023.

However, this is a small number of buildings, not on the reported scale of losses. I have also looked at the Planet imagery from just before the landslide. There is no evidence of a large settlement in this area.

Thus, a reported death toll of 1,000-1,500 seems highly improbable, and the loss of 1,500 houses is not supported by the imagery or photography.

So, what are the possible explanations? We might consider the following:

1. The location is wrong – and the reports describe an event somewhere else. I consider this to be low probability, given the images that have been published;
2. Dan Shugar suggested that perhaps some sort of social event occurred in the path of the landslide – a wedding or suchlike. But the local reports are of 1,500 houses lost, and in general my experience is that this type of circumstance is reported, given the nature of the tragedy. Again, low probability.
3. The losses occurred a long distance down the channel. I cannot find such a site on the imagery, and I would expect that the location would be reported, and that photography would show the devastation. Again, this would seem to be low probaility.
4. There is mis-reporting, either accidentally or deliberately. I consider this to be high probability.

In remote areas, we have previously seen vastly inflated estimates of loss of life – a recent example was the 24 May Kaokalam landslide in Papua New Guinea. This can simply be a misunderstanding or the result of rumours. On the other hand, it can be deliberate. For example, loss of life can be inflated to attract additional resource for a population that is suffering extreme poverty, or it can be an attempt by the local forces to attract supplies for their own purposes. Others will be able to judge the most likely explanation.

I’m struggling to understand why reputable international agencies would appear to support these inflated reports, especially where they appear to provide testimony from the site.

Reference

Planet Team 2025. Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://www.planet.com/

Return to The Landslide Blog homepage Text © 2023. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

In 2018, the Hayabusa2 mission successfully encountered asteroid Ryugu. The Japan Aerospace Exploration Agency (JAXA) spacecraft arrived at, touched down on, collected samples of, and lifted off from the asteroid. It returned samples to Earth in 2020.

With plenty of fuel left for its extended mission, called Hayabusa2# or “Hayabusa2 Sharp,” the spacecraft raced off to its next objective, a high-speed flyby of asteroid 98943 Torifune in 2026. If all goes well with that rendezvous, the craft will attempt its final objective: an encounter with and touchdown on asteroid 1998 KY26 in 2031.

But that final objective may prove more difficult than initially imagined. New ground-based observations of 1998 KY26 have revealed that the asteroid is 3 times smaller than previously thought and spins twice as fast.

“We found that the reality of the object is completely different from what it was previously described as,” Toni Santana-Ros, lead author of a new study on 1998 KY26 and an asteroid researcher at Universidad de Alicante and Universitat de Barcelona in Spain, said in a statement.

Small and Fast

Astronomers discovered 1998 KY26 in 1998 when it came within 2 times the Earth-Moon distance. Radar and visual observations shortly after discovery estimated that the asteroid was about 30 meters across and rotated once every 10.7 minutes, the fastest-rotating asteroid known at that time. As the asteroid moved away, it became too faint to see for more than 2 decades. When Hayabusa2’s mission scientists selected targets for its extended mission, they relied on those 1998 calculations.

The asteroid completes one spin every 5 minutes and 21 seconds, less time than it takes to listen to Queen’s “Bohemian Rhapsody.”

Finally, in 2024, 1998 KY26 came close enough to Earth—12 times the distance to the Moon—to observe again. Using four of the most powerful ground-based telescopes available, Santana-Ros and his colleagues watched the diminutive asteroid tumble and spin from multiple angles, allowing them to calculate a more accurate spin rate than was possible with the limited radar and photometry in 1998.

They calculated that the asteroid completes one spin every 5 minutes and 21 seconds, less time than it takes to listen to Queen’s “Bohemian Rhapsody.” The team then combined those new observations with the 1998 radar data to recalculate the asteroid’s size. They found that instead of being roughly 30 meters in diameter, 1998 KY26 is just 11 meters, or about the length of a telephone pole. The team published these results in Nature Communications on 18 September.

Asteroid Ryugu (left) is roughly 82 times the size of asteroid 1998 KY26 (right). Credit: ESO/M. Kornmesser; Asteroid models: T. Santana-Ros, JAXA/University of Aizu/Kobe University, CC BY 4.0

“The smaller the asteroids get, the more abundant they are—but that also means that they are harder to find,” explained Teddy Kareta, a planetary scientist at Villanova University in Pennsylvania who was not involved with the new discovery. “The fact that this new paper finds such a small size for KY26 is tremendously interesting on its own—Hayabusa2 will be able to explore an extremely understudied population—but it also means that we might not have a tremendous number of known objects to compare to as well.”

A Challenge and an Opportunity

The new size and spin measurements of 1998 KY26 will make Hayabusa2’s planned touchdown more challenging, the researchers wrote. However, this is not the first time that an asteroid rendezvous mission has had to adjust its expectations mid-flight. Both Ryugu and Bennu, the first target of NASA’s Origins, Spectral Interpretation, Resource Identification, Security–Regolith Explorer (OSIRIS-REx) mission, had rougher surfaces than expected, requiring the respective missions to adjust their sample collection methods. Too, the OSIRIS-REx team learned that Bennu was actively spitting out material only when the spacecraft got close, which led them to change their plan for orbital insertion.

Despite the new challenges with 1998 KY26, Hayabusa2#’s team has a big advantage: 6 years to rework their game plan.

“The Hayabusa2 team are incredibly smart, hardworking, and have a ton of experience under their belts, but I’m sure that this kind of result is causing a bit of hand-wringing and concern for them even if the spacecraft is fully capable,” Kareta said.

“We have never seen a ten-metre-size asteroid in situ, so we don’t really know what to expect and how it will look.”

“I’m sure even the team has their doubts about whether or not the original plan was possible, but if I had to bet money, I still think the team will try [to touch down],” they added. “You set yourself up for success by building a great spacecraft and collecting a great team of engineers and scientists to staff it, but it’s still a bet every time you try something new.”

Even if a touchdown on 1998 KY26 ultimately proves impossible and Hayabusa2# simply flies on by, asteroid scientists will still gain valuable information about an incredibly common but hard-to-spot type of small asteroid.

“We have never seen a ten-metre-size asteroid in situ, so we don’t really know what to expect and how it will look,” Santana-Ros wrote.

“In many ways, a spacecraft visit to it now is even more exciting than it was before,” Kareta said.

—Kimberly M. S. Cartier (@astrokimcartier.bsky.social), Staff Writer

Citation: Cartier, K. M. S. (2025), Hayabusa2’s final target is 3 times smaller than we thought, Eos, 106, https://doi.org/10.1029/2025EO250353. Published on 18 September 2025. Text © 2025. AGU. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Source: AGU Advances

Streamflow drought—when substantially less water than usual moves through rivers—can seriously disrupt the welfare of nearby communities, agriculture, and economies. Synchronous drought, in which multiple river basins experience drought simultaneously, can be even more severe and far-reaching.

Recent observations and modeling suggest that on the Indian subcontinent, where major rivers support more than 2 billion people, the likelihood of synchronous drought is increasing as summer monsoons weaken, the Indian Ocean warms, and anthropogenic emissions and excessive groundwater pumping continue. However, little is known about the long-term patterns of synchronous drought in India, in part because streamflow data don’t offer information about the distant past.

By combining several decades of streamflow measurements from 45 gauge stations along India’s major rivers with high-resolution temperature and precipitation data and data from a range of paleoclimate proxies, Chuphal and Mishra have now reconstructed streamflow records across more than 800 years.

To look farther back in time, the researchers turned to the Monsoon Asia Drought Atlas, which comprises tree ring data indicating summer drought conditions across Asia between 1200 and 2012. They also considered historical records of climate patterns like El Niño, the Pacific Decadal Oscillation, and the Indian Ocean Dipole to explore connections among drought frequency, reoccurrence, and synchronicity. And they used two models from the Paleoclimate Modeling Intercomparison Project Phase 4 (PMIP4) that are part of the Coupled Model Intercomparison Project Phase 6 (CMIP6) to simulate precipitation and temperature data, as well as a hydrological model to simulate streamflow from 1200 to 2012.

With all this information, the researchers created their own reconstruction model that captured historical droughts driven by monsoon failures and connected low river levels to periods of drought-induced famine. Their findings revealed an increased frequency in synchronous drought between 1850 and 2014 compared with preindustrial centuries—an increase they surmise was likely caused by the warming climate. The researchers also suggest that future synchronous droughts may threaten water security throughout India. (AGU Advances, https://doi.org/10.1029/2025AV001850, 2025)

—Rebecca Owen (@beccapox.bsky.social), Science Writer

Citation: Owen, R. (2025), Droughts sync up as the climate changes, Eos, 106, https://doi.org/10.1029/2025EO250324. Published on 18 September 2025. Text © 2025. AGU. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

In recent years, the North American beaver (Castor canadensis) has been increasingly recognized as a valuable on-site engineer to help communities meet water management goals. Beavers are famously “eager” to build dams, which slow the flow of streams and allow wetland areas to grow.

Until now, however, land managers didn’t have a way to estimate how much water beaver reintroduction could actually bring to a habitat. Not every beaver dam results in a sprawling ponded complex; sometimes they result in smaller areas with less water retention than meets the needs of the community.

In a study published last month in Communications Earth and Environment, researchers from Stanford University and the University of Minnesota were able to link the amount of surface water in beaver ponds across the western United States to the features in those landscapes that make beaver ponds bigger.

Big, Beautiful…Beaver Ponds

Oftentimes, beavers will chain together multiple dams and ponds to form beaver pond complexes. The complexes increase an area’s water retention, cool water temperatures, and provide natural firebreaks. These wetland habitats also give the semiaquatic rodents ample room to roam and allow other species (such as amphibians, fish, and aquatic insects) to flourish.

Beaver pond complexes like the one in Happy Jack Recreation Area create habitat for wetland creatures big and small, like this (very large) moose. Credit: Emily Fairfax

“Our models highlight the landscape settings where ponds grow largest, helping target nature-based solutions under climate stress.”

The advantages of beaver pond complexes aren’t going unnoticed—the reintroduction of beavers to the North American landscape is an increasingly popular strategy for land managers looking to naturally improve a waterway.

“Managers need to know where beaver activity—or beaver-like restoration—will store the most water and maximize the environmental benefits, such as providing cooling and enhancing habitat quality” said Luwen Wan, a postdoctoral scholar at Stanford and the new study’s lead author. “Our models highlight the landscape settings where ponds grow largest, helping target nature-based solutions under climate stress.”

While improving water retention is a goal of many watershed management projects, especially in the increasingly drought-prone western United States, the researchers also emphasized that creating the largest possible ponds might not be the right solution for every area.

“It’s worth thinking about what we are actually asking of these beavers, and is that reasonable?”

“Bigger ponds are not always better,” said Emily Fairfax, coauthor on the study and assistant professor at the University of Minnesota. Fairfax explained that larger ponds are great for when the goal of the project involves water retention, but smaller ponds could be a better fit for a project in which the goals are pollution removal or increasing biodiversity. “It’s worth thinking about what we are actually asking of these beavers, and is that reasonable?”

How to Design a Dream Stream

Speaking on the main findings of the study, Wan said that she and her colleagues “found a clear link between the total length of beaver dams and the size of the ponds they create.” Additionally, they observed that the biggest ponds were found “where dams are longer, stream power is lower to moderate, and woody vegetation is of moderate [6–23 feet, or 2–7 meters] height.”

Included in the study were 87 beaver pond complexes across the western United States, encompassing almost 2,000 dams. Using high-resolution aerial imagery from the National Agriculture Imagery Program (NAIP), the team was able to connect the observed ponded area to different landscape measurements like soil characteristics, stream slope, vegetation metrics, and more.

The researchers chose NAIP imagery for its high spatial resolution and ability to cover large areas (visiting every beaver pond in the field would take too much time). Wan noted that while NAIP aerial imagery was the right fit for this project, it isn’t perfectly beaver proof. The imagery is updated every 2–3 years during the growing season, which may introduce some errors, like missing ponds even when dams have already been constructed.

Using remote sensing to predict where beaver reintroduction would be a successful match to the needs of a watershed isn’t a new idea. One frequently used model mentioned in the study is the Beaver Restoration Assessment Tool (BRAT). BRAT allows researchers to identify how many dams a given stream would likely be able to host. “That’s really important information to have,” said Fairfax, “but that doesn’t tell us how big the dams are, or how much water they could be storing.”

When Beavers Aren’t Best

Findings from this study are also helpful when selecting sites for beaver dam analogs (BDAs). These human-made structures are alternatives to beaver reintroduction that mimic beaver dams to achieve the same ecosystem benefits the beavers would bring. They are often the right tool when a waterway is too degraded to host a beaver population.

BDAs raise water levels and allow the preferred foods of beavers (such as willows and alders) to take root, giving “a little push” to the process of reestablishing a beaver population, explained fluvial geomorphologist and associate professor Lina Polvi Sjöberg from Umeå University in Sweden. Polvi was not involved in the new study.

Fairfax added that BDAs are a useful tool but are not equivalent to actual beaver dams. With beaver dams, a living animal is always present, so the land managers can count on the “maintenance staff on-site” to constantly update and monitor the waterway.

The Beavers Are Back in Town

North American beaver populations are still on the rebound from a long history of trapping and habitat loss that came with European colonization of the continent. “We are at maybe 10% of the historic population, and we actually don’t know if it’s still growing,” Fairfax said. Modern threats to beaver populations include highways and man-made dams, she added, which prevent beavers from freely moving back to places they once were.

Not everyone is quick to welcome North America’s largest rodent back to their neighborhood with open arms. Though public perceptions of beavers are shifting from pest to watershed management partner, the potential for contention still remains. Beavers occasionally build their dams in less-than-ideal locations, a situation that can result in flooded private properties and damaged infrastructure. The study notes that human influence (like trapping and land use conflicts) is a factor that land managers must consider but is not captured in statistical models.

Beavers Worldwide

The researchers found what makes beaver dams bigger in the western United States, but scientists say it will be important to replicate this study in different regions of North America, especially as beaver habitat expands northward as a result of climate warming.

“North American beavers are all one species, Castor canadensis. A beaver in Arizona is the same species as a beaver in Alaska. They all have the same instincts,” said Fairfax, “but beavers also do learn and adapt to their environments pretty strongly.”

She added that beavers will use the materials available to them, such as a colony in Yukon, Canada, that has been observed using rocks as dam-building material. “Whenever we build a model that describes what beavers are doing, there is a chance that it’s going to have a strong geospatial component to it,” Fairfax said.

Polvi agreed, stating that she hadn’t seen many studies using remote sensing methods to estimate the suitability of a stream for beaver reintroduction outside of the western United States. Putting things into a wider perspective, she added that some defining features of the American West, like the semiarid climate and large expanses of undeveloped public land, aren’t applicable to other regions of the world.

In an email, Wan said the next steps from this study include further exploring beavers’ ponded complexes across larger areas and “quantifying the ecosystem services these ponds provide, such as enhancing drought resilience.”

—Mack Baysinger (@mack-baysinger.bsky.social), Science Writer

Citation: Baysinger, M. (2025), What makes beaver ponds bigger?, Eos, 106, https://doi.org/10.1029/2025EO250341. Published on 18 September 2025. Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Source: Global Biogeochemical Cycles

This is an authorized translation of an Eos article. 本文是Eos文章的授权翻译。

多年来，海洋学家们一直困惑于为何算法会在南极偏远海域的卫星图像中检测到神秘的高浓度颗粒无机碳 (PIC)。在其他地区，高PIC是单细胞浮游植物大量繁殖的标志，这种浮游植物被称为颗石藻(coccolithophores)，这些植物闪亮的碳酸钙外壳会将光线反射回卫星。然而，长期以来人们一直认为这些极地水域温度过低，不适合颗石藻生长。

如今，得益于 Balch 等人的最新船载测量数据，这个谜团终于解开了。他们发现了一种名为硅藻(diatoms)的不同类型的浮游植物，当硅藻的反射性硅质外壳（或称硅藻壳，frustules）浓度极高时，其反射率可以模拟 PIC 的反射率。这种反射率可能导致卫星算法将这些遥远的南部海域错误地归类为高 PIC 区域。

同一研究团队此前的船上观测已证实，来自颗石藻的PIC是大方解石带的成因——大方解石带是一个巨大的、季节性的、反射性的水环，环绕南极洲北部较温暖的水域。然而，在更南端，南极大陆周围异常明亮的区域仍然无法解释，推测的成因包括松散的冰块、气泡或反射性的冰川“粉”（被侵蚀的岩石颗粒）被释放到海洋中。

研究人员乘坐R/V Roger Revelle号从夏威夷向南航行，进入较少被探索的水域，这里以冰山和波涛汹涌的大海而闻名。他们测量了PIC和二氧化硅的含量，确定了光合作用速率，进行了光学测量，并在显微镜下观察了微生物。这些数据表明，这些偏远地区的高反射率主要是由硅藻壳引起的。

然而，研究人员也惊讶地发现极地水域中有一些颗石藻，这表明这些浮游植物可以在比以前想象的更冷的海水中生存。

由于颗石藻和硅藻在海洋碳固定中都发挥着重要作用，因此这些发现可能对地球的碳循环具有重要意义。研究人员表示，这项研究还可以为改进卫星算法提供参考，以便更好地区分PIC和硅藻壳。(Global Biogeochemical Cycles, https://doi.org/10.1029/2024GB008457, 2025)

—科学撰稿人Sarah Stanley

This translation was made by Wiley. 本文翻译由Wiley提供。

Read this article on WeChat. 在微信上阅读本文。

Text © 2025. AGU. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.