EOS

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.