Feed aggregator

A team of atmospheric scientists at the University of Washington has found evidence that weather forecasters may be able to look ahead for up to 30 days when making predictions. In their study, posted on the arXiv preprint server, the group tested Google's GraphCast AI-based weather modeling and predicting system using a technique to improve initial weather conditions to improve its accuracy.

Editors’ Highlights are summaries of recent papers by AGU’s journal editors. Source: Water Resources Research

Rocks are important in subsurface engineering, but they are mostly invisible, extremely heterogeneous, and difficult to access. Thus, data on rock properties is scarce and uncertainties are large.

Liu et al. [2025] provide new AI technologies to augment existing rock data sets, which maintain important geometric characteristics. Now realistic rock images can be generated with this technology that are useful in the quantification of uncertainties. In addition, an analysis workflow is proposed to check the quality of the generated images, lending confidence in the obtained results. While currently the methods are limited to 2D images, the approaches could be applicable in 3D in the future.

Citation: Liu, L., Chang, B., Prodanović, M., & Pyrcz, M. J. (2025). AI-based digital rocks augmentation and assessment metrics. Water Resources Research, 61, e2024WR037939. https://doi.org/10.1029/2024WR037939  

—Stefan Kollet, Editor, Water Resources Research

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.

Frank Marks remembers the Diet Coke can floating in front of his face as the plane pitched violently. After several attempts to grab it, he gave up and focused on avoiding the other debris ricocheting around the cabin. Then an engine flamed out, and the pilots dumped 15,000 pounds (6,800 kilograms) of fuel in a last-ditch effort to climb to relative safety without overheating the three working engines. The flight miraculously landed safely in Barbados a few hours later.

Rather than swearing off flying forever, many of the flight’s passengers were back in the air 2 days later for another chance to chase the storm that very nearly killed them.

Most pilots give storms a wide berth, but those flying NOAA’s two WP-3D Orion aircraft—known as Hurricane Hunters—head right for them. Those flights yield important data about storm structure and intensity that can help protect people on the ground. “There’s only so much you can learn from remote sensing,” said Todd Lane, an atmospheric scientist at the University of Melbourne in Australia who was not involved in the research. So scientists, pilots, and crew keep flying, despite the risks that severe turbulence poses.

New research published in the Bulletin of the American Meteorological Society shows that Marks’s memorable flight through Hurricane Hugo in 1989 was rightly infamous—it ranks as the most turbulent NOAA Hurricane Hunter mission to date. Data from that and other bumpy NOAA Hurricane Hunter flights could make future trips safer.

The Bumpiest of Them All

Josh Wadler, a meteorologist at Embry-Riddle Aeronautical University in Daytona Beach, Fla., had a wild ride aboard a NOAA Hurricane Hunter aircraft in September 2022. He and his colleagues were flying through Hurricane Ian to study how energy was being transferred from the ocean to the atmosphere and, ultimately, into the hurricane. That Hurricane Hunter flight was by far the bumpiest of the 20 or so he’d been on, with extreme turbulence lasting for an unprecedented 10 minutes or so.

“We’re scientists—let’s try to figure this out.”

When the team finally emerged into smooth air, Wadler and others on board couldn’t help but wonder how their experience stacked up to the infamous 1989 flight through Hurricane Hugo. Being scientists, they decided to throw data at the question. “We’re scientists—let’s try to figure this out,” Wadler said.

Wadler and his colleagues mined in-flight data collected automatically by onboard navigation systems for every NOAA Hurricane Hunter flight into a tropical cyclone from 2004 to 2023. Those data, recorded every second, were already digitized and freely available online. But amassing data from two earlier flights for comparison—through Hurricane Hugo and another notoriously bumpy storm, Hurricane Allen, in 1980—required a bit more finesse. “There’s no record of them online,” Wadler said. “They’re just on tapes.”

Enter the data-wrangling skills of Neal Dorst, a meteorologist with the Hurricane Research Division of NOAA’s Atlantic Oceanographic and Meteorological Laboratory in Miami. “Back in the day we would record the flight-level data on magnetic tapes,” Dorst said. Reels of magnetic tape sit in a room just down the hall from Dorst’s office. He’s digitizing them all and processed the Hurricane Hugo and Hurricane Allen data out of sequence after a special request for this project.

For each NOAA Hurricane Hunter flight of interest, the team analyzed six different aircraft motions: three translational (forward and back, side to side, and up and down) and three rotational (roll, pitch, and yaw). For every second, the team calculated the aircraft’s acceleration and jerk—that is, the rate of change of acceleration in time—in each of those six dimensions.

“There’s a lot of folklore about that flight.”

Because rotational motion depends on position relative to an axis of rotation, the team also considered a passenger’s seat position when determining what acceleration and jerk someone on board would have experienced. “The farther away from the axis of rotation you are, the more you feel,” Wadler said. “You’re going to feel the rotational motions more in the front or back of the plane.”

When the researchers tabulated a “bumpiness index” that took into account both acceleration and jerk, Wadler discovered that his memorable flight through Hurricane Ian in 2022 ranked second to the flight through Hurricane Hugo. That finding wasn’t wholly surprising, Wadler said. “There’s a lot of folklore about that flight.”

That infamous Hurricane Hugo flight pierced the storm just 1,600 feet (500 meters) above the Atlantic Ocean. That left dangerously little airspace for maneuvering and sent the plane directly into a region of the storm known for its extreme winds. (Nowadays, NOAA Hurricane Hunters fly roughly 6 times higher.)

Different Kinds of Bumpy

The in-flight data also corroborated something that Marks and his colleagues aboard the 1989 flight remember well: Their wild ride was characterized by extreme up and down motions. “Within a minute, we went through these huge three updraft/downdraft couplets,” said Marks, a meteorologist who retired last year from the Hurricane Research Division of NOAA’s Atlantic Oceanographic and Meteorological Laboratory. Wadler’s trip through Hurricane Ian, on the other hand, involved strong turbulence directed largely sideways. “The side to side motions were unique,” Wadler said.

Hurricanes Irma (2017), Sam (2021), and Lane (2018) rounded out the top five positions. Wadler and his collaborators found that turbulence tended to be stronger for large storms that went on to weaken in the next few hours. Bumpiness was also most pronounced near the inner edge of a storm’s eyewall and near features known as mesovortexes, which are basically storms within a storm.

Beyond satisfying a personal curiosity, the finding could help make future NOAA Hurricane Hunter flights safer. “We know what to look for on radar when we’re going into a mission,” Wadler said. He hopes to take this new work in the direction of crew performance and cognition. “Is there a threshold of turbulence where humans are bad at making decisions?” he wondered. But instead of taking willing participants up on flights, Wadler plans to do laboratory experiments mimicking turbulence.

—Katherine Kornei (@KatherineKornei), Science Writer

Citation: Kornei, K. (2025), The wildest ride on a Hurricane Hunter aircraft, Eos, 106, https://doi.org/10.1029/2025EO250194. Published on 21 May 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: AGU Advances

The Chesapeake Bay is the continental United States’ largest estuary, spanning approximately 320 kilometers (200 miles) between northeastern Maryland and Virginia Beach. Like many coastal ecosystems, its water chemistry is affected by agricultural runoff, chemical weathering, and increasing atmospheric carbon dioxide.

Although rising carbon dioxide levels have led to ocean acidification, land use changes and chemical weathering from acid rain have made inland rivers and streams generally more alkaline. But long-term pH trends in coastal waters, such as the Chesapeake Bay, are less clear.

Li et al. ran a simulation to analyze pH trends in the Chesapeake Bay between 1951 and 2010, revealing a complex web of factors that altered the bay’s pH over that 60-year period.

Nutrient runoff into the Chesapeake Bay increased between 1950 and 1980 before dropping in the 1990s, thanks primarily to decreased atmospheric deposition of nitrogen and to upgrades in wastewater treatment systems. Agricultural lime application and intensified chemical weathering, which also decrease acidity, became more common over the study period. In contrast, coal mining, drainage from which can increase water acidity, declined over the study period. Weather played a role as well: Typical spring rainfall, as well as particularly wet decades such as the 1970s, pushed the upper bay freshwater plume farther into the middle of the bay and increased the area’s pH.

The researchers examined all these factors and found that overall, the upper bay generally became more alkaline over time but that deeper waters in the middle and lower bay became more acidic. No long-term trend in the pH of the surface waters of the middle and lower bay was observed, as the effects of river alkalinization and ocean acidification mixed and essentially canceled each other out.

They found that river alkalinization had twice the effect on the Chesapeake Bay’s long-term pH trends compared with ocean acidification. Both processes played a greater role than coastal eutrophication did.

The researchers say their results suggest the potential effectiveness of ocean alkalinity enhancement, a geoengineering technique that adds alkaline minerals to the ocean, for increasing carbon dioxide removal from the atmosphere. (AGU Advances, https://doi.org/10.1029/2024AV001350, 2025)

—Madeline Reinsel, Science Writer

Citation: Reinsel, M. (2025), River alkalinization and ocean acidification face off in coastal waters, Eos, 106, https://doi.org/10.1029/2025EO250196. Published on 21 May 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.

A mysterious and highly active undersea volcano off the Pacific Coast could erupt by the end of this year, scientists say.

The Landslide Blog is written by Dave Petley, who is widely recognised as a world leader in the study and management of landslides.

News this morning from Blatten in Switzerland is that the landslide on Kleiner Nesthorn has continued to develop. Blue News reports that:

“The situation in Blatten in the Valais Lötschental remained tense during the night to Wednesday. According to a spokesperson for the Lötschental regional command post, there were further small rockfalls. The pile of rubble on the Birch Glacier had grown.“

“There is still a lot of movement on the Kleiner Nesthorn. A constant rumbling could be heard during the night, said the spokesperson at the request of the Keystone-SDA news agency.”

During yesterday, rockfalls occurred continuously, and it appears that a substantial volume of material has now been evacuated from the site.

Swisstopo has made aerial photography of the site publicly available – the best way to view this is on their visualisation tool.

The Google Earth image below shows the site, with Blatten in the valley and the marker located in the path of the part of the slope that has failed to date:-

Google Earth image from 2022 showing the site of the landslide on Kleiner Nesthorn above Blatten in Switzerland.

There is much interest in the potential behaviour of the small ice sheet – the  Birch Glacier – just below the marker on the image above, which has accelerated during this period. Rockfall debris falling onto glaciers can cause a change in their behaviour. Whilst I don’t have full details, the concern is likely to focus on either the glacier destabilising and collapsing into the valley, or a large landslide entraining the glacier to form a rock and ice avalanche.

Neither of these scenarios is inevitable (indeed, very little is truly inevitable at this stage), but they would be at the upper end of the range of severity of potential events.

Melaine Le Roy continues to provide excellent updates on the events via his BlueSky feed. Yesterday morning, he posted this animation of the development of the failure:-

https://bsky.app/profile/subfossilguy.bsky.social/post/3lplsxku2a222

This provides a fantastic illustration of the scale of the landslide that is developing above Blatten.

Finally, AZPost has some amazing footage of “smaller” collapses occurring on the mountain:-

This includes this still of the upper part of the collapsing slope:-

The upper part of the landslide on Kleiner Nesthorn above Blatten in Switzerland. Still from a video posted to Youtube by AZPost.

The video shows that there is a long way to go before this event is over.

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.

SummaryThe South China block (SCB) formed after the collision between the Yangtze craton and the Cathaysia block along the Jiangnan orogenic belt at 980–820 Ma. Afterward, intense tectonic activity occurred in the SCB in the early Paleozoic and early Mesozoic. In the Mid–Late Jurassic and Early Cretaceous, the SCB experienced vigorous magmatic activity, which resulted in assemblage of mineralogenetic materials in the Youjiang basin (YB), Southeast coastal metallogenic belt (SCMB) and Wuyishan metallogenic belt (WMB). The mechanisms involved in the formation of various types of metallic ore deposits have attracted considerable attention. However, the crustal and uppermost mantle dynamics of the metallogenic mechanisms are still controversial. To address this issue, we conducted seismic tomography to image the velocity and azimuthal anisotropy of the crust and uppermost mantle beneath the SCB. In this study, an eikonal equation-based traveltime tomography method was used to invert a total of 143,473 high-quality P-wave first arrivals, which were obtained by manually picking the seismic waveforms of 3615 regional earthquakes recorded by 892 broadband seismic stations. After the inversion, we derived high-resolution images, in which we identified a strong low-velocity anomaly and weak azimuthal anisotropy in the uppermost mantle of the northern YB. Below the SCMB, a low-velocity body extends from the uppermost mantle to the bottom of the crust; the azimuthal anisotropy of the uppermost mantle is weak and does not exhibit a consistent fast-velocity direction (FVD). These characteristics can be attributed to the upwelling of hot materials and crustal partial melting. For both the northern YB and SCMB, the low-velocity anomaly is probably related to hot property of the uppermost mantle and weak azimuthal anisotropy may be due to the nearly vertical α-axis of olivine. These features indicate the upwelling of hot materials beneath the YB. The upwelling of hot materials carried metal elements from deep mantle to shallow crust, resulting in metal deposits. Beneath the WMB, the lower crust and uppermost mantle show high-velocity anomalies and moderate strong azimuthal anisotropy with a consistent NE–SW-oriented FVD. The high-velocity anomalies reflect cold and rigid properties of the lower crust and the uppermost mantle beneath the WMB; consistent FVD of azimuthal anisotropy may indicate ancient fossil anisotropy. These features suggest ancient continental relicts of the Cathaysia block under the WMB.

SummaryNumerical simulations of earthquakes and seismic wave propagation require accurate material models of the solid Earth. In contrast to purely elastic rheology, poroelasticity accounts for pore fluid pressure and fluid flow in porous media. Poroelastic effects can alter both the seismic wave field and the dynamic rupture characteristics of earthquakes. For example, the presence of fluids may affect cascading multi-fault ruptures, potentially leading to larger-than-expected earthquakes. However, incorporating poroelastic coupling into the elastodynamic wave equations increases the computational complexity of numerical simulations compared to elastic or viscoelastic material models, as the underlying partial differential equations become stiff. In this study, we use a Discontinuous Galerkin solver with Arbitrary High-Order DERivative time stepping (ADER-DG) of the poroelastic wave equations implemented in the open-source software SeisSol to simulate 3D complex seismic wave propagation and 3D dynamic rupture in poroelastic media. We verify our approach for double-couple point sources using independent methods including a semi-analytical solution and a finite-difference scheme and a homogeneous full-space and a poroelastic layer-over-half-space model, respectively. In a realistic carbon capture and storage (CCS) reservoir scenario at the Sleipner site in the Utsira Formation, Norway, we model 3D wave propagation through poroelastic sandstone layers separated by impermeable shale. Our results show a sudden change in the pressure field across material interfaces, which manifests as a discontinuity when viewed at the length scale of the dominant wavelengths of S- or fast P-waves. Accurately resolving the resulting steep pressure gradient dramatically increases the computational demands, requiring high-resolution modeling. We show that the Gassmann elastic equivalent model yields almost identical results to the fully poroelastic model when focusing solely on solid particle velocities. We extend this approach using suitable numerical fluxes to 3D dynamic rupture simulations in complex fault systems, presenting the first 3D scenarios that combine poroelastic media with geometrically complex, multi-fault rupture dynamics and tetrahedral meshes. Our findings reveal that, in contrast to modeling wave propagation only, poroelastic materials significantly alter rupture characteristics compared to using elastic equivalent media since the elastic equivalent fails to capture the evolution of pore pressure. Particularly in fault branching scenarios, the Biot coefficient plays a key role in either promoting or inhibiting fault activation. In some cases, ruptures are diverted to secondary faults, while in others, poroelastic effects induce rupture arrest. In a fault zone dynamic rupture model, we find poroelasticity aiding pulse-like rupture. A healing front is induced by the reduced pore pressure due to reflected waves from the boundaries of the poroelastic damage zone. Our results highlight that poroelastic effects are important for realistic simulations of seismic waves and earthquake rupture dynamics. In particular, our poroelastic simulations may offer new insights on the complexity of multi-fault rupture dynamics, fault-to-fault interaction and seismic wave propagation in realistic models of the Earth’s subsurface.

SummaryIn this work we apply a dedicated 4D full waveform inversion workflow to short offset streamer data from the Sleipner CO2 storage field in the North Sea. We consider a baseline dataset acquired in 1994 and a monitor dataset acquired in 2008. Accessing to only short offset data raises significant difficulties for full waveform inversion. In this case the penetration of diving waves, which controls the depth where quantitative updates of the velocity can be expected, do not reach the zone of interest where the CO2 is injected. For this reason, we propose to combine an efficient time-lapse full waveform inversion strategy, which we call simultaneous, with a reflection oriented full waveform inversion workflow. The latter has been introduced in the literature as a way to circumvent short-offset limitation and increase the ability of full waveform inversion to update the macro-velocity model at depth by exploiting the reflection paths, using a prior step of impedance reconstruction. We first illustrate the interest of this combined strategy on a 2D synthetic model inspired from the Sleipner area. Then we apply it to the Sleipner field data, using as baseline model the one we present in a companion paper, where our reflection oriented workflow is presented. Our combined approach yields reliable estimates of the changes due to the CO2 injection, characterized by velocity reductions of up to 400 m.s−1 and strong impedance contrasts at depths of 800-1000 m, which consistent with previous FWI studies. Furthermore, the spatial distribution of CO2 changes aligns with conventional seismic time-migration results from earlier studies, following a north-south migration trend.

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

Officials from the Food and Drug Administration (FDA) announced on Tuesday that only adults older than 65 and people with specific medical conditions will be considered eligible for COVID-19 vaccinations this fall.

Healthy Americans younger than 65 may be eligible for vaccine boosters depending on the outcomes of new clinical trials.

FDA commissioner Marty Makary and Vinay Prasad, director of the FDA’s Center for Biologics Evaluation and Research, published the agency’s new framework in the New England Journal of Medicine. They wrote that that the United States’ existing policy is more aggressive than those in European nations and Canada, most of which recommend COVID-19 boosters primarily for older adults and those classified as high-risk.

“The FDA will approve vaccines for high-risk persons and, at the same time, demand robust, gold-standard data on persons at low risk,” they wrote.

Makary and Prasad used the Center for Disease Control and Prevention’s (CDC) definition of “high risk” conditions, which includes asthma, cancer, cystic fibrosis, diabetes, heart disease, pregnancy, and tuberculosis.

The FDA also approved a new COVID-19 vaccine from Novavax on 17 May, with similar limitations.

 
Related

• FDA tightens requirements for COVID vaccine, adding trials for healthy adults
•  New Trump vaccine policy limits access to COVID shots
•  F.D.A. Poised to Restrict Access to Covid Vaccines
•  Get Involved: AGU Science Policy Action Center

Fewer Americans are opting into COVID boosters, with CDC data reporting that just 23% of adults and 13% of those under 18 had received the 2024-2025 vaccine as of 26 April.

However, COVID-19 still presents a danger. The CDC estimates that between 1 October 2024 and 10 May 2025, there were 260,000 to 430,000 COVID-19 hospitalizations and 30,000 to 50,000 COVID-19 deaths. Research has found that environmental factors, such as exposure to air pollution and proximity to gas and oil wells, can increase the likelihood or severity of the disease.

“This an anti-science move that will kill more Americans,” Lucky Tran, a scientist and public health communicator based in New York, wrote on Bluesky.

Under its new leadership — who spread misinformation throughout the pandemic — the FDA announced it will limit all Covid vaccines to adults over 65 and those with certain medical conditions.Covid continues to spread and cause harm. This an anti-science move that will kill more Americans.

— Dr. Lucky Tran (@luckytran.com) 2025-05-20T17:43:11.744Z

—Emily Dieckman (@emfurd.bsky.social), Associate Editor

These updates are made possible through information from the scientific community. Do you have a story about how changes in law or policy are affecting scientists or research? 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.

As wildfires continue to ravage regions from Los Angeles to South Korea, a new study featured on the cover of Advances in Atmospheric Sciences sheds light on the large-scale climate patterns influencing these devastating global extreme events.

The Arctic is one of the coldest places on Earth, but in recent decades, the region has been rapidly warming, at a rate three to four times faster than the global average. However, current climate models have been unable to account for this increased pace.

Source: Geophysical Research Letters

Artificial intelligence (AI) algorithms can produce weather predictions more quickly than traditional algorithms for a fraction of the computational cost. But because training AI takes such large amounts of data, it has so far been most successful at producing global-scale forecasts. Until recently, researchers lacked the data needed to train algorithms to predict small-scale weather patterns such as thunderstorms.

Flora and Potvin extended AI-based weather forecasting to thunderstorm-scale events by training Google’s neural network GraphCast on data from NOAA’s Warn-on-Forecast System. The Warn-on-Forecast research project generates high-resolution forecasts for areas likely to experience extreme weather with the aim of issuing earlier warnings for tornadoes, severe thunderstorms, and flash floods.

The AI model, named WoFSCast, learned the dynamics of key thunderstorm features, including updrafts, which feed thermodynamic energy into storms, and cold air pockets that form beneath storms, which influence how storms move and grow.

The model yielded largely accurate predictions of how storms would evolve for up to 2 hours; these predictions matched 70% to 80% of those generated by the Warn-on-Forecast system. The process of generating a prediction took only 30–40 seconds using one graphical processing unit. That’s at least a factor of 10 faster than using the current Warn-on-Forecast System to generate forecasts without AI.

With additional training data, the researchers suggest that WoFSCast could become even more versatile, predicting surface winds and rainfall within landfalling tropical cyclones, as well as how wildfires will spread, for instance. By using an AI-enhanced system, the National Weather Service may be able to issue severe weather warnings more quickly and reduce the harm caused by these extreme events. (Geophysical Research Letters, https://doi.org/10.1029/2024GL112383, 2025)

—Saima May Sidik (@saimamay.bsky.social), Science Writer

Citation: Sidik, S. M. (2025), Storm prediction gets 10 times faster thanks to AI, Eos, 106, https://doi.org/10.1029/2025EO250159. Published on 20 May 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.

For decades, wounds have surfaced in the Peruvian Amazon where the Rio Inambari merges with the Rio Madre de Dios, carving thick gashes into ancient tree stands. These almost lunar scars of barren rubble were formed at the hands of a growing enterprise capitalizing on gold that the rivers deposit throughout their floodplains.

A new study published in Environmental Research Letters shows that the spread of this devastation has quickened and increasingly affected a unique Amazonian ecosystem. Peatland swamps safeguard meters-deep deposits of carbon accumulated over millennia and contain unique assemblages of life distinct from the surrounding rain forest.

Around 15 years ago, “there was mining in peatlands, but it was the exception.”

Around 15 years ago, Ethan Householder first visited the Madre de Dios region, where 70% of artisanal gold mining takes place in Peru. Peatlands are dispersed in small pockets throughout the enormous Madre de Dios floodplain, which stretches into Bolivia, where the river’s confluence with the Mamore forms the Madeira and eventually empties into the Amazon itself.

Householder, a community ecologist at Germany’s Karlsruher Institut für Technologie and a study coauthor, recalled that at that time, “there was mining in peatlands, but it was the exception.”

Now, he and his colleagues have found that peat mining is surging in the region.

To determine how artisanal gold mining (defined as subsistence or small scale and often illegal) has affected peatlands along the Rio Madre de Dios, the team searched through decades of satellite imagery for sudden drops in the greenness of the forest canopy that might signal deforestation. Then, they looked for spectral information indicative of mining activity: the buildup of gravel and sand, for example, and the presence of water-filled pits at the site of formerly forested land.

Using a machine learning algorithm trained to pick out pixels that met those criteria, the team combed through imagery from 1985 to 2023. They found that more than 11,000 hectares of forest along the river had been converted to mines, with most of the growth taking place since the mid-2010s.

Studies like this “are all puzzle pieces of evidence saying this is a huge issue and still not resolved.”

The analysis found more than 550 hectares associated with mining activity in peatlands. Though that’s just a fraction of the total area that’s been mined, research showed that peatland mines have expanded faster than mining in the forest at large in recent years. (Fifty-five percent of peatland loss occurred within the past 2 years.)

Already, digging up these peatlands has released anywhere from 200,000 to 700,000 metric tons of carbon stored belowground—in addition to the carbon released from the loss of trees and plant life above it. If all the peat in the Madre de Dios region is lost, some 17 million metric tons of long-sequestered carbon could be released.

Mining in the Amazonian peatlands “is basically the entire force of global capitalism on top of one of the most carbon-rich habitats on Earth,” Householder said.

The new work continues and expands the story of how gold mining is degrading the Amazon, said Greg Asner, a conservation ecologist at Arizona State University who has been studying the effects of gold mining on the Peruvian Amazon since the early 2010s. To him, studies like this “are all puzzle pieces of evidence saying this is a huge issue and still not resolved.”

—Syris Valentine (@shapersyris.bsky.social), Science Writer

Citation: Valentine, S. (2025), Artisanal gold mining is destroying Amazonian peatlands, Eos, 106, https://doi.org/10.1029/2025EO250195. Published on 20 May 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: Earth’s Future

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

随着气候变化持续推动全球海平面上升，许多生活在沿海地区的居民已感受到其影响。海岸侵蚀正在加速，海岸线向内陆移动，风暴潮也愈演愈烈。但隐藏在地表之下的还有另一个迄今为止鲜为人知的严重后果：地下水位上升。

有证据表明，在一些地势低洼、地下水位较浅的沿海地区，海平面上升将导致地下水位同步上升，这可能会给住宅、企业和其他基础设施带来严重风险。

在一篇聚焦新西兰沿海城市达尼丁的新论文中，Cox等人展示了一种预测海平面上升如何改变地下水位，从而增加内陆洪涝灾害的方法。达尼丁南部已经经历了周期性洪涝灾害，随着海平面上升，洪涝灾害将变得更加严峻。研究人员将该城市描述为新西兰社区应对和适应气候变化和海平面上升的典范。

研究人员使用了2019年至2023年的数据，这些数据来自安装在达尼丁低洼沿海地区的35个地下水传感器网络，该市的大部分基础设施都位于该区域。他们将传感器数据与潮汐、降雨和其他因素的数据进行比较，来预测未来海平面上升对地下水的影响。

研究结果表明，海平面上升首先会导致地下水位上升，从而降低土地吸收降雨的能力。随着海平面继续上升，地下水位可能会进一步上升，并在地下水位以下造成问题，例如淹没污水处理系统、渗入地下室以及破坏建筑物地基。最终，地下水可能会上升到足够高的地方，形成泉水，引发洪水。

研究人员得出结论，地下水位上升造成的洪水灾害可能向内陆延伸到比许多人预期的更远的地方。此外，假设达尼丁沙丘屏障的防护地形不发生重大变化，这些地下水效应将比海平面上升直接造成的洪水更早发生。

研究人员指出，他们的方法包含关键的假设和不确定性——例如，地下水和海平面将以相同的速度上升，地下水位将保持大致相同的形状，但保守的预测对于达尼丁的规划和灾害管理具有重要价值。他们表示，由于该方法相对简单且成本低廉，因此也可以应用到世界各地类似的沿海地区。 (Earth’s Future, https://doi.org/10.1029/2024EF004977, 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.

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

River discharge is an important variable for a wide range of water management applications, yet many rivers (even in the United States) remain ungauged or under-gauged due to the difficulty of conventional field methods, especially in regions of complex terrains. Existing remote sensing methods need gauge data for calibration and are subject to other limitations.

Legleiter et al. [2025] propose a new image-based method to infer river discharge based on critical flow theory. Specifically, slope transitions (from steep to mild) or channel constrictions can induce critical flow conditions, causing undular hydraulic jumps in the form of well-defined standing wave trains. For critical flow (with Froude number equal to 1), the spacing of the waves, the velocity of the flow, and the depth of the water are all uniquely related to one another, so the discharge can be calculated from basic measurements of wavelength and channel width.

The newly proposed method is used to derive discharges from 82 Google Earth images, which agreed closely with gauge records, providing preliminary confirmation for the reliability of the method. Although the method is only applicable to rivers with standing waves (which typically occur on steep slopes or near channel constrictions), these conditions are frequently met in mountainous regions, exactly where new monitoring methods are the most needed due to the lack or sparsity of gauge stations. This study provides a foundation for further evaluation and refinement of the theoretically grounded approach to river discharge estimation.

Citation: Legleiter, C. J., Grant, G., Bae, I., Fasth, B., Yager, E., White, D. C., et al. (2025). Remote sensing of river discharge based on critical flow theory. Geophysical Research Letters, 52, e2025GL114851. https://doi.org/10.1029/2025GL114851   

—Guiling Wang, Editor, Geophysical Research Letters

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.

Efforts to limit the global temperature increase to 1.5°C under the Paris Climate Agreement may not go far enough to save the world's ice sheets, according to a study published in the journal Communications Earth & Environment.

The Landslide Blog is written by Dave Petley, who is widely recognized as a world leader in the study and management of landslides.

In Switzerland, a dramatic rock slope failure is developing above Blatten [46.4128, 7.7987], a village located in Vallais Canton.

Blue News is providing regular updates on a dedicated website. The drama started at the weekend with a major landslide in the Petit Nesthorn area, which impacted and entrained a part of the Birch glacier. This has resulted in evacuation of the majority of the population of Blatten.

There is little doubt that a major instability has developed. The estimated scale of the instable mass is up to 5 million cubic metres. At least 17 metres of displacement have been recorded in the last few days.

Melaine Le Roy is providing detailed coverage of the evolution of the event on BlueSky. Hopefully, you’ll be able to view one of his posts below, which shows the extraordinary scale of the mobile mass:-

INSANE !!

The Amazon Basin lost about 27,000 square kilometers of forest each year from 2001 to 2016. By 2021, about 17% of the basin had been deforested.

The near-bottom water in the U.S. Northeast continental shelf provides a critical cold-water habitat for the rich regional marine ecosystem. This "cold pool" preserves winter temperatures, even when waters become too warm or salty elsewhere during the summer.