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

After 18 years of data collection, quality control, processing, and archiving, the United States Magnetotelluric Array (USMTArray) data set was completed in 2024. A new article in Reviews of Geophysics introduces this unprecedented data set and a new high-resolution model of the Earth’s crust and upper mantle that was made possible because of it. Here, we asked the authors to give an overview of magnetotellurics, how the USMTArray was developed, and future directions for research.

In simple terms for a non-specialist, what is the science of magnetotellurics?

Magnetotellurics (MT) is a passive geophysical technique capable of imaging the subsurface from hundreds of meters to hundreds of kilometers depth using the Sun and global lightning as sources. The science behind MT is largely based on Faraday’s law of induction, where external magnetic field variations induce telluric (from the Latin word ‘tellus’ meaning Earth) currents in the conducting Earth. These magnetic field variations are constantly occurring and happen over a wide range of time scales ranging from milliseconds to hours. And they are tiny – typically on the order of 0.1% of Earth’s magnetic field amplitude and even during intense magnetic storms rarely exceed 1%. 

By measuring these magnetic variations, and the induced electric field variations at Earth’s surface, we can constrain the 3D distribution of conductivity in the Earth. MT is an elegant method – we exploit powerful and distant energy sources which we have no control over and can mathematically remove the stochastic source spectrum to recover reliable estimates of Earth impedance. Impedance can be thought of as the Earth filter – a complex, frequency dependent set of functions that encapsulates all the information about the 3D conductivity structure beneath our feet. Through numerical inversion of impedance data at an array of sites, we build up 3D models of electrical conductivity.

What are some of the applications of the magnetotelluric method?

MT is applied across a broad spectrum of the Earth and space sciences ranging from mineral and geothermal resource investigations, to fundamental geologic and tectonic studies, to imaging the magmatic plumbing systems of active volcanoes, and to hazard mapping centered upon geomagnetically induced currents and the risk they pose to power grids.

Studies using MT are performed on every continent and in all tectonic settings, on land and on the ocean floor, on the Antarctic ice sheet, and even on the Moon. Because of its ability to image the entire lithospheric column, MT studies have made important contributions to our understanding of continental assembly by revealing ancient orogens and rifts. Moreover, MT is uniquely able to constrain the stability of cratonic roots by mapping hydration of the mantle lithosphere. MT studies are key to understanding active tectonic processes, including constraining the water budget in subduction zones, imaging melt zones beneath orogenic plateaus, and mapping the extent of crustal extension – for example beneath the western U.S.

With the rise of computational power and 3D modeling and inversion codes, MT is now routinely used to study complex 3D systems, such as active volcanoes, geothermal systems, and mineral deposits. The sensitivity of MT to minor conductive phases – be it partial melt, clay, or conductive minerals such as graphite and metallic sulfides – make it ideal for studying these types of systems. As a result, MT is commonly employed within the resource sector at both the district and deposit scale. Many of the world’s iconic volcanoes have also been imaged with MT, where they constrain the geometry of crustal melt reservoirs – especially their volume and melt fraction which is in turn linked to the eruptibility of a subsurface magma. These analyses are especially powerful because they are sensitive to a distinct physical parameter – resistivity – of Earth materials. MT therefore provides unique and complementary information about the subsurface across a wide range of scales and is a particularly invaluable tool when other methods yield non-unique interpretations. 

One somewhat unexpected application of MT has been to space weather hazards. It was recognized a little over a decade ago that MT impedances are key to estimating surface electric fields generated during intense geomagnetic storms that can impact electric power grids. Past storms have knocked out power to vast areas and damaged critical infrastructure such as transformers. The importance of MT data to scenario analysis, in which power grid components are ‘stress tested’ against past geomagnetic storms, cannot be overstated. Regional to national-scale geoelectric hazard maps, both in the U.S. and internationally, are also informed by MT data, as are real-time geoelectric hazard estimates.

What is the United States Magnetotelluric Array (USMTArray)?

The USMTArray was an ambitious program begun in 2006 under the NSF-funded EarthScope program and completed in June of 2024 under USGS funding. The USMTArray collected long-period MT soundings on a 70-km grid across the contiguous U.S. – totaling more than 1,800 stations – each collected with uniform instrumentation, acquisition parameters, data processing, archiving, and metadata. Funded throughout its 18-year lifetime by three different federal agencies (the NSF, NASA, and USGS working closely with the Incorporated Research Institutions for Seismology and Oregon State University), the data – time series, response functions and metadata – were released incrementally to the public without data embargo or usage restriction.

In broad terms, how was the USMTArray developed?

The USMTArray had humble beginnings – being mentioned in early planning documents as having value in understanding subduction zones and characterizing volcanic systems. Funded by NSF in 2003, the MT component of EarthScope was modeled after the much larger seismic component, with a transportable array of instruments to march across the U.S. on a 70-km spaced grid and a backbone array of seven instruments to study deep mantle structure. The USMTArray started off small and before dedicated instruments were even available. In 2006, a pilot study collected the first 30 stations in eastern Oregon using borrowed instruments, while subsequent years expanded what became known as the ‘northwest’ footprint, a 331-sites array completed in 2011 encompassing the Yellowstone-Snake River Plain, the Northern Rocky Mountains, the Cascades magmatic arc, and the northern Basin and Range province. Subsequent footprints in the midcontinent and the eastern U.S. continued to expand coverage.

What were some of the challenges in developing the USMTArray?

The biggest challenge by far was money. Within the EarthScope program, the USMTArray was never funded at the level needed to cover the contiguous U.S. The MT component was instead carried out as a series of footprints in areas deemed most scientifically advantageous. This limitation, however, led to one of the big successes of the USMTArray – active community engagement. Siting workshops held in 2008 and 2013 brought together participants from academia, government, and industry to discuss and prioritize where the array would go next, while a community working group provided scientific and operational guidance throughout the life of the array. The success of the USMTArray was recognized early on by the community governance of the EarthScope facility activities, with the ‘full-48’ concept endorsed in 2009, leading to modest increases in funding and an acceleration of station completions. In 2018, by the end of NSF-sponsored activities, roughly 2/3 of the contiguous U.S. had been covered. Seeing the array to completion, however, required additional funding, a challenge met by NASA (2019-2020) and the USGS (2020-2024), in large part due to recognition of the importance of USMTArray data to space-weather hazards and supported through executive orders in 2016 and 2019.

Another notable challenge that we faced while developing the USMTArray operations was the absence of established data sharing practices within the magnetotelluric community. Indeed, the concept of FAIR data was only introduced in 2016. Back in 2006 when this program commenced, the concepts of open data and systematic data sharing were largely unfamiliar, and no widely adopted, sustainable data formats existed. Available data formats were lacking in flexibility, consistency, and self-descriptive metadata. As the project progressed, our team developed such formats and accompanying databases, which have now reached maturity and are helping to drive more sustainable MT data‑sharing practices internationally.

How has the development of the USMTArray advanced the scientific field?

The USMTArray, along with parallel advances in modeling capabilities and increased computational power, ushered in a jump to 3D MT and to interrogating the Earth at regional to national scales. National-scale conductivity models, such as those developed from the USMTArray, now join the ranks of other data sets like magnetic, gravity, and seismic, and are a new lens with which to view the architecture of the North American continent. Numerous contributions to continental architecture and assembly and to understanding active tectonic processes have come from the USMTArray.

The USMTArray also serves as a framework for more detailed studies, allowing Principal Investigators (PIs) to derisk future surveys and industry to investigate anomalous or unexpected structure. Studies of the Cascadia subduction zone and the adjacent magmatic arc and geothermal energy prospectivity studies in the Oregon Cascades and Great Basin have been built upon the USMTArray while new MT surveys along the eastern seaboard are collecting high-resolution MT data to improve space-weather hazard maps over areas identified as particularly at risk from the analysis of USMTArray data.

Beyond the data and models derived from them, the USMTArray has motivated methodological advances, led to an investment in MT instrumentation and open-source software for researchers within the NSF-supported National Geophysical Facility, and served as a model for other regional and continental scale MT experiments.

What are some of the future directions for research in continental scale magnetotellurics?

With completion of the USMTArray, and the 3D conductivity models derived from it, there are numerous avenues for future research. Most models of continental evolution, for example, were developed prior to the advent of this rich data set. Critically evaluating such models in light of this new data set is paramount, and initial studies are already forcing a reexamination of certain paradigms.  

Multi-disciplinary studies incorporating geochronology, geochemistry, and rapidly evolving seismic models is another promising area as is the coupling of geophysical models to geodynamic models to examine the evolution of newly imaged model structure. Similarly, advancements in integrated and joint inversion are promising directions to leverage the wealth of public data sets available at regional to continental scales.  

Geology doesn’t stop at national borders or the land-sea interface – additional opportunities exist for cross-border arrays and onshore/offshore MT studies. Investigation of subduction zone processes and rifted continental margins by their very nature demand an amphibious approach.

On the applied front, resource assessments increasingly are applied at national and even global scales and demand data support at these same scales. Mineral resource assessments, for example, in the U.S., Canada, and Australia are exploring machine learning approaches to map prospectivity for various deposit types and incorporate a range of geophysical data layers to do so. Similarly, geothermal assessments can benefit from the consistent and synoptic data coverage offered by USMTArray data and models.

Finally, on the space-weather hazards front, partnering with power-system engineers to investigate data scale and uncertainty shows promise in generating accurate hazard maps and in improving upon operational, near real-time geoelectric field models. For all these future research directions the USMTArray remains both a framework and a benchmark upon which to build.

—Paul A. Bedrosian (pbedrosian@usgs.gov; 0000-0002-6786-1038), U.S. Geological Survey, United States; Anna Kelbert (anna.kelbert@cfa.harvard.edu; 0000-0003-4395-398X), Center for Astrophysics | Harvard & Smithsonian, United States; Adam Schultz (adam.schultz@oregonstate.edu; 0000-0003-1663-1547), Oregon State University, United States; and Gary D. Egbert (gary.egbert@oregonstate.edu; 0000-0003-1276-8538), Oregon State University, United States

Editor’s Note: It is the policy of AGU Publications to invite the authors of articles published in Reviews of Geophysics to write a summary for Eos Editors’ Vox.

Citation: Bedrosian, P. A., A. Kelbert, A. Schultz, and G. D. Egbert (2026), Mapping the hidden electrical anatomy of a continent, Eos, 107, https://doi.org/10.1029/2026EO265021. Published on 26 May 2026.

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 © 2026. 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: Geophysical Research Letters  

Real-time hydrologic forecasting predicts river level and flooding inundation by combining continuously updated rainfall measurements, river gauge readings, and weather forecasts. Most of these flood forecasting systems depend on human interpretation and adjustments, or a “forecasters-in-the-loop” approach, which pairs computer models with a human expert on flood dynamics and local conditions. In contrast, in a “forecasters-over-the-loop” system, humans supervise automated forecasts and intervene only if necessary.

Recently, artificial intelligence (AI) and machine learning (ML) have become more integrated into flood prediction, and many of these systems are faster at processing large datasets and learning complex patterns from historical records than traditional models alone. But these new technologies also come with limitations—AI and ML require extensive data and may struggle to capture extreme, rare events.

Even though ML and AI are often touted as the future of flood forecasting, most studies have tested this technology against models that provide historical simulations, not the real-time operational systems that would be used during a flood. These simplified models may lack local details or are tested at daily rather than hourly resolution. Their effectiveness may be overestimated. 

Tran et al. produced the first study comparing the performance of ML models to an actual flood forecasting system used at the California Nevada River Forecast Center (CNRFC) that uses professional forecasters and traditional hydrologic models. The study suggests that a forecasters-in-the-loop approach outperforms the ML models in several key ways, including streamflow predictions and flood event detection, because forecasters can recognize model errors and account for poor input data—actions models cannot take on their own.

The researchers used data gathered from CNRFC river stage forecasts across 50 California and Nevada locations between 2012 and 2022 and river condition lead times from 1 to 96 hours. Compared to the ML models, the Community Hydrologic Prediction System used at CNRFC generally performed better at predicting stream flow and flood peaks, especially with longer lead times. Though the ML models could perform better at very short lead times, their accuracy declined quickly. Though automated forecasting options may seem promising, they are not yet a suitable replacement for human expertise when it comes to protecting lives and livelihoods from damaging floods, the researchers say. (Geophysical Research Letters, https://doi.org/10.1029/2025GL118317, 2026)

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

Citation: Owen. R. (2026), Keeping humans in the loop improves flood forecasting, Eos, 107, https://doi.org/10.1029/2026EO260161. Published on 19 May 2026.

Text © 2026. 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: Tectonics

Scientific progress rarely follows a straight path. Instead, it develops through open discussion, critical evaluation, and the testing of new ideas. The exchange between authors and colleagues illustrates how this process unfolds in modern Earth sciences and provides a valuable example of constructive scientific debate.

At the center of the discussion lies a fundamental question about one of Earth’s most remarkable geological features: how did the Himalaya and the Tibetan Plateau become the highest and largest mountain system on the planet?

In their paper “Raising the Roof of the World: Intra-Crustal Asian Mantle Supports the
Himalayan–Tibetan Orogen,” Sternai et al. [2025] address this question using numerical geodynamic modeling. These computer simulations reproduce the physical behavior of large rock masses deep inside the Earth and allow researchers to investigate the long-term evolution of this vast orogenic system.

Their study specifically explores the possibility that, during the collision between the Indian and Asian plates, layers of mechanically strong Asian mantle rock became embedded within the thickened Indian continental crust beneath the Tibetan Plateau. According to this hypothesis, these mantle layers could help sustain the elevation of the Plateau by effectively withstanding stresses over long geological timescales: the Indian crust would provide buoyancy (raising the roof), while the Asian mantle would contribute mechanical strength to support the Himalayan–Tibetan topography.

Hetényi and Cattin disagree with and challenge this interpretation in their Comment. Drawing on a large body of well-established geophysical and geological observations, they argue that the structure beneath southern Tibet is better explained by underthrusting, the process by which the Indian plate slides beneath the Tibetan Plateau. Seismic imaging studies, including receiver-function analyses that use earthquake waves to map subsurface structures, consistently reveal features interpreted as Indian crust and upper mantle extending far north beneath Tibet.

In their Reply, Sternai and colleagues clarify that their models were not intended to accurately reproduce the present-day structure of the region in detail. Instead, they were designed as process-oriented experiments to test whether existing and/or alternative mechanisms for crustal thickening and plateau support are mechanically and rheologically viable.

This exchange highlights an important aspect of contemporary geoscience—observations of Earth’s interior such as seismic images, gravity data, and geological records often allow multiple, non-unique interpretations. Numerical modeling provides a complementary approach by evaluating whether proposed geological mechanisms are physically plausible.

Equally significant is the tone of the discussion itself. The Comment and Reply show how scientists, while strongly disagreeing about interpretations, can maintain a constructive and respectful dialogue. Such approach fuels scientific advance by encouraging the community to re-examine established assumptions, refine models, and integrate new observations.

Debates like this one, therefore, extend well beyond a specific geological question. They illustrate how scientific understanding advances through the interplay of observations, theoretical reasoning, and modeling experiments.

In this way, the dialogue highlighted here contributes not only to our understanding of the Himalayan–Tibetan mountain system but also to the broader methodology of Earth science.

Citations

Sternai, P., Pilia, S., Ghelichkhan, S., Bouilhol, P., Menant, A., Davies, D. R., et al. (2025). Raising the roof of the world: Intra-crustal Asian mantle supports the Himalayan-Tibetan orogen. Tectonics, 44, e2025TC009057. https://doi.org/10.1029/2025TC009057

Hetényi, G., & Cattin, R. (2026). Comment on “Raising the roof of the world: Intra-crustal Asian mantle supports the Himalayan-Tibetan orogen” by Sternai et al. Tectonics, 45, e2025TC009214. https://doi.org/10.1029/2025TC009214

Sternai, P., Pilia, S., Ghelichkhan, S., Bouilhol, P., Menant, A., Ostorero, L., et al. (2026). Reply to comment by Hetényi and Cattin on: “Raising the roof of the world: Intra-crustal Asian mantle supports the Himalayan-Tibetan orogen”. Tectonics, 45, e2026TC009436. https://doi.org/10.1029/2026TC009436

—Giulio Viola, Editor-in-Chief, Tectonics

Text © 2026. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Editors’ Highlights are summaries of recent papers by AGU’s journal editors.

Source: AGU Advances

The evolution of rivers that split into multiple channels is a scientific challenge in terms of modeling and prediction. On the other hand, these rivers are widespread and play a key role for ecosystems’ life, groundwater recharge, and therefore, water security. They are also extremely sensitive to hydroclimatic changes, leading to shifts in precipitation, erosion and sediment transport.

Zhao et al. [2026] investigate the drivers of river evolution for 97 multithread river reaches worldwide, spanning diverse climates and morphologies. The study reveals the key role of intermittency for river evolution. In particular, higher flow intermittency could lead to more even flow partitioning among threads, therefore impacting hydrology and ecosystems. With flow variability increasing after climate change, rivers are likely to increase their thread count, thus impacting livelihoods and ecosystems.

Citation: Zhao, F., Ganti, V., Chadwick, A., Greenberg, E., McLeod, J., Liu, Y., et al. (2026). Global hydroclimatic controls on multithread River dynamics. AGU Advances, 7, e2025AV002166. https://doi.org/10.1029/2025AV002166

—Alberto Montanari, Editor-in-Chief, AGU Advances

Text © 2026. 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.

When it comes to thriving at high elevation, diminutive plants are always a safe bet. And low-lying vegetation is in fact colonizing higher and higher reaches as the climate changes, new results reveal. Researchers analyzed more than 2 decades’ worth of satellite data and showed that the vegetation line in the Himalayas is moving upward, in some cases by up to several meters per year. These changes have implications for the hydrology of the region and therefore for water resources for the population centers located downstream, the team reported last month in Ecography.

Mountains and People

The Himalayas, with their massive stores of frozen water, are part of a region known as the planet’s “Third Pole.” Nearly a billion people rely on water sourced from this area, but the Himalayas aren’t immune to climate change—shifts in temperature and precipitation patterns are causing glaciers to melt and permafrost to thaw, among other effects. “The Himalayan mountains are experiencing a lot of ecosystem changes,” said Ruolin Leng, an Earth scientist who led this new research while at the University of Exeter in the United Kingdom. She currently works at H2Tab, a wellness company.

And while the macroscopic effects of climate change in mountainous regions—the melting of the aforementioned glaciers, for example—have been readily studied, shifts in vegetation are often overlooked, said Leng. That’s a problem because plant cover affects everything from soil moisture levels to water runoff to the albedo of the planet’s surface, all of which have consequences for how water moves through the larger system, she said. “It’s a very important factor in the hydrological system.”

Leng and her colleagues focused on six sites, each roughly 40,000 square kilometers in size, in Bhutan, Nepal, and politically disputed areas farther west. Altogether the locales spanned roughly 15° in longitude (about the width of a U.S. time zone). The choice to analyze several locations along an east-west gradient was deliberate, said Stephan Harrison, a climate scientist also at the University of Exeter and a member of the research team. “The western Himalayas are very different from the eastern Himalayas in terms of climate. If you’re going to understand climate change across the Himalayas, you can’t just look at one location.”

Spotting Vegetation from Space

For each of those sites, the researchers mined satellite observations collected from 1999 to 2022 by the NASA/U.S. Geological Survey Landsat program. The researchers focused on visible and near-infrared observations to calculate a metric known as the normalized difference vegetation index (NDVI). Vegetation tends to reflect relatively little visible light while reflecting much more near-infrared light, and that fact can be exploited to infer the presence of vegetation in remote sensing data, said Karen Anderson, a remote sensing scientist at the Environment and Sustainability Institute at the University of Exeter and a member of the research team.

After masking out pixels too obscured by clouds or snow to correctly analyze, Leng and her colleagues calculated the NDVI for each 30- × 30-meter Landsat pixel within their study regions. The team retained pixels with NDVI levels above a minimum threshold and used those data, combined with topography information, to estimate the maximum elevation that was reliably vegetated each year. All six sites exhibited upward trends in the elevations of their vegetation lines over time, the researchers found. A site in central Nepal straddling the country’s northern border recorded the largest changes: From 1999 to 2022, the elevation of its vegetation line rose from roughly 5,520 meters to 5,670 meters, an increase of just under 7 meters per year on average. The five remaining sites all recorded annual upward shifts ranging from about 1 to 6 meters per year on average.

“Broadly speaking, plants are moving up mountains,” said Anderson. But different regions are responding differently, she added. (And while similar results have been previously noted in the Himalayas, not all plant life everywhere is moving up—recent research has shown that some tree lines are in fact moving downslope.)

A Climatic Culprit?

To investigate the potential drivers behind these changes, the team studied correlations with three climatic parameters: temperature, total precipitation, and snow depth. These data came from the European Centre for Medium-Range Weather Forecasts reanalysis dataset, which has a spatial resolution of roughly 30 kilometers.

Leng and her collaborators found that their site with the fastest-changing vegetation line also recorded the most rapid increase in snow depth over time. These two changes might therefore be linked, but more work is needed, Anderson admitted. “We haven’t addressed the causal link here. We’ve simply looked for patterns.”

There’s also a significant mismatch in the spatial resolution of the team’s meteorological data and their Landsat data, said Trevor Keenan, an ecosystem scientist at the University of California, Berkeley not involved in the research. Such a discrepancy can be particularly problematic in complex landscapes like mountain ranges because the coarse meteorological data might not be capturing the true microclimates that are bound to persist in such places, he said. “With heterogenous terrain and large elevational gradients, you really need that microclimate information.”

Anderson knows the geographical complexity of the Himalayas firsthand—in 2017 and 2022, she and other scientists conducted fieldwork in Nepal that informed this research. Those trips were a special opportunity to see plants like dwarf rhododendron thriving in tough conditions, she said. And it was a good lesson in appreciating some of the most diminutive members of the plant kingdom, Anderson added. “People neglect the little plants.”

—Katherine Kornei (@KatherineKornei), Science Writer

Citation: Kornei, K. (2026), Vegetation moves upslope across the Himalayas, Eos, 107, https://doi.org/10.1029/2026EO260149. Published on 14 May 2026.

Text © 2026. 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.

Tiny saltwater channels have a big influence on sea ice.

Sea ice typically includes pockets or channels of brine that allow salt water to flow vertically through the ice. When those channels align neatly, they need to make up only about 5% of the ice volume before the water can flow. But in more disordered, granular ice, salt water starts to flow only when the brine channels take up more space—roughly 10% of the ice volume, according to a new study published in Scientific Reports.

This higher threshold could slow the drainage of surface melt ponds, as well as the transport of nutrients to microbial communities inside the ice.

“If we’re trying to find predictive models about how these ice cores are responding under climate change, it’s going to be necessary to take into account these structural and microstructural conditions,” said Stephen Ackley, a sea ice researcher at the University of Texas at San Antonio who was not involved in the study.

Disorderly Constructs

As seawater freezes, it forms a mixture of ice crystals and brine. In calm conditions, the ice slowly grows into long, parallel crystals separated by orderly brine channels. This columnar sea ice is common in the Arctic, and its properties have been widely used in sea ice models.

But in choppy waves or when the ice’s snow-covered surface floods and refreezes, new ice can’t grow into these ordered columns. Instead, it forms small, randomly oriented grains separated by more complex pores containing brine and gases. Called granular ice, this form is more common in Antarctica but is becoming increasingly prevalent in the Arctic as temperatures rise and ice cover thins.

In 1998, University of Utah mathematician Kenneth Golden established the first estimate of the point at which the brine channels are connected enough to allow water to flow in columnar ice, called the percolation threshold. The new work, also led by Golden, extends a similar analysis to granular sea ice.

“It’s the sequel we’ve been waiting decades for,” said Don Perovich, a sea ice researcher at Dartmouth who was not involved in the new work.

To quantify the percolation threshold for granular ice, Golden and his colleagues collected sea ice samples during two expeditions off the eastern coast of Antarctica in 2007 and 2012. They measured how quickly water moved through the brine channels in the ice. After the 2012 expedition, they also mapped the arrangement of ice crystals within the ice blocks to correlate those permeability measurements with the microscale structure of the ice.

The finding that in granular ice, about twice as much of the ice volume needs to be brine for water to flow compared to columnar ice suggests that brine channels within granular ice are much less interconnected.

With the higher threshold, “you have to reassess all these models, anything that relies on fluid flow through sea ice,” if granular ice is present, said Golden. Granular ice will require warmer or saltier conditions to leave enough brine in the ice structure to meet the percolation threshold and allow water to flow vertically.

For example, the new value could influence models of how meltwater ponds behave atop an underlying ice sheet. If meltwater ponds form above a base of granular sea ice, those ponds will require warmer temperatures before they start draining than melt ponds on columnar ice will.

If these melt ponds remain on the surface longer waiting for those warmer temperatures, they could lower the albedo, or reflectivity, of the ice sheet. That could cause the ice sheet to absorb more heat, leading to a feedback loop that could accelerate melting.

The higher percolation threshold could also affect algae that lives within the ice. Ice algae make up an important food source for krill and crustaceans, which in turn become food for fish, penguins, and whales. Algae rely on water flowing through the ice to deliver nutrients. Because granular ice requires warmer temperatures for that flow to start, it could affect the depth at which algae can live inside the ice, Golden said.

Percolation Consideration

Still, experts say more data are needed to establish percolation thresholds across both Arctic and Antarctic ice. The size of the grains in granular ice can vary substantially at different temperatures, under different formation conditions, and between the poles. Larger grains could lower the percolation threshold, allowing water to flow even when the ice contains much less than 10% brine by volume, said Sønke Maus, a scientist studying ice microstructure at the Norwegian University of Science and Technology who was not involved in the study.

“The data that we have at the moment for the granular sea ice is sparse,” Maus said. “You need a big campaign to collect such data.”

Golden said that in future work he also plans to develop models to compute the electromagnetic properties of both columnar and granular sea ice. Knowing these properties can help scientists determine the thickness and age of an ice sheet from satellite data.

—Skyler Ware (@skylerdware), Science Writer

Citation: Ware, S. (2026), Changes in sea ice microstructure could affect climate models, Eos, 107, https://doi.org/10.1029/2026EO260164. Published on 20 May 2026.

Text © 2026. 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: Journal of Geophysical Research: Solid Earth

Magnetic rocks with iron oxide concentrations act as natural chroniclers of Earth’s past continental movements. Using small samples of rocks, scientists can isolate magnetic grains that were frozen in orientation as the rock solidified. The magnetization of these grains acts as a miniature compass needle, pointing toward ancient magnetic poles. This same principle applies to extraterrestrial samples, such as meteorites and lunar rocks, which preserve evidence of the early solar nebula’s evolution.

However, traditional bottle cap–sized bulk samples often contain a mixture of reliable and unreliable magnetic signals, resulting in complex data that hamper interpretation. To improve accuracy, researchers have turned to magnetic microscopy. This technique maps magnetic fields at submillimeter to submicrometer scales in thinly sliced rock sections using advanced tools like a quantum diamond microscope (QDM) or a cryogenic superconducting quantum interference device microscope. By creating high-resolution maps of individual magnetic particles, scientists can reconstruct ancient fields with much higher precision while filtering out muddy signals from unstable grains.

Despite its potential, magnetic microscopy is an emerging field with its own set of uncertainties. To help constrain measurement data, Bellon et al. combined QDM observations with computer modeling to analyze how a magnetic particle’s stray field—the magnetic flux that leaks into the surrounding space—decays as it moves away from the source. They specifically investigated how a particle’s internal magnetic structure and external measurement noise affect the accuracy of these reconstructions.

The study found that in iron oxides, the smallest and most magnetically stable particles produce signals that are strong at the source but fade rapidly with distance. In contrast, larger particles produce signals that remain detectable farther away. This creates a challenge: The most stable grains for long-term geological data (the smallest ones) are the hardest to detect if the sensor is not perfectly positioned or if sensor interference is present.

By quantifying measurement error, the authors provide a road map for the field of micropaleomagnetism. Their findings could allow researchers to better account for uncertainty, leading to more robust reconstructions of Earth’s magnetic history and a deeper understanding of planetary evolution. (Journal of Geophysical Research: Solid Earth, https://doi.org/10.1029/2025JB033133, 2026)

—Aaron Sidder, Science Writer

Citation: Sidder, A. (2026), Navigating the past with ancient stone compass needles, Eos, 107, https://doi.org/10.1029/2026EO260122. Published on 16 April 2026.

Text © 2026. 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.

The volcano Ol Doinyo Lengai in Tanzania is unique on Earth: Its lava is rich in carbon compounds that melt at significantly lower temperatures than typical silicon-rich lavas from other terrestrial volcanoes.

It is possible, however, that carbon volcanoes could exist elsewhere, including on exoplanets, or—as suggested in a recently published article in Icarus—perhaps even on planet Mercury.

Despite being known from antiquity, Mercury is very hard to study because of its closeness to the Sun. As a result, the best data so far were gathered within the past 20 years by NASA’s MESSENGER (Mercury Surface, Space Environment, Geochemistry, and Ranging) probe. In particular, scientists identified mysterious pits they dubbed “hollows” scattered across Mercury’s surface. The hollows’ relatively bright appearance indicates they were formed in recent geological times, and could even be still forming today. The origins and geochemical makeup of these hollows are unknown.

“Mercury looks like the Moon a little bit, so we don’t expect large volcanoes,” said Maximilian Paul Reitze, a planetologist at Universität Münster’s Institut für Planetologie who is first author of the Icarus study. Without volcanic conditions like those on Earth or even on Jupiter’s moon Io, researchers expect Mercury to be largely geologically dormant. In other words, to explain hollows, “we need some volcanism under the conditions we expect on Mercury,” Reitze said.

Hence the interest in Ol Doinyo Lengai, known as the Mountain of God to the Maasai and Sonjo peoples. This volcano produces lava made up of carbonatites, igneous rocks composed of more than half carbon (and which are known to host critical minerals). These lavas flow at temperatures roughly 100°C lower than Mercury’s blazingly hot daytime temperature of 424°C. If the planet has a carbon-rich subsurface, as Reitze and his collaborators proposed, then the hollows could be Mercury’s version of Ol Doinyo Lengai.

This theory, however, has its skeptics.

“We know that there is carbon in [Mercury’s] crust, but the amount is very low,” said Paul Byrne, a planetary scientist at Washington University in St. Louis, who was not involved in the Icarus study. He also pointed out that the surface regions where carbon is most concentrated don’t correspond to higher concentrations of hollows. “For this to be some kind of carbon-based lava, it would imply a lot more carbon than we might think, given how widespread the hollows are.”

The Making of a Weird Planet

Mercury’s proximity to the Sun means that NASA’s Mariner 10 spacecraft provided humanity’s first-ever views when it flew by in 1974 and 1975. Three decades later, the MESSENGER mission was the first probe to orbit Mercury, mapping the planet’s full surface and turning up unexpected features like the hollows. The BepiColombo mission, a joint project of the European Space Agency and the Japan Aerospace Exploration Agency, is only the third mission ever to visit the planet, so when its two spacecraft settle into orbit in November 2026, it will almost inevitably reveal something unexpected, because it’s a weird planet.

Unlike Earth, Mars, or the Moon, Mercury has a freakishly large core and a thin mantle.

“Basically, Mercury is a molten ball bearing wrapped in a thin blanket of rock,” Byrne said. “One explanation is that early in the planet’s life, either one large or several smaller impacts stripped the outer portion away.”

The question then becomes what got vaporized, and what was left behind, particularly when trying to understand hollows. Many planetary researchers proposed that sulfides in the mantle could drive volcanism, but Reitze had doubts.

“The problem with sulfides I see is that they’re stable up to 1,000°C or so, which cannot explain the explosive volcanism that’s needed to form those hollows,” he said.

Instead, he and his coauthors contacted a colleague working on Ol Doinyo Lengai, who obtained a sample of the lava for laboratory study while it was still molten. Because carbonatite lava reacts chemically with Earth’s air very quickly, the researchers needed to isolate it to understand how the unaltered materials might behave under conditions on Mercury, particularly infrared spectra that could be confirmed by the BepiColombo mission.

In the hypothesis proposed by Reitze and colleagues, impacts from meteorites heat the carbon-rich magma below Mercury’s surface, melting it and driving eruptions. The hollows, which are found frequently on the slopes of Mercury’s craters or their central peaks, are the remains of those eruptions. Over time, further meteorite bombardments and intense solar radiation destroyed older hollows, which is why the ones in MESSENGER data were all formed within the past 270 million years—a short time ago, geologically speaking.

“The carbonatite angle is an interesting one, and I certainly wouldn’t rule it out,” Byrne said. “Anytime people have been confident about anything in planetary science, [planets have] shown you wrong. I’m certainly open to it, but is it the only explanation for all of the hollows? I am skeptical of that.”

Byrne and Reitze both dream of a future Mercury lander, a very challenging and expensive proposition nobody expects will happen soon. In the meantime, they agreed that BepiColombo data will help settle the question of whether the most Mercury-like place on Earth is a volcano in Tanzania.

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

Citation: Francis, M. R. (2026), A unique African volcano could solve a mystery on Mercury, Eos, 107, https://doi.org/10.1029/2026EO260176. Published on 2 June 2026.

Text © 2026. 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.

Magma on the move can cause the ground around a volcano to heave in measurable ways. But surface deformation doesn’t always point to an impending eruption—new results show that the terrain around a volcano can also shift during episodes of heavy rainfall. Researchers studying Japan’s Mount Fuji spotted instances of centimeter-level ground deformation tied to intense precipitation. Fortunately, such events can be readily differentiated from deformation caused by magmatic activity, the team reported in Geology.

Keeping an Eye on Volcanoes

Volcanoes around the world, from Kīlauea in the United States to Calbuco in Chile, are outfitted with arrays of sensors. Mount Fuji is no exception—the region around the edifice is equipped with dozens of instruments to detect ground movement, infrasound, and other signs of potential volcanic unrest. All that monitoring is warranted: Shin-Fuji (“Younger Fuji”)—the youngest of Mount Fuji’s three overlapping volcanoes—is currently active.

Shuo Zheng, a hydrological geodesist at Hong Kong Polytechnic University in China, and his colleagues recently mined some of those Mount Fuji data. The team focused on Global Navigation Satellite System (GNSS) observations—otherwise known as GPS data—collected daily from 2017 to 2023.

Rain and Rise

Zheng and his collaborators found several instances in which the two GNSS stations located within 10 kilometers of the summit of Mount Fuji recorded clear signs of uplift. Those signals, reflecting changes of roughly 1–2 centimeters, far exceeded the sensors’ millimeter-level precision. And when the team correlated the timing of that uplift with rain gauge records, they found that the ground often tended to rise almost immediately during periods of heavy precipitation (defined as several tens of millimeters of rain falling per day).

There’s likely a physical link behind that correlation, the researchers surmised. The explanation involves the so-called clinkers that cap each of Mount Fuji’s subterranean layers of lava. Clinkers are layers of small rocks that form when the surface of a lava flow rapidly cools, and these structures persist in the shallow subsurface of Mount Fuji. “They can store and transmit groundwater, acting like aquifers,” Zheng said.

When water fills up the pore space within a clinker, there’s no place for the overlying ground to go but up. It therefore makes sense that GNSS stations located atop old lava layers would exhibit uplift in response to intense rainfall, the team concluded.

When Zheng and his collaborators analyzed data from the nine GNSS stations located between 25 and 40 kilometers from the summit, however, they found that the ground actually tended to subside during periods of heavy precipitation. “There are two different responses,” said Kosuke Heki, a geophysicist and geodesist at Hokkaido University in Japan and a member of the research team. That subsidence is a known effect, and it’s been observed in a variety of locales. The subsidence doesn’t dominate closer to the summit of Mount Fuji because of the presence of the clinker layers there, the team reasoned.

Long-Lasting Magma

The uplift that the team recorded close to the summit of Mount Fuji tended to last just a day or two; it disappeared when the rainfall ceased. That timing is key for differentiating precipitation-induced uplift from magma-induced uplift. “Uplift by rain easily terminates when it stops raining,” said Heki. “But magma has a much longer timescale. It continues for weeks or months.”

That difference is critical, said Luca Caricchi, a volcanologist at the Université de Genève who was not involved in the research. There’s long been the mindset that ground deformation means that an eruption is imminent, but these new findings show that a heaving volcano doesn’t always mean that magma is on the move, said Caricchi. If the deformation is short-lived, the explanation might just be precipitation, he said. “You don’t need to worry.”

Zheng and his colleagues have looked for a similar effect for other volcanoes in Japan. They didn’t find any conclusive trends when they analyzed a chain of island volcanoes south of Tokyo, however. Perhaps that’s because the clinker layers beneath those edifices are so close to the sea that water efficiently drains out of them, the team hypothesized.

—Katherine Kornei (@KatherineKornei), Science Writer

Citation: Kornei, K. (2026), Heavy rainfall inflates Mount Fuji, Eos, 107, https://doi.org/10.1029/2026EO260169. Published on 26 May 2026.

Text © 2026. 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.

Off Track, On Purpose

Iceland, Chile, Kenya, Antarctica, Papua New Guinea, and the Great Salt Lake. That ambitious lineup covers (most of) the destinations where scientists featured in our annual fieldwork collection have ventured to test innovative instruments and answer pressing questions about natural processes on—and off—Earth.

Read along to learn about some fascinating field science and to hit all these hot spots and cool destinations for yourself.

In “Discovering Venus on Iceland,” scientists describe a multiweek effort traversing three rugged and rocky sites to collect samples and validate airborne radar measurements. Iceland’s basaltic lava fields are about the closest analogue to the surface of Venus that Earth has to offer, and the team’s data collection is helping to test the performance of instruments that will be a part of NASA’s VERITAS (Venus Emissivity, Radio Science, InSAR, Topography, and Spectroscopy) mission in several years’ time.

From Iceland, travel west and south to Chile, Guatemala, and Idaho to learn how researchers have been building and using their own inexpensive, lightweight sensors to detect infrasound emanating from volcanoes, earthquakes, and wildfires in “Sensing the Sounds from Earth’s Hazardous Environments.” At Villarica volcano in the Chilean Andes, for example, they have deployed sensor clusters on, around, and even hanging from a cable above the volcano’s summit crater to better understand how infrasound may be useful for eruption monitoring.

Meanwhile, at Lake Turkana in Kenya, scientists have been partnering with local industries to map the subsurface and better understand how the continent is unzipping along the East African Rift System, as Kimberly Cartier describes in “Eastern Africa Is Splitting Apart, but Not Where We Expected.”

Stick with Cartier for another leg of our fieldwork trip as she relates how researchers have instrumented an underwater volcanic vent off Papua New Guinea to track effects of ocean acidification on corals in “Coral Diversity Drops as Ocean Acidifies.”

From there, head to the decidedly less tropical climes of the South Pole, where a team recently installed a pair of seismometers deep in the Antarctic ice, completing a challenging and years-long feat of engineering, reports Grace Van Deelen in “These South Pole Seismometers Will Detect Vibrations 1.5 Miles Under the Ice.”

Finally, journey to the North American interior to learn what scientists found when they installed electrodes on the now-desiccated surface of Utah’s Great Salt Lake in Carolyn Wilke’s—spoiler alert—“What’s Below the Great Salt Lake? More Water.”

We’ll understand if you need a break after all that globe-trotting. But you’re always welcome to join us for more adventures in the field.

—Timothy Oleson, Eos Senior Science Editor

Citation: Oleson, T. (2026), An off-road itinerary, Eos, 107, https://doi.org/10.1029/2026EO260181. Published on 1 June 2026.

Text © 2026. 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: Journal of Geophysical Research: Solid Earth

Enhanced Geothermal Systems (EGS) can expand low-carbon energy production, but fluid injection may trigger earthquakes whose locations and mechanisms are difficult to predict. Feng et al. [2026] investigate induced seismicity at China’s first EGS site in the Gonghe Basin using a comprehensive observational dataset. Machine learning processing of data from 20 surface seismic stations produced a high-resolution earthquake catalog with well-constrained locations and focal mechanisms. Stress inversion and modeling, constrained by borehole stress measurements, reveal mechanically weak faults with low friction coefficients, indicating that low-to-moderate fluid overpressure can trigger seismic slip. Site-scale analysis shows that seismicity reflects shear reactivation of pre-existing natural faults, rather than the creation of new tensile fractures. Further integration with borehole image logs reveals a fine-scale relationship between the main seismogenic zones and stress heterogeneity, expressed as rotations of the principal stress axes that likely reflect localized lithological contrasts and fault-damage zones.

Together, these integrated analyses show that geothermal-induced seismicity is controlled by inherited fault architecture at the site scale and localized stress heterogeneity at the borehole scale. By linking seismic observations to borehole stress and image-log evidence, the study provides a more physically constrained framework for seismic-hazard assessment and stimulation design in enhanced geothermal reservoirs.

Citation: Feng, P., Wang, R., Zhang, H., Zhang, C., Schultz, R., & Yang, L. (2026). Pre-existing structures and stress variations jointly control the induced seismicity in enhanced geothermal system of Gonghe Basin, China. Journal of Geophysical Research: Solid Earth, 131, e2025JB033158. https://doi.org/10.1029/2025JB033158  

—Xiaowei Chen, Associate Editor, JGR: Solid Earth

Text © 2026. 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.

Ensuring the sustainability of water resources and ecosystems in a changing world requires a thorough understanding of how water moves through Earth’s Critical Zone, a dynamic interface where air, water, soil, plants, and rocks interact. Researchers can track and model this movement of water using naturally occurring markers or “tracers.”

A recent article in Reviews of Geophysics explores the latest advancements in tracer-aided mixing models and how they can help us to better understand the Critical Zone. Here, we asked the authors to give an overview of the Critical Zone, how tracer-aided mixing modeling works, and future directions for research.

What is the Critical Zone (CZ)?

The Critical Zone is Earth’s “living skin”—the dynamic layer where the atmosphere, hydrosphere, biosphere, and lithosphere interact. It stretches from the top of the vegetation canopy and, in cold regions, from the surface of snowpacks and glaciers, down through soils and into the deeper aquifers. It encompasses lakes, streams, and wetlands at the surface, and extends beyond the soil layer to underlying groundwater aquifers. It is where rainfall, snowmelt and glacier melt become soil moisture, where plants take up water and return it to the atmosphere, where aquifers get recharged, and where streamflow is generated. In short, the Critical Zone is where most processes that sustain terrestrial life and freshwater resources unfold.

Why is it important to understand how water moves through the Critical Zone?

Virtually every freshwater resource we rely on (e.g., drinking water, irrigation) passes through the Critical Zone at some point. Global warming, land-use changes, and intensifying water demand emerging from rapid urbanization and changes in agriculture are reshaping how water is stored and released within the Critical Zone, often in ways we cannot yet predict. Understanding how much water is stored within the Critical Zone, how this water is both recharged from rainfall and snowmelt and eventually discharged into streams, and the timescale of these dynamic processes is essential for protecting ecosystems, safeguarding water supplies, and adapting to a changing climate.

How would you explain a tracer-aided mixing model to a non-specialist?

Imagine mixing a glass of orange juice with a glass of apple juice, and trying afterwards to work out how much of each went into the glass. If the juices had distinctive “fingerprints” (imagine its color, sugar content, or a specific chemical) and these fingerprints primarily changed because of the mixing of these two juices, you can then measure the fingerprint in the final mixture and back-calculate the proportion of its distinct sources.

Tracer-aided mixing models work in a similar way but can track the entire water cycle. Different water sources (e.g., rainfall, snowmelt, glacier melt, soil water, groundwater) can have distinct “fingerprints” in a naturally occurring tracer, such as stable isotopes of water or specific dissolved elements. By measuring these fingerprints in the streamwater or groundwater and in its potential sources for example, hydrologists can estimate how much each source contributed to the streamwater or groundwater.

What are some of the most significant and exciting recent advances in tracer-aided mixing models?

Classical mixing models relied on demanding assumptions: that all water sources can be identified and sampled, and that their signatures were distinct and constant in time. Much of the recent progress has been about relaxing these assumptions.

Bayesian approaches now estimate full probability distributions and provide a more realistic picture of uncertainty. Methods like Convex Hull End-Member Mixing Analysis (CHEMMA) use machine learning to infer the distinct sources directly from data, while ensemble hydrograph separation exploits tracer fluctuations over time, thereby making un-mixing feasible even when multiple sources have overlapping signatures. Perhaps the most conceptually novel advance is end-member splitting, which flips the question from “where does streamflow come from?” to “where does precipitation go?”

Alongside these modeling advances, there have been immense advances in how tracers are measured. Portable laser and mass spectrometers now enable high-frequency, in-situ tracer measurements which allows us to capture critical hydrological events such as storms and snowmelt in near-real time.

What are stable water isotope tracers and what are their advantages?

Stable water isotopes are naturally occurring non-radioactive atoms of hydrogen and oxygen that make up a water molecule but have slightly different molecular masses. The two stable isotopes widely used in hydrology are 2H (deuterium) and 18O (oxygen-18). Because these isotopes are part of the water molecule itself, they directly travel with the water molecule. Their key advantages are: (1) they are conservative, meaning they do not react chemically as water moves through soils and aquifers, and (2) they carry distinct signatures resulting from climatic variables such as air temperature.

Consequently, in the European Alps, winter precipitation has a different isotopic signature than summer precipitation because winters are cooler than summers. Other hydrological processes such as evaporation and sublimation leave a recognizable fingerprint on the remaining water, thereby allowing us to estimate how much evaporation or sublimation occurred. Stable water isotopes can be measured in essentially every water compartment, from atmospheric vapor and precipitation to snowpack, plant xylem, soil water, streams, and groundwater. Together, these properties make stable water isotopes the most versatile and widely used tracer in Critical Zone hydrology.

What are the current limitations of tracer-aided mixing models?

Despite their power, mixing models still face many constraints. End-member signatures vary in space and time, are sometimes too similar to distinguish, and some sources may be overlooked entirely. Non-conservative tracers such as nitrate or sulfate can react with their environment along their journey, thereby biasing results if these reactions are not explicitly accounted for.

Sampling is another major bottleneck. Capturing the spatial heterogeneity of soils, snowpacks, and groundwater requires a lot of measurements that are often logistically or financially prohibitive, especially in remote regions. Many of the newer, more powerful tracers such as noble gases or stable isotopes of trace elements, can only be analyzed by a handful of specialized laboratories. As a result, global coverage remains highly uneven, with key regions such as the Arctic and the global South still under-sampled.

What are some of the major unsolved questions and where is more research needed?

There are several fronts where more research is needed. Source signatures are not static, and methods that explicitly capture their variability in time are still underdeveloped. Embedding tracers within global Earth System Models would, in theory, enable more accurate assessment of hydrological partitioning e.g., how rainfall, snowmelt, and glacier melt are split between sublimation, evapotranspiration, groundwater, and streamflow. These will directly inform more robust climate projections, but this remains technically demanding.

Expanding data coverage in under-sampled regions is critical, and citizen science and low-cost sensors may help. Machine learning is a promising approach for uncovering non-linear relationships and gap-filling sparse datasets, but requires training data that often do not yet exist. Greater interdisciplinary integration, e.g., combining tracers with remote sensing, ecological indicators, and biogeochemical data, could yield a more holistic view of the Critical Zone. Finally, the field would benefit from shared protocols and open data practices to enhance progress.

—Andrea L. Popp (andrea.popp@smhi.se; 0000-0003-3911-8105), Swedish Meteorological and Hydrological Institute, Sweden; Harsh Beria (hberia@ethz.ch; 0000-0003-2597-9449), ETH Zurich, Switzerland

Editor’s Note: It is the policy of AGU Publications to invite the authors of articles published in Reviews of Geophysics to write a summary for Eos Editors’ Vox.

Citation: Popp, A. L., and H. Beria (2026), Tracing water’s hidden journey through the Earth’s living skin, Eos, 107, https://doi.org/10.1029/2026EO265019. Published on 13 May 2026.

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 © 2026. 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.