EOS

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

A recent article in Reviews of Geophysics explores land subsidence drivers, rates, and impacts across the globe. It also discusses the need for improved process representations and the inclusion of the interplay among land subsidence and climatic extremes, including their effects in models and risk assessments. Here, we asked the lead author to explain the concept of land subsidence, its impacts, and future directions needed for improved mitigation.

What is land subsidence? 

Land subsidence (LS) refers to the relative sinking or lowering of the Earth’s land surface. LS is a pressing global issue that warrants action since subsidence can adversely impact infrastructure, humans, and the environment across various landscapes and climates (Figure 1). It may be driven by one or more natural processes and/or human activities that compound to cause localized or expansive ground deformation. Differential LS causes structures and roadways to crack and buckle. LS can also reduce the water storage capacity of aquifers. Notably, LS can be recoverable (e.g., natural variations in groundwater levels) or permanent (e.g., overdraft causing irreversible compaction).  

Figure 1. Reported LS rates and drivers around the world based on literature. (a) Map of primary LS drivers (colors) indicating mean (circles) and maximum (triangles) rates (shape sizes). A shared color scheme (shown in (b)) demarcates the main causes of LS in (a) and (b). (b) 50 largest mean LS rates for global locations (numbered along x-axis and listed above). LS rates are often nonlinear, temporally dependent, and occur at various time scales. Rates shown were not all observed or estimated over the same time period. Credit: Huning et al. [2024], Figure 1.

Why is it important to understand and monitor land subsidence? 

Various LS drivers and physical processes exist and interact with one another (Figure 1). LS is often closely related to natural resources demand, which increases with growing urbanization and megacities. The proximity of LS to critical infrastructure like water conveyance, transportation, and utility systems is a significant concern since LS could cause catastrophic lifeline failures, outages, and/or loss of life. Also, feedbacks between climatic extremes (e.g., droughts, floods, wildfires, heatwaves) and LS impacts exist, but are not fully understood.

Although a chronic hazard, LS may initially go unnoticed as sinking typically occurs slowly. This influences perceived risk and contributes to reactive policies, regulations, and mitigation steps targeting LS and its implications rather than proactive measures. Furthermore, the compounding effects of extreme events and their impacts can exacerbate LS. More pronounced interactions are likely with projected rises in climate extremes.

How do scientists monitor and measure land subsidence across the globe? 

Scientists use various techniques and technologies to measure LS, including ground-based surveys, subsurface instrumentation, and satellite-based observations. Satellite-based Synthetic Aperture Radar (SAR) has revolutionized LS monitoring and mapping. It is an active remote sensing system that emits microwave pulses and receives echoes. Such systems can operate under various conditions (e.g., day and night, in cloudy skies) and produce high-resolution imagery. With SAR-based information, scientists can infer surface deformation by computing phase differences between SAR snapshots over a region using techniques like interferometric SAR (InSAR). SAR-based observations commonly inform impact assessments for agriculture, structural health, and resource management.

What are the major natural and anthropogenic drivers of land subsidence? 

Naturally-occurring processes and human activities can independently drive LS or enhance existing LS rates (Figure 2). Some examples of natural drivers of LS include: natural consolidation, volcanic or tectonic activity, seasonal groundwater level variations, and soil organic material decomposition. Extraction of natural resources (e.g., fossil fuels, groundwater), removal of wetlands and peatlands, and loading from rapid urbanization serve as examples of human-related activities contributing to LS. Natural resource extraction is a leading anthropogenic driver of LS (Figure 1), which often rises with increasing population. Also, extreme events such as wildfires or heatwaves can trigger LS in permafrost areas by thawing the permafrost layer, altering the soil structure, and releasing greenhouse gases that accelerate warming.

Figure 2. Schematic illustrating feedbacks and effects of land subsidence, extreme events, and human activities. Credit: Huning et al. [2024], Figure 3.

How is land subsidence projected to change in the future? 

Estimating future LS rates is challenging. Projecting human activities driving LS and the effectiveness of restoration and mitigation efforts is complicated, uncertain, and variable. LS projections also depend on other factors (e.g., infrastructure investments, land use-land cover changes). They are further complicated by uncertain projected hydrologic variables like precipitation. Yet, more people are expected to be exposed to LS with greater economic losses anticipated in the future.

Sea level rise (SLR), rising temperatures, and extreme events often compound LS. Subsiding coastal areas and deltas face higher inundation risk from the compounding effect of SLR. Extreme events and LS impacts are expected to increasingly affect one another (Figures 2-3) as extremes (e.g., drought) intensify with warming. Amidst drought, groundwater levels drop through decreased recharge and increased pumping, often leading to soil compaction and LS. As soils dry and crack, heightened microbial processes decompose soil organic matter and release carbon. Such processes can enhance warming while triggering LS and feedbacks. As temperatures rise, permafrost thaw-driven LS is also expected to expand, increasing the infrastructure at risk for damage and failure.

Figure 3. Example feedback loops involving land subsidence, climatic trends, extreme events, infrastructure, and cascading hazards. (a) Peatland‐carbon, (b) permafrost‐carbon, and (c) salinization‐subsidence feedbacks and (d) infrastructure‐subsidence, (e) flood‐subsidence, and (f) drought‐subsidence cascading hazards. Black (orange) arrows denote a positive feedback (strengthening of impacts). Credit: Huning et al. [2024], Figure 4.

What additional research, data, or modeling is needed to help track and mitigate land subsidence and its impacts? 

Integrated models incorporating multiple LS drivers and processes are necessary for better estimating LS rates, extent, and ramifications at the spatiotemporal resolutions essential for mitigation, adaptation, and policy. Additional data and research are needed to understand the interplay of extreme events, infrastructure, climatic trends, and human activities with LS dynamics and effects (Figure 3), and inform LS mitigation efforts.

Improved climate modeling, management practices, and risk assessments require better representations of LS feedbacks, carbon emissions, and LS processes. Such advancements necessitate accurate, longer, and spatial observations and analyses with improved process understandings. Global adoption of consistent monitoring and reporting frameworks will also support such efforts by leading to new insight into LS observations and regions at-risk for LS, LS-enhanced flooding, etc. Interdisciplinary efforts will help transform science into action focused on LS hazard and risk mitigation.

—Laurie S. Huning (laurie.huning@csulb.edu, 0000-0002-0296-4255), California State University, Long Beach, 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: Huning, L. S. (2025), Rising concerns of climate extremes and land subsidence impacts, Eos, 106, https://doi.org/10.1029/2025EO255019. Published on 9 June 2025. This article does not represent the opinion of AGU, Eos, or any of its affiliates. It is solely the opinion of the author(s). Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

A new paper in the journal Landslides has presented a review of a large landslide that killed 29 people in Sichuan Province.

On 8 February 2025, a large rock avalanche occurred in Junlian County in Sichuan Province, China. I wrote about this event, now known as the Junlian rock avalanche, at the time. With remarkable and commendable pace, Bo Zhao and colleagues have published an initial review of the event (Zhao et al. 2025) in the journal Landslides. Whilst the paper is behind a paywall, this link should allow readers to access the full text.

The landslide is located at [27.99885, 104.60801]. The Google Earth image below shows the site in 2020 – the marker is on the source area of the Junlian rock avalanche:-

Google Earth image of the site of the 8 February 2025 Junlian rock avalanche, China.

The image below, published by Xinhua, shows the aftermath of the landslide:-

The aftermath of the 8 February 2025 Junlian rock avalanche in Sichuan, China. Image by Xinhua.

Zhao et al. (2025) have determined the key statistics for this landslide. The initial failure was 370,000 m3, increasing to 600,000 m3 through entrainment. The landslide had a runout distance of 1,180 metres and a vertical elevation change of 440 m, giving a landslide mobility index of 0.37. This is a typical value for a rock avalanche of this volume.

Zhao et al. (2025) show that the initial failure was structurally controlled, which is no surprise. It occurred in a Triassic interbedded sandstone and mudstone formation. They estimate that the average velocity was 19.3 m/second.

The authors consider in some detail the triggering event. The site experienced 10 days of low intensity rainfall prior to the failure. Zhao et al. (2025) suggest that this led to the build up of pore water pressure, initiating the failure. Total rainfall in the month proceeding the collapse was in the order of 85 mm. This rainfall seems somewhat unexceptional, suggesting to me that a progressive failure mechanism was in play.

The Junlian rock avalanche killed 29 people and left two people injured. It is a fascinating example of a major failure with high consequences in a remote mountainous area. Anticipating such events remains a major challenge in landsldie research. Many thanks to the authors for providing such a rapid description of this event.

Reference

Zhao, B., Zhang, Q., Wang, L. et al. 2025. Preliminary analysis of failure characteristics of the 2025 Junlian rock avalanche, China. Landslides. https://doi.org/10.1007/s10346-025-02556-1.

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.

On 10–11 May 2024, the strongest solar storm since 2003 hit Earth. The storm caused spectacular aurorae around the world, including as far south as Kansas in the midwestern United States. Unfortunately, it also had negative effects, such as days-long disruptions in GPS signals needed by farm tractors that, in turn, caused delays in planting operations at a critical time in the spring.

Solar storms, which throw torrents of protons, neutrons, and other particles at our planet, have had severe effects in decades past. A massive storm in May 1967, for example, significantly disrupted military communications (and ultimately led the United States to strengthen its space weather capacity) [Knipp et al., 2016]. Another, in March 1989, disabled power grids, hitting Quebec, Canada, especially hard [Boteler, 2019].

The biggest recorded modern event took place in February 1956. Were it to be repeated today, such an event could disrupt aircraft electronics and expose passengers to substantially elevated radiation doses.

The largest known solar event in history, 50–100 times larger than the one that happened in 1956, occurred in 774 CE [Miyake et al., 2012]. An event on par with the 774 storm is considered a worst-case scenario for modern aviation [Mishev et al., 2023].

With the 11-year solar cycle approaching its maximum in 2025, we are in a time of heightened potential for such events to disrupt daily life.

Fortunately, technology for observing solar storms and the particle showers they rain down on Earth has developed significantly over the past several decades. Both ground-based and satellite observations are critical for measuring solar storms and their effects [National Academies of Sciences, Engineering, and Medicine, 2024] and for generating space weather forecasts (e.g., by NOAA’s Space Weather Prediction Center (SWPC)). The global aviation sector, for example, uses these forecasts to predict solar radiation storm warning levels and radiation dosage levels to help keep flights safe.

The small number of high-energy neutron monitoring stations used to observe the effects of solar events at Earth’s surface limits data availability and thus the accuracy and spatial resolution of forecasts.

Good predictions rely on the availability of high-quality and comprehensive data. However, the small number of high-energy neutron monitoring stations currently used to observe the effects of solar events at Earth’s surface limits data availability and thus the accuracy and spatial resolution of forecasts. But solutions are within reach.

In addition to space weather scientists, hydrologists use data from these monitoring stations, albeit for a different purpose: They rely on the high-energy neutron detections to calibrate the low-energy neutron detectors they use as one way to collect snow cover and soil moisture measurements that are important for hydrological modeling and agricultural applications. Recent studies showed that the larger networks of low-energy neutron detectors used by hydrologists can supplement and effectively increase the coverage of the smaller network of high-energy neutron monitors [Baird, 2024]. Now, scientists are devising a strategy to combine forces for their mutual benefit.

Wanted: Better Observational Capabilities

Massive lead-lined neutron monitors (NMs) are typically used to monitor the arrival of cosmic ray particles at Earth’s surface. These particles include high-energy secondary neutrons (carrying energies of ~50–100 megaelectron volts) that are generated by collisions of primary solar and galactic cosmic rays with other particles in the atmosphere, a process that can be reconstructed using NM data and numerical models [Mishev et al., 2014].

This 18-tube neutron monitor is housed in a Quonset hut on the campus of the University of New Hampshire in Durham. Credit: James Ryan, University of New Hampshire

Satellites, including those in the GOES (Geostationary Operational Environmental Satellite) system, also provide operational data about primary cosmic rays in real time, but they cannot resolve particle energies in the detail required for estimating radiation doses affecting aviation or for modeling solar particle energy spectra [National Academies of Sciences, Engineering, and Medicine, 2024].

A global network of NMs, each run by different universities or other entities, has been in operation for the past 7 decades [Väisänen et al., 2021]. Unfortunately, today, only 20 NM sites around the globe provide real-time data; another roughly 30 NMs have been shut down because of a lack of long-term funding to maintain them. Geopolitical factors and closed data policies in some parts of the world additionally limit data quality and access internationally.

The U.S. Senate’s 2020 Space Weather Research and Forecasting Act emphasized the need for better observational capabilities to address this crisis of critical infrastructure. The 2020 PROSWIFT Act and the most recent National Academies’ solar and space physics decadal survey [National Academies of Sciences, Engineering, and Medicine, 2024] further underscored the challenges and need for supporting long-term operational NM networks.

Hydrologists Have Their Own Networks

Hydrologists have, in the past 15 years, deployed networks of detectors similar to neutron monitors (NMs) to measure snow and soil moisture.

Applying methods developed beginning several decades ago [e.g., Kodama et al., 1979], hydrologists have, in the past 15 years, deployed networks of detectors similar to NMs to measure snow and soil moisture [Zreda et al., 2012]. These cosmic ray neutron sensors (CRNSs) are, however, much smaller than NMs, and they are sensitive to much lower neutron energies (~0.025 to 100 kiloelectron volts).

At these lower energies, the number of detected neutrons depends not only on incoming secondary cosmic rays but also on the abundance of hydrogen in the surrounding environment (e.g., in the form of snow or soil moisture). In soil, for example, cosmic ray neutrons collide with hydrogen atoms, lose energy in the process, and become thermalized (i.e., they slow down). CRNSs are designed to count these water-sensitive neutrons.

The sensors can measure these low-energy neutrons within a roughly 20-hectare circular area and up to about 30 centimeters above the ground surface, an extraordinarily large volume relative to their size. Figure 1 shows how example CRNS measurements of neutron counts and soil water content from central Nebraska clearly respond to rainfall, as measured by the local Mesonet station, and match potential evapotranspiration data well.

Fig. 1. Neutron counts (corrected for variations in air pressure, water vapor, and high-energy neutron intensity, top left) and estimates of soil water content (bottom left) from 4 April to 2 May 2021 collected from a cosmic ray neutron sensor (CRNS) in central Nebraska are shown. Rainfall data from a local Nebraska Mesonet station (top right) and potential evapotranspiration (ET) data (bottom right) over the same period match the soil water content changes measured by the CRNS.

Area-averaged estimates of snow and soil moisture like this match scales relevant for hydrological modeling and agricultural management (e.g., irrigation and fertilizer application, crop yield prediction), providing a big advantage compared with estimates from point-scale measurements, given the high spatial variability that naturally exists from one meter to another. CRNS detectors offer other benefits as well. Their measurements, collected roughly hourly, are nondestructive; they have extremely low maintenance costs; and they can be deployed outdoors for long-term environmental monitoring.

Today, more than 300 CRNS instruments are operating across all seven continents, with networks in Australia, China, Europe, India, South Africa, the United Kingdom, and the United States. These networks have led to exciting advances in hydrology.

A CRNS measures soil moisture and snow water equivalent at a study site in eastern Nebraska. Graduate students Sophia Becker and Tanessa Morris are collecting soil samples for calibration. Credit: Trenton Franz, University of Nebraska–Lincoln

For example, CRNSs have been shown to be excellent sources of ground validation data for remote sensing soil moisture data products like SMAP (Soil Moisture Active Passive) and SMOS (Soil Moisture and Ocean Salinity) that support weather and agricultural forecasting efforts, among other applications [Montzka et al., 2017]. CRNS data have also been shown to significantly improve predictions of streamflow by catchment models by improving estimates of near-surface water storage [Dimitrova-Petrova et al., 2020]. Mobile CRNSs have also been deployed on commuter trains in Europe, providing soil moisture and snow observations across unprecedented scales [Schrön et al., 2021].

Despite their clear utility, CRNS networks, like the global NM network, often lack long-term funding. Moreover, in the United States, no single federal agency is mandated to monitor soil moisture, a void that hinders the development of a national coordinated soil moisture monitoring network.

An Exciting Opportunity

The CRNS research community has been highly dependent on the NM network because real-time reference data are required to correct CRNS measurements for variations in incoming cosmic radiation. In a recent advance bridging the two neutron monitoring communities, Baird [2024] showed that potential benefits also extend in the other direction.

He used 50 CRNS stations in the United Kingdom to investigate whether they can inform space weather monitoring, concluding that they “can identify persistent space weather periodicities, transient space weather periodicities, and transient aperiodic space weather signals” and that these capabilities are “largely unaffected by the influence of soil moisture in the data.” Although these identifications are not as reliable as those from neutron monitors, the much larger number of CRNSs compared with NMs offers promise for expanding data collection.

Baird also found that the CRNS data recorded some medium to large solar events, such as Forbush decreases (FDs), which are decreases in galactic cosmic rays reaching Earth following solar coronal mass ejections. The CRNSs detected 4 out of 28 FDs that had been identified by NMs between 2014 and 2022.

An exciting opportunity exists to use cosmic ray neutron sensor (CRNS) networks globally to augment the roughly 20-station NM network.

CRNS data have also been used to simulate ground level enhancements (GLEs) of radiation levels at Earth’s surface caused by bombardments of intense solar cosmic rays. These emitted particles, primarily protons, are accelerated to high energies during solar flares or coronal mass ejections. GLEs are rarer than FDs, occurring once per year on average, but are more detrimental to humans and aviation. GLEs are also nearly impossible to predict and prepare for because they arrive at Earth only minutes after a solar flare or coronal mass ejection occurs, whereas FDs take several days to arrive.

Given the newfound connection between low-energy neutron observations and space weather phenomena, an exciting opportunity exists to use CRNS networks globally to augment the roughly 20-station NM network. This ability would offer an unprecedented number of ground monitors to help researchers understand and analyze larger FD and GLE events and their impacts all around Earth.

Two Communities Join Forces

The hydrology and space weather communities have worked together informally since the 2010 launch of the Cosmic-Ray Soil Moisture Observing System in the United States [Zreda et al., 2012]. But the need for additional collaboration has been identified in the literature and during joint sessions at AGU and European Geoscience Union meetings.

In response to this increased interest, the first Coordinated Cosmic-Ray Observation System Conference was held in October 2024 at the University of Nebraska–Lincoln. The hybrid event gathered 50 experts from academia, government, and industry to explore both the scientific potential of ground-based neutron monitoring across energy spectra and opportunities for productive cross-disciplinary partnerships.

Conference participants produced a concept paper identifying key issues on which the participating communities can work together. These issues involve critical needs for improved infrastructure and enhanced data accessibility.

Documenting soil moisture conditions more comprehensively and meeting data needs for environmental modeling and operational products, for example, require the deployment of additional CRNS stations globally—ideally, 30 stations per 1 million square kilometers. In the United States, this level of coverage equates to about 250 stations spread across the country’s roughly 8 million square kilometers.

With respect to space weather, NOAA’s SWPC has stated a need for real-time NM data (1-minute resolution with 5-minute latency) and additional NM monitoring sites to improve the spatial resolution of aviation forecasts. More NM sites are also needed to better understand the anisotropy (uneven distribution) of incoming cosmic ray particles globally, particularly during GLEs and other perturbed geomagnetic conditions, and how it may influence space weather impacts experienced around the planet.

By collaboratively addressing these and other gaps in the neutron-detecting networks used for space weather and soil moisture monitoring, we can advance scientific understanding of critical environmental and planetary processes and better serve the needs of operational systems designed to foster safety and prosperity.

References

Baird, F. (2024), The potential use of hydrological neutron sensor networks for space weather monitoring, Ph.D. thesis, University of Surrey, Guildford, U.K., https://doi.org/10.15126/thesis.901065.

Boteler, D. H. (2019), A 21st century view of the March 1989 magnetic storm, Space Weather, 17(10), 1,427–1,441, https://doi.org/10.1029/2019SW002278.

Dimitrova-Petrova, K., et al. (2020), Opportunities and challenges in using catchment-scale storage estimates from cosmic ray neutron sensors for rainfall-runoff modelling, J. Hydrol., 586, 124878, https://doi.org/10.1016/j.jhydrol.2020.124878.

Knipp, D. J., et al. (2016), The May 1967 great storm and radio disruption event: Extreme space weather and extraordinary responses, Space Weather, 14(9), 614–633, https://doi.org/10.1002/2016SW001423.

Kodama, M., et al. (1979), An application of cosmic-ray neutron measurements to the determination of the snow-water equivalent, J. Hydrol., 41(1–2), 85–92, https://doi.org/10.1016/0022-1694(79)90107-0.

Mishev, A. L., L. G. Kocharov, and I. G. Usoskin (2014), Analysis of the ground level enhancement on 17 May 2012 using data from the global neutron monitor network, J. Geophys. Res. Space Phys., 119(2), 670–679, https://doi.org/10.1002/2013JA019253.

Mishev, A., S. Panovska, and I. Usoskin (2023), Assessment of the radiation risk at flight altitudes for an extreme solar particle storm of 774 AD, J. Space Weather Space Clim., 13, 22, https://doi.org/10.1051/swsc/2023020.

Miyake, F., et al. (2012), A signature of cosmic-ray increase in AD 774–775 from tree rings in Japan, Nature, 486, 240–242, https://doi.org/10.1038/nature11123.

Montzka, C., et al. (2017), Validation of spaceborne and modelled surface soil moisture products with cosmic-ray neutron probes, Remote Sens., 9(2), 103, https://doi.org/10.3390/rs9020103.

National Academies of Sciences, Engineering, and Medicine (2024), The Next Decade of Discovery in Solar and Space Physics: Exploring and Safeguarding Humanity’s Home in Space, Natl. Acad. Press, Washington, D.C., https://doi.org/10.17226/27938.

Schrön, M., et al. (2021), Neutrons on rails: Transregional monitoring of soil moisture and snow water equivalent, Geophys. Res. Lett., 48(24), e2021GL093924, https://doi.org/10.1029/2021GL093924.

Väisänen, P., I. Usoskin, and K. Mursula (2021), Seven decades of neutron monitors (1951–2019): Overview and evaluation of data sources, J. Geophys. Res. Space Phys., 126(5), e2020JA028941, https://doi.org/10.1029/2020JA028941.

Zreda, M., et al. (2012), COSMOS: The Cosmic-ray Soil Moisture Observing System, Hydrol. Earth Syst. Sci., 16, 4,079–4,099, https://doi.org/10.5194/hess-16-4079-2012.

Author Information

Trenton Franz (tfranz2@unl.edu), School of Natural Resources, University of Nebraska–Lincoln; Darin Desilets, Hydroinnova LLC, Albuquerque, N.M.; Martin Schrön, Helmholtz Centre for Environmental Research UFZ, Leipzig, Germany; Fraser Baird, University of Surrey, Guildford, U.K.; and David McJannet, Commonwealth Scientific and Industrial Research Organisation, Canberra, Australia

Citation: Franz, T., D. Desilets, M. Schrön, F. Baird, and D. McJannet (2025), Two neutron-monitoring networks are better than one, Eos, 106, https://doi.org/10.1029/2025EO250212. Published on 6 June 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: GeoHealth

Wildfires are creeping into urban environments with alarming frequency, and they are connected to health problems ranging from respiratory illnesses to hypertension to anxiety. Studying the links between wildfires in these areas and health is challenging because wildfire smoke and ash contain a mix of chemicals from buildings, cars, and electronics, leaving researchers and communities with many unanswered questions.

Barkoski et al. recently published the GeoHealth Framework for Wildland Urban Interface Fires to help researchers quickly visualize the relationships between urban wildfires and health outcomes, as well as identify data gaps and future research priorities. It also aims to improve the coordination among different groups working to support wildfire preparedness, response, and recovery. The researchers built the framework using the example of the 2020 Walbridge Fire, which burned more than 55,000 acres (about 22,258 hectares) in Sonoma County, California. This example helped them understand the types of geoscience and health data that are available and that are needed after a wildland-urban interface fire.

To apply the framework, users define a question and then map various wildfire and health factors and the ways they are connected. For example, they may select environmental factors preceding a specific fire, such as land use and recent weather patterns; characteristics of the fire, including its size and the kinds of materials it burned; and factors that influenced its spread, such as firefighter response, wind, and topography. The team suggests pulling data from sources such as the U.S. Geological Survey, NASA, NOAA, EPA, electronic health records, and public surveys.

These inputs and the known and hypothesized connections among them help users to identify which pollutants a fire may generate, how humans may encounter these pollutants (such as through the air or drinking water), and how these encounters may affect the likelihood of physical or mental health consequences.

The researchers also note that the framework can be expanded and adapted to apply to new research questions. For instance, if researchers want to better understand how wildfire exposure affects the biological mechanisms of disease, they could incorporate epidemiological, toxicological, and clinical research studies into the framework. These studies might include more detailed information about how wildfire smoke harms health, such as gene variants that predispose people to asthma. (GeoHealth, https://doi.org/10.1029/2025GH001380, 2025)

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

Citation: Sidik, S. M. (2025), Charting a path from fire features to health outcomes, Eos, 106, https://doi.org/10.1029/2025EO250214. Published on 5 June 2025. Text © 2025. AGU. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Source: Geophysical Research Letters

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

土壤湿度是温度和湿度的关键调节器，易受气候变化的显著影响。尽管土壤湿度至关重要，但其建模工作涉及数十个约束不充分的参数，而且不同的模型对土壤湿度水平在全球变暖背景下的变化往往存在分歧。

Gallagher 和 McColl 采取了一种“极其简化”的方法，仅根据降水量和地表净辐射来模拟土壤湿度。该模型在使用欧洲中期天气预报中心第五代大气再分析数据(ERA5) 和第六次耦合模式比较计划(CMIP6) 气候数据集进行测试时，效果良好。

研究人员表示，这令人惊讶，因为这个简单的模型排除了近期许多文献关注的测量数据：水汽压差（空气能够容纳的水分量与实际容纳的水分量之间的差值）和大气二氧化碳 (CO2) 水平。预计这两者都将随着温室气体排放的增加而上升。

研究人员认为，他们的模型之所以仍然有效，是因为水汽压差无法准确衡量大气对水的需求；而模型中包含的地表净辐射才是更佳的衡量指标。关于二氧化碳，研究人员表示，之前的一些研究高估了这种气体的作用。

这个简单的模型为两个关于土壤湿度的基本问题提供了可能的答案：(1)为什么土壤湿度呈W型纵向剖面，赤道和两极的湿度高，两极之间的湿度低；(2)为什么土壤湿度在某些地区随温度升高而增加，而在另一些地区则降低？

W型分布可能是降水率和辐射强度共同作用的结果。赤道附近的高降水量在模型中占主导地位，并导致高土壤湿度。中纬度地区和两极地区的降水量都处于中等水平。但中纬度地区比两极地区接收到更强烈的辐射，导致中纬度地区的土壤相对干燥。

至于第二个问题，研究人员认为，气候变暖可能对土壤湿度有不同的影响，因为气候变暖既可能伴随降水增加（导致土壤湿度升高），也可能伴随地表净辐射增加（导致土壤湿度降低）。这两个变量在不同地区会以不同的程度相互抵消，这意味着气候变暖有时会提高土壤湿度，有时则会降低土壤湿度。(Geophysical Research Letters, https://doi.org/10.1029/2025GL115044, 2025)

—科学撰稿人Saima May Sidik (@saimamay.bsky.social)

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

The El Niño Southern Oscillation (ENSO) is a natural climate phenomenon driven by interactions between the ocean and atmosphere in the tropical Pacific. In recent decades, major advances in observing and modeling ENSO have greatly improved our understanding, yet important challenges remain.

A recent article in Reviews of Geophysics highlights the recharge oscillator (RO) conceptual model, a simple mathematical representation of ENSO fundamental mechanisms. Here, we asked the lead author to provide an overview of ENSO, discuss the strengths and limitations of the RO model, and outline key open questions.

Why is the El Niño Southern Oscillation (ENSO) important to understand? 

ENSO events typically last around a year and occur in two phases: El Niño, when the central and eastern Pacific Ocean becomes unusually warm, and La Niña, when it becomes cooler than normal. These temperature shifts disrupt wind patterns and rainfall, triggering anomalies such as droughts, floods, tropical cyclones, and marine or terrestrial heatwaves. These impacts strongly affect ecosystems, agriculture, and economies around the world.

Although ENSO originates in the tropical Pacific, its influence extends globally.

Although ENSO originates in the tropical Pacific, its influence extends globally through atmospheric “teleconnections.” Because of its widespread effects, understanding and predicting ENSO is essential. Today, coupled ocean–atmosphere models and statistical methods allow scientists to forecast ENSO events up to a year in advance, making ENSO a key pillar of global seasonal climate prediction.

Over the past few decades, what advances have been made in observing and modeling ENSO?

Two major breakthroughs in the 1990s greatly advanced our ability to observe and model ENSO. First, on the observational side, the TAO mooring array across the equatorial Pacific and satellite altimetry provided continuous measurements of surface meteorological and subsurface ocean conditions—key data for understanding ENSO dynamics. Second, modeling evolved from simplified “intermediate” coupled models of the 1980s to more sophisticated coupled general circulation models (CGCMs), which simulate the full complexity of ocean–atmosphere interactions.

These advances provided deeper insight into the mechanisms driving ENSO. Importantly, subsurface observations also became essential for initializing ENSO forecasts improving their accuracy. Together, these observational and modeling tools laid the groundwork for modern ENSO research and prediction systems.

What are the benefits of using conceptual models to understand ENSO compared to other modeling methods?

Conceptual models of ENSO are simple mathematical representations that distill the phenomenon into just a few key variables—such as sea surface temperature in the central Pacific or equatorial ocean heat content. These models use basic equations to capture the core dynamics of ENSO, including the Bjerknes feedback (a positive loop that amplifies temperature anomalies) and slower equatorial ocean adjustment processes that help shift ENSO from one phase to another.

Conceptual models offer clarity and insight that complement the realism of full-scale simulations.

Because they focus on essential mechanisms, conceptual models are powerful tools for teaching and for gaining physical intuition. They also allow researchers to test hypotheses about ENSO dynamics in a controlled, simplified setting. Despite their simplicity, they can make useful quantitative predictions about ENSO features like amplitude or period, and are often used to diagnose biases in more complex climate models. In short, conceptual models offer clarity and insight that complement the realism of full-scale simulations.

What is the “recharge oscillator” model and why did you choose to focus on it?

The Recharge Oscillator (RO) is a conceptual model of ENSO introduced in the mid-1990s by Fei-Fei Jin. Unlike earlier models, it includes an explicit equation for subsurface ocean heat content, capturing ENSO’s “memory.” Its flexible mathematical structure has allowed researchers to gradually increase its realism while preserving simplicity and interpretability.

In our review, we show that the RO can now reproduce key ENSO characteristics, including its amplitude, dominant period, seasonal synchronization, and the tendency for El Niño events to be stronger than La Niña events. Remarkably, recent studies show that it can even rival complex dynamical models in terms of forecast skill. Thanks to its clarity, predictive power, and widespread use in the research community, the Recharge Oscillator was a natural focus for a dedicated review.

How does the recharge oscillator model aid in understanding ENSO response to climate change?

Climate models generally project increased near-surface ocean stratification under climate change. Most predict a weakening of the equatorial Pacific trade winds, though some show a strengthening—closer to observed trends in recent decades. These shifts in the background mean state can significantly affect ENSO behavior.

The Recharge Oscillator (RO) helps explore these effects by providing quantitative links between the mean state and ENSO characteristics such as amplitude, period, and asymmetry. This makes the RO a useful tool for understanding how future changes in stratification or winds might influence ENSO—and why model projections sometimes disagree. However, using the RO to study climate change impacts is still a developing field, partly because the way mean state changes affect RO parameters is not yet fully understood. Addressing this gap is highlighted in our review as a key direction for future research.

What are the primary challenges or limitations of the recharge oscillator model?

Klaus Wyrtki famously noted that “no two El Niño events are alike.” This insight underpins the challenge of ENSO diversity—the fact that some events peak in the eastern Pacific, while others peak farther west, with differing global impacts. Capturing this diversity remains a key limitation of the RO. While recent studies have proposed promising ways to represent these variations within the RO framework, more work is needed to develop a community consensus on a physically consistent approach.

Overcoming these limitations will strengthen the Recharge Oscillator’s relevance for studying both ENSO diversity and its links to broader climate variability.

Another challenge lies in modeling two-way interactions between ENSO and other climate modes, such as the Indian Ocean Dipole or Atlantic variability, which can influence ENSO through atmospheric teleconnections. These interactions are not accounted for in the RO. However, recent work introducing an extended Recharge Oscillator (XRO) offers a promising path forward. Overcoming these limitations will strengthen the RO’s relevance for studying both ENSO diversity and its links to broader climate variability.

What are some of the remaining questions where additional modeling, data, or research efforts are needed? 

In our review, we highlight 10 open research questions—many of which are well-suited for PhD or postdoctoral projects—centered on improving the RO and using it to explore broader ENSO dynamics. These include previously mentioned challenges such as understanding ENSO behavior in a warming climate, accounting for ENSO diversity, and modeling interactions with other climate modes. Several of these topics are already being actively explored, reflecting the vitality of the field.

To support future research, we will soon release open-source Python and Matlab versions of the RO, accompanied by a technical article detailing its numerical implementation and parameter fitting methods. This will make it easier for researchers to use and extend the RO framework to address today’s pressing ENSO questions—ultimately helping bridge conceptual models and complex Earth system simulations.

—Jérôme Vialard (jerome.vialard@ird.fr, 0000-0001-6876-3766), LOCEAN-IPSL, IRD-CNRS-MNHN-Sorbonne Universités, France; with feedback provided by review co-authors.

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: Vialard, J. (2025), Two equations that unlock El Niño, Eos, 106, https://doi.org/10.1029/2025EO255018. Published on 5 June 2025. This article does not represent the opinion of AGU, Eos, or any of its affiliates. It is solely the opinion of the author(s). Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

A recent Facebook post has highlighted a reasonably large slump landslide in a remote area of Alaska. Satellite images suggest that this occurred in late October or early November 2024.

Loyal reader Andrew McNown kindly highlighted a recent Facebook post that provided some images of a landslide that has partially blocked the Lakina River in Alaska. This is one of the images, posted by John Matthews:-

The landslide on the Lakina River in Alaska. Photograph posted to Facebook by John Matthews.

This image provides a more detailed view:-

The landslide on the Lakina River in Alaska. Photograph posted to Facebook by John Matthews.

A quick review of the Planet image catalogue suggests that the location of the landslide is [61.46578, -143.27085]:-

Satellite image of the landslide on the Lakina River in Alaska. Image copyright Planet, used with permission. Image dated 19 May 2025.

The landslide is about 350 m from crest to toe and 300 m wide, with a surface area of about 0.085 km2. From the images, it appears to be a rotational slump in fine-graimed (presumably) glacial materials. The event blocked the river but has breached; a small lake remains on the upstream side.

In terms of timing of the event, the landslide appears to be present on a Planet image dated 4 November 2024, but it appears to be absent on one dated 24 October 2024, so it occurred sometime in that window. The trigger is unclear – this seems to be an unusual time for a landslide of this type, but perhaps there was a rapid snowmelt event.

There is a large displaced rotational block in the images in which there is erosion of the toe. This provides some potential for a further valley-blocking landslide, although this is far from inevitable. Fortunately, there are few assets at risk in the immediate downstream area, but there could be some threat to groups using or camping beside the Lakina River.

Reference

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

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

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

Mountain landscapes are as much a product of erosion as they are of uplift. It is certainly true that glaciers can carve uplifted regions, increasing their topographic relief.

Using numerical modeling that integrates both river and glacial erosion across a time span that includes glacial-interglacial cycles, Bernard et al. [2025] flip the script on how we think glaciers shape mountains. The authors show that a “glacial sheltering” effect can lead to the development of extensive low-relief surfaces at moderate elevations, and they review the existence of candidate surfaces in Scandinavia and other locations.

A key finding is that such surfaces can not only be preserved by glaciation, but they can also emerge from it, and at variable elevations that are a function of ice volume. This is significant not just because humans are inspired by mountains and their topography: flat or low-relief surfaces play a large role as a reference elevation in explaining landscape evolution and in tectonic studies of uplift that make assumptions about where, when, and how such surfaces originated.

Citation: Bernard, M., van der Beek, P. A., Pedersen, V. K., & Colleps, C. (2025). Production and preservation of elevated low-relief surfaces in mountainous landscapes by Pliocene-Quaternary glaciations. AGU Advances, 6, e2024AV001610.  https://doi.org/10.1029/2024AV001610

—Peter Zeitler, Editor, AGU Advances

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

Shortly after President Joe Biden took office in 2021, he nominated Asmeret Asefaw Berhe, then a biogeochemist at the University of California, Merced, to oversee the Department of Energy’s (DOE) Office of Science. After a 15-month vetting process involving interviews, a mountain of paperwork, and, ultimately, a Senate confirmation, the AGU medalist became the first person of color and the first Earth scientist to hold the position. She served in the position for just under 2 years.

Now, with science and diversity programs under attack, she reflects on her path to leadership in a new commentary in AGU Advances. Berhe became familiar with DOE’s science program as a graduate student at the University of California, Berkeley. She later went on to receive DOE funding, collaborate with researchers from various national laboratories, and mentor scientists who went on to secure DOE positions. She says that combined with guidance from her mentors, these experiences helped her develop the skills she needed for her DOE appointment, not only in science but in managing, accounting, mediation, and ethical guidance.

Berhe, who was born in Eritrea and was one of only a few undergraduate women at Asmara University studying soil science, prioritized basic research, robust science communication, and promoting diversity in STEM (science, technology, engineering, and mathematics) in her DOE role. Providing opportunities in STEM for people from all walks of life starts with equalizing the distribution of funding, she writes. She cited an American Physical Society report that found, in 2018, 90% of federal research funding went to the top 22% of institutions, even though the vast majority of students—especially those from low-income backgrounds—attend other schools. Under Berhe’s tenure, the DOE began asking grant applicants to demonstrate plans for collaborating with schools less likely to receive funding, enabling scholars from diverse backgrounds to access DOE resources.

Berhe thinks recent efforts by some politicians to end diversity, equity, and inclusion (DEI) programs are partly because of a misconception around what DEI means. These programs are often misconstrued as serving only gender or racial minorities from urban environments, when, in fact, many are intended to serve a much wider range of Americans, she writes.

Today’s political climate sometimes leaves Berhe with feelings of despair. But she remains hopeful that with time, the next generation of scientists will benefit from opportunities like those she’s had. “Together, we will weather this storm,” she writes. (AGU Advances, https://doi.org/10.1029/2025AV001757, 2025)

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

Citation: Sidik, S. M. (2025), Former Department of Energy leader reflects on a changing landscape, Eos, 106, https://doi.org/10.1029/2025EO250211. Published on 4 June 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.

What’s Next for Science?

A 200,000 cubic metre rockslide in a remote area of Tibet on Sunday has left ten people dead or missing.

On 1 June 2025 a large rockslide occurred in Muta township in Chamdo (Qamdo) metropolitan area in Tibet. Note that Chinese media sources call this area Xizang Autonomous Region, but it is what most of us know as Tibet. Chinese media reports, which can be unreliable from Tibet, indicate that three people are confirmed to have been killed with a further seven reported to be missing. Two people were injured.

CGTN has a video online showing the landslide, which includes drone footage. The area has a dusting of snow, which makes interpretation difficult. CCTV also has the same footage posted to Youtube:-

This video includes imagery of the head scarp of the landslide:-

The head scarp of the 1 June 2025 rockslide at Muta in Tibet. Image from a video posted to Youtube by CCTV.

There is also a good image of the full length of the rockslide:-

The full extent of the 1 June 2025 rockslide at Muta in Tibet. Image from a video posted to Youtube by CCTV.

This landslide has a slightly unusual morphology, with much of the material from the upper portion of the slope stalled on the hillside. However, the mass of material in the valley floor is large, as this image shows:-

The lower portion of the 1 June 2025 rockslide at Muta in Tibet. Image from a video posted to Youtube by CCTV.

The landslide has blocked the valley and a small lake has started to develop. This will need to be managed. Note the run up of the landslide deposit on the opposite slope, which indicates that the mass was moving comparatively quickly. There are two people on the left of the image for scale.

The CGTN video suggests that the landslide was about 200,000 m3, which would be around 500,000 tonnes.

The precise location of this event is unclear to me. Chamdo is a large area centred on [31.1362, 97.2359]. A report by Xinhua suggests that the landslide occurred in Dengqen County (Dêngqên County), which is in the northwest of Chamdo, centred on [31.5396, 95.4156]. Wikidata indicates that Muta is located at [32.30957, 95.09376], and Google maps has this location as “Mutaxiang”, with “Muta town” a little to the west, so this is credible. We shall have to wait for a clear day to obtain satellite imagery to confirm this – given the limited loss of life, the landslide has probably not struck Muta itself.

As usual for China, especially when it comes to Tibet, the media footage includes lots of images of the response of the authorities to the disaster. Sadly, the likelihood of the missing people being recovered alive is very low.

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.

Antarctica is Earth’s most remote continent, barely touched by human activities.

It is, however, not immune to the kind of environmental damage that plagues more populated parts of the world. In a new study, researchers found chemicals originating from everyday personal care products (PCPs), such as cosmetics, detergents, pharmaceuticals, and deodorants, in Antarctic snow.

Contaminants in PCPs—loosely defined as semivolatile organic compounds that are industrially produced at a global scale, used in large volumes, and relatively persistent in the environment—are increasingly being recognized as pollutants. Both the Arctic Monitoring and Assessment Programme and the Scientific Committee on Antarctic Research have encouraged further research on PCP ingredients and the creation of monitoring plans for tracking their presence at the poles.

Looking for these pollutants, researchers collected 23 surface snow samples from 18 sites along the Ross Sea coast during the Antarctic summer of 2021–2022. Though some sampling locations were near areas with human activity, including Italy’s seasonally occupied Mario Zucchelli research station, the majority were situated hundreds of kilometers from human settlements.

The scientists reached these remote locations by piggybacking on helicopter rides transporting other teams maintaining weather stations or deploying scientific instruments. “This way we halved the impact of our sampling, because they needed to go there in any case,” said Marco Vecchiato, an analytical chemist at Ca’ Foscari University in Venice, Italy, who led the study.

Back in Italy, Vecchiato and his colleagues analyzed the snow samples under clean-room conditions to prevent contamination.

“This very different behavior during the season means that [PCPs] are very sensitive to the environmental conditions.”

They found PCP chemicals in every sample, with varying chemical concentrations suggesting different capacities for atmospheric transport. Of the 21 chemicals analyzed, three compound families were particularly notable. Salicylates, commonly used as preservatives in cosmetics (including lotions, shampoos, and conditioners) and pharmaceutical products, were the most prevalent, followed by UV filters associated with sunscreens. Fragrances such as musks were also detected.

Most of these substances were dissolved in the snow. The UV filter octocrylene, however, which has been associated with coral reef damage and banned in places like the U.S. Virgin Islands and Palau, was found bound to solid particles within the snow.

The researchers observed an unexpected seasonal variation in the amount of PCPs within the samples: Samples collected later in the summer had about 10 times higher PCP levels than those collected earlier in the season, though the relative proportions of each pollutant within a sample remained consistent.

Seasonal fluctuation suggests that Antarctic summer air circulation plays a role in transporting pollutants from distant sources to the continent’s interior. During summer, oceanic winds blowing inland dominate over winds originating from the polar plateau, which are stronger during the rest of the year. That shift may push pollutants far inland.

“This very different behavior during the season means that [PCPs] are very sensitive to the environmental conditions,” Vecchiato said.

One of the researchers presented the team’s preliminary findings at the European Geosciences Union General Assembly in May, and the scientists have a more comprehensive analysis currently underway, according to Vecchiato.

A Local or Distant Source

Finding organic pollutants in seemingly pristine polar environments isn’t surprising. In the 1960s, scientists found large concentrations of persistent organic pollutants (POPs), including the widely used pesticide DDT (dichlorodiphenyltrichloroethane), in Antarctica. POPs don’t degrade naturally and travel thousands of kilometers through the atmosphere, with some eventually getting trapped in snow and ice. Permanently frozen places such as glaciers and polar regions become natural traps. Starting in the early 2000s, the United Nations’ Stockholm Convention on Persistent Organic Pollutants established international cooperative efforts to eliminate or restrict the production and use of POPs.

Though they might travel by a mechanism similar to that used by persistent organic pollutants, unlike POPs, PCPs “do break down in the environment,” said Alan Kolok, a professor of ecotoxicology at the University of Idaho. However, “if those fragrances are not coming from the [research] stations themselves,” he asked, “where are they coming from?”

“Thousands of people are currently accessing the Antarctic continent, and my conclusion is that wherever we humans go, we bring contaminants.”

To rule out a local origin for the PCP pollutants, researchers analyzed sewage from the Mario Zucchelli research station. The outpost did contribute some pollution, but the relative abundance of each compound in the sewage differed from that found in the snow, suggesting that the PCPs detected in the broader Antarctic environment likely originated from more distant sources.

“My suspicion is that for these types of compounds—personal care products, pharmaceutical products—there must be a local source,” said Ricardo Barra Ríos, an environmental scientist at the Universidad de Concepción in Chile who has analyzed PCP pollution in Antarctic coastal waters related to research stations. “Thousands of people are currently accessing the Antarctic continent, and my conclusion is that wherever we humans go, we bring contaminants.”

Vecchiato disagreed. In a separate study published earlier this year, he and other colleagues found PCPs, including fragrances and UV filters, in the snows of the Svalbard archipelago in the Arctic. In that study, the researchers linked the presence of these compounds to atmospheric patterns that carried pollution from northern Europe and the northwestern coast of Russia.

“Most of these contaminants should have a limited mobility, but actually, we found them in remote regions,” Vecchiato said. “Does that mean that the models are wrong or that our analysis is wrong?” he asked. “No, probably we are missing a piece [of the puzzle], or maybe the use of these contaminants is so huge that we still have a relevant concentration in remote areas, even if they should not be prone to this kind of transport.”

—Javier Barbuzano (@javibar.bsky.social), Science Writer

Citation: Barbuzano, J. (2025), Is your shampoo washing up in Antarctica?, Eos, 106, https://doi.org/10.1029/2025EO250209. Published on 3 June 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.

This is an authorized translation of an Eos article. Esta es una traducción al español autorizada de un artículo de Eos.

En 2017, Paulo Tarso Oliveira, profesor de hidrología en la Universidad de São Paulo, se encontró con una noticia sobre una pequeña aldea a orillas del río São Francisco, uno de los principales ríos del noreste de Brasil. El artículo informaba que los habitantes estaban presentando tasas inusualmente altas de hipertensión arterial, y relacionaba esta anomalía con el clima seco de la región y el bajo caudal del río. A medida que el nivel freático descendía, el agua oceánica comenzaba a infiltrarse hacia el agua subterránea de la región, elevando los niveles de sal en el suministro y provocando problemas de salud entre la población.

“Muchas veces, la gente no se da cuenta, pero las aguas superficiales y subterráneas están conectadas y deben considerarse como un todo”.

Intrigado, Oliveira investigó más a fondo. Más adelante descubrió que el flujo del río estaba disminuyendo porque los pozos estaban extrayendo agua del acuífero subyacente. “Muchas veces, la gente no se da cuenta, pero las aguas superficiales y subterráneas están conectadas y deben considerarse como un todo”, señaló Oliveira.

En lugares donde el nivel freático se encuentra bajo el lecho de un río, el río puede filtrar agua hacia el acuífero subyacente. Este proceso, conocido como filtración del caudal fluvial, ocurre de forma natural dependiendo de las formaciones geológicas subyacentes y los niveles de agua subterránea. Sin embargo, la construcción de pozos que extraen agua en exceso de los acuíferos puede intensificar este fenómeno.

Oliveira y sus colegas descubrieron que la situación en la cuenca del São Francisco no es un caso aislado. Al evaluar pozos en todo Brasil, los investigadores encontraron que en más de la mitad de ellos el nivel del agua estaba por debajo del nivel de los arroyos cercanos.

Mapeo de pozos

En 2023, Oliveira y el estudiante de maestría José Gescilam Uchôa comenzaron a mapear los ríos de Brasil para identificar zonas en riesgo de pérdida de agua. Se basaron en datos públicos sobre niveles de ríos y ubicación de pozos, proporcionados por el Servicio Geológico de Brasil. Sin embargo, la mayoría de los pozos registrados carecían de información suficiente. Como resultado, se enfocaron en 18,000 pozos con datos completos, distribuidos a lo largo de miles de ríos en el país.

Los investigadores compararon el nivel del agua en cada pozo con la elevación del arroyo más cercano. En el 55 % de los casos, el nivel del agua en los pozos era inferior a la elevación de los arroyos vecinos.

José Uchôa realiza mediciones en un río de São Paulo. Crédito: Laboratorio de Hidráulica Computacional, Universidad de São Paulo

“Nuestros datos sugieren que el uso de aguas subterráneas está afectando significativamente el caudal de los ríos”, señaló Uchôa. “Este es, y seguirá siendo, un motivo de creciente preocupación para la gestión del agua en el país”.

El estudio, publicado en Nature Communications, también identificó regiones críticas, incluida la cuenca del São Francisco, donde más del 60 % de los ríos podrían estar perdiendo agua debido a la intensa extracción subterránea. Esta extracción se asocia principalmente con actividades de irrigación.

En la cuenca del Verde Grande, en el este de Brasil, donde la irrigación representa el 90 % del consumo de agua, el 74 % de los ríos podrían estar perdiendo agua hacia los acuíferos.

Oliveira considera que los resultados son conservadores y que la situación podría ser aún peor, ya que los investigadores no tomaron en cuenta los pozos ilegales. Un estudio realizado en 2021 por el geólogo Ricardo Hirata, de la Universidad de São Paulo, estimó que alrededor del 88 % de los 2.5 millones de pozos en Brasil son ilegales, al carecer de licencia o registro para operar.

Hirata, quien no participó en la nueva investigación, advirtió que el estudio se basó únicamente en el 5 % de los pozos existentes, ubicados principalmente en regiones donde la explotación de aguas subterráneas es más intensa.

“Quizá esto también esté ocurriendo en otras regiones del país con alta demanda de irrigación, y simplemente no lo sabemos por falta de datos”.

Hirata también subrayó que, aunque los investigadores identificaron ríos que potencialmente están perdiendo agua hacia los acuíferos, esos datos por sí solos no son suficientes para determinar si los ríos realmente se están secando. Para evaluar eso, se deben considerar otros factores, como la cantidad de agua extraída del acuífero en comparación con el caudal del río, el grado de conexión entre el acuífero y el río, y cuánta agua se extrae del acuífero en relación con las variaciones estacionales del caudal.

“El hecho de que el nivel de agua de un pozo esté por debajo del de un río cercano no necesariamente afecta al río o al acuífero”, explicó Hirata.

Las áreas identificadas como críticas por el estudio se ubican principalmente en regiones áridas, donde ya se esperaba que ocurriera filtración del caudal de manera natural, señaló André F. Rodrigues, hidrólogo de la Universidad Federal de Minas Gerais, quien no participó en la investigación.

El estudio es relevante porque resalta un problema creciente, dijo Rodrigues, pero se necesitan análisis más locales para obtener una imagen más detallada del problema y considerar, por ejemplo, los efectos del clima y los cambios estacionales. “Quizá esto también esté ocurriendo en otras regiones del país con alta demanda de irrigación, y simplemente no lo sabemos por falta de datos”, comentó.

Un problema en crecimiento

La expansión descontrolada de pozos y la extracción excesiva de agua subterránea no solo afectan la salud de las personas, el abastecimiento de agua y la agricultura, sino que también pueden desestabilizar el suelo, provocando hundimientos (subsistencia). Fenómenos similares se han observado en regiones de China, Estados Unidos e Irán.

El panorama no es nada alentador para Brasil. Es probable que la cantidad de pozos se multiplique, ya que se espera que las áreas de riego se incrementen en más del 50 % en los próximos 20 años, según la agencia nacional del agua de Brasil.

“Probablemente veremos un círculo vicioso de degradación, en el que la disminución en la cantidad y calidad del agua superficial, combinada con el aumento de los períodos de sequía, obligará a los agricultores a perforar más pozos para mantener la producción de alimentos, intensificando aún más la extracción de aguas subterráneas y agravando el problema”, advirtió Oliveira.

La sobreexplotación de aguas subterráneas es una preocupación a nivel mundial. La mayoría de los acuíferos han mostrado un descenso en lo que va del siglo XXI, y los estudios por modelado sugieren que la filtración de caudales será más común en las próximas décadas. Aun así, este problema ha sido en gran medida ignorado en regiones tropicales como Brasil, que alberga el 12 % de los recursos de agua dulce renovables del planeta.

Esta falta de atención se debe en parte al escaso financiamiento y vigilancia, y en parte a una creencia persistente de que en los países tropicales y húmedos los ríos suelen ganar agua de los acuíferos y no perderla, mencionó Oliveira. “Debemos actuar ahora si queremos evitar que regiones enteras queden devastadas en el futuro”.

Los investigadores hacen un llamado a realizar más estudios y establecer un monitoreo sistemático de los pozos para identificar las zonas más secas y evaluar el impacto de nuevos pozos sobre los ríos. Actualmente, Brasil solo cuenta con 500 pozos de observación monitoreados constantemente por el gobierno, en comparación con los 18,000 que existen en Estados Unidos, a pesar de que ambos países tienen extensiones territoriales similares. “La vigilancia es extremadamente importante y está tremendamente subestimada”, enfatizó Uchôa.

—Sofia Moutinho (@sofiamoutinho.bsky.social), Escritora de ciencia

This translation by Saúl A. Villafañe-Barajas (@villafanne) was made possible by a partnership with Planeteando and Geolatinas. Esta traducción fue posible gracias a una asociación con Planeteando y Geolatinas.

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

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

Rock glaciers are debris landforms found in many mountain ranges on Earth. They represent the movement of permanently frozen ground over long periods of time and can be used to understand how climate change is affecting permafrost.  

A new article in Reviews of Geophysics explores the use of “Rock Glacier Velocity” to measure how fast these landforms move each year, and its relationship with climatic factors. Here, we asked the authors to give an overview of Rock Glacier Velocity, how scientists measure it, and what questions remain.

What makes rock glaciers unique landforms? 

Rock glaciers primarily form where the ground temperature ranges from approximately -3 to 0°C. Generated by gravity-driven deformation of permafrost, rock glaciers exhibit distinct morphologies indicative of a cohesive flow. The motion mechanism, known as rock glacier creep, involves shearing in one or more layers (i.e., shear horizons) at depth within the permafrost and deformation of the frozen materials above. Changes in rock glacier creep rates depend primarily on changes in ground temperature. Rock glaciers provide a unique opportunity to indirectly document the evolution of permafrost temperatures in mountainous regions.

Remote sensing and field photos of rock glaciers. Credit: Hu et al. [2025], Figure 1

What is “Rock Glacier Velocity” and why is it important to measure? 

“Rock Glacier Velocity (RGV)” refers to the time series of annualized surface velocity reflecting the movement related to rock glacier creep. Since 2022, RGV has been accepted by the Global Climate Observing System (GCOS) as an Essential Climate Variable (ECV) Permafrost Quantity. An ECV is defined as “a physical, chemical, or biological variable (or group of linked variables) that is critical for characterizing the Earth’s climate.” An ECV Quantity is a measurable parameter necessary for characterizing an ECV. Rock Glacier Velocity is instrumental in assessing the state of permafrost under climate change, especially in places where direct monitoring is scarce. From a climate-oriented perspective, relative changes in Rock Glacier Velocity are significant.

What are the main factors that control Rock Glacier Velocity? 

Rock Glacier Velocity is collectively controlled by the geomorphologic features such as slope and landform geometry, as well as the thermo-mechanical properties of the frozen ground, such as ice content, subsurface structure, temperature, and the presence of unfrozen water under permafrost conditions. On a given rock glacier, relative changes in surface velocity over time usually reflect the climatic impacts, with temperature forcing being the dominant factor, especially when temperatures approach 0°C.

How do scientists observe and monitor Rock Glacier Velocity at different spatial scales? 

An illustration showing different survey methods for quantifying Rock Glacier Velocity. Credit: Hu et al. [2025], Figure 5a

Rock Glacier Velocity can be observed and monitored using in-situ and remote sensing methods. Global Navigation Satellite System (GNSS), theodolite, and total station surveys, provide point-based in-situ measurements. Regional-scale surveys typically employ remote sensing techniques, such as laser scanning, photogrammetry, radar interferometry, and radar offset tracking. In-situ RGV time series’ are rare and have mostly been provided from the European Alps, but they can be more than 20 years long. The goal is to leverage the experience gained from the systematic compilation of those in-situ time series to expand the RGV collection to regional-scale surveys using remote sensing techniques.

What kinds of patterns have been observed in Rock Glacier Velocity? 

According to the Rock Glacier Velocity data from across the European Alps, rock glaciers have generally accelerated alongside increasing air temperatures over the past three decades. At the interannual scale, RGV exhibits a regionally synchronous pattern with distinct acceleration phases (i.e., 2000–2004, 2008–2015, and 2018–2020) which are interrupted by deceleration or a steady kinematic state. However, systematic monitoring and documentation of Rock Glacier Velocity is currently lacking in many parts of the world.

How is climate change expected to influence Rock Glacier Velocity? 

Among the climatic factors, multi-annual air temperature changes primarily influence Rock Glacier Velocity by altering the ground thermal state of rock glaciers. Snow cover acts as an insulating layer whose development varies from year to year, causing the ground temperature to deviate from the air temperature on an interannual scale.

In general, warmer ground temperatures favor rock glacier movement. This pattern is expected to occur in many rock glaciers in the future as the climate continues to warm.  When the ground temperature reaches 0°C, some rock glaciers experience drastic acceleration. However, consequent thawing at the tipping point of 0°C causes the rock glacier creep to decline.

What are some of the remaining questions where additional modeling, data, or research efforts are needed? 

First, a standardized strategy for monitoring Rock Glacier Velocity using different methods is under development. We call for more systematic and consistent velocity measurements that can be used to generate Rock Glacier Velocity data products.

Second, the mechanisms linking climatic factors to Rock Glacier Velocity still need to be explored further, such as whether water infiltrates the partially frozen body of a rock glacier and how cold temperatures influence winter deceleration.

Additionally, an in-depth understanding of the relationship between Rock Glacier Velocity, environmental factors, and permafrost conditions requires observations combined with laboratory work and numerical modeling. This is necessary in order to incorporate rock glacier processes into land surface models and predict future changes in a warming climate.

—Yan Hu (huyan@link.cuhk.edu.hk, 0000-0001-8380-276X), University of Fribourg, Switzerland; and Reynald Delaloye (0000-0002-2037-2018), University of Fribourg, 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: Hu, Y., and R. Delaloye (2025), Rock Glacier Velocity: monitoring permafrost amid climate change, Eos, 106, https://doi.org/10.1029/2025EO255017. Published on 3 June 2025. This article does not represent the opinion of AGU, Eos, or any of its affiliates. It is solely the opinion of the author(s). Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.

Nine people have been killed in a series of landslides, triggered by heavy rainfall, that have struck an army camp.

At about 7 pm local time on 1st June 2025, a series of landslides struck an army camp at Chaten in the Lachen District of Sikkim in India. It is believed that nine people have been killed, although at the time of writing six of these people were still missing, including an army officer, his wife and daughter.

Chaten is located at [27.7188, 85.5581]. This is a Google Earth image of the site, collected in March 2022:-

Google Earth image of the site of the 1 June 2025 landslide at Chaten in Sikkim, India.

The best imagery of the landslides that I have found is on a Youtube video posted by Excelsior News:-

This still captures the site well:-

The 1 June 2025 landslides at Chaten in Sikkim, India. Still from a video posted to Youtube by Excelsior News.

The image shows two main landslide complexes (plus one in the background). Each consists of a series of shallow slips on steep terrain – the one on the left has at least three initial failures, on the right there are also at least three). These have combined to create open hillslope landslides that have stripped the vegetation and surficial materials. Note the very steep lower slopes to the river.

These shallow landslide complexes are characteristic of extremely intense rainfall events, which saturate the soil and regolith from the boundary with the underlying bedrock. This causes a rapid loss of suction forces and a reduction in effective stress, triggering failure. The high water content of the soil then promotes mobility.

It is interesting to note that the natural vegetation has been removed from these slopes. It would be premature to assert that this was an underlying cause of the landslides, but it may have been a factor.

It appears that there has also been erosion of the riverside cliffs, which has left other parts of the camp in severe danger.

Sadly, given the terrain and the availability of people to participate in a rescue (which is one advantage of an event in an army camp), the prospects for those who are missing are not postive.

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.

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.

In a move that worried politicians and space scientists alike, President Trump announced on 31 May that he will withdraw his nomination of Jared Isaacman for the position of NASA administrator, according to Semafor. Isaacman’s nomination received bipartisan support and he was expected to easily pass a Senate confirmation vote in a few days.

 
Related

•  White House to pull NASA nominee Isaacman
•  Trump withdraws Jared Isaacman’s nomination as NASA administrator
• The Planetary Society reissues urgent call to reject disastrous budget proposal for NASA
• Get Involved: AGU Science Policy Action Center

This is seismic.Isaacman had clearly articulated a strong support for science, and the withdrawal of his nomination yet further imperils NASA's Science Mission Directorate.www.semafor.com/article/05/3…

— Paul Byrne (@theplanetaryguy.bsky.social) 2025-05-31T20:49:52.860Z

Trump cited a “thorough review of prior associations” as the reason for withdrawing the nomination. It was not immediately clear whether he was referring to Isaacman’s past donations to Democrats or his ongoing associations with former DOGE head and SpaceX CEO Elon Musk, who spent the weekend distancing himself from the president. Both of these associations were public at the time of Isaacman’s nomination.

Isaacman, a billionaire, private astronaut, and CEO of credit processing company Shift4 Payments, was questioned by the Senate Committee on Commerce, Science, and Transportation in a nomination hearing in April. Despite a few contentious moments regarding Isaacman’s association with Musk and some waffling over NASA’s Moon-to-Mars plan, the committee ultimately approved Isaacman’s nomination with strong bipartisan support.

When Trump announced Isaacman’s nomination in December 2024, very early for a NASA administrator, space scientists greeted the news with cautious optimism. Isaacman had vocally expressed support for the imperiled Chandra X-ray Observatory, and is a known space enthusiast.

Now, with the withdrawal of his nomination just days after a president’s budget request that would devastate Earth and space science, scientists fear for the future of NASA.

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

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.

Source: AGU Advances

In the southern flanks of the Indian Ocean and the central and eastern Pacific, just north of the Antarctic Circumpolar Current, lie the Subantarctic Mode Waters. As part of the global ocean conveyor belt, these large masses of seawater transfer substantial amounts of heat and carbon northward into the interiors of the Indian and Pacific Oceans. These waters hold about 20% of all anthropogenic carbon found in the ocean, and their warming accounted for about 36% of all ocean warming over the past 2 decades—making them critical players in Earth’s climate system.

Prior research has suggested Subantarctic Mode Waters form when seawater flowing from warm, shallow subtropical regions mixes with water flowing from cold, deep Antarctic regions. But the relative contributions of each source have long been debated.

Fernández Castro et al. used the Biogeochemical Southern Ocean State Estimate model to investigate how these water masses form. The model incorporates real-world physical and biogeochemical observations—including data from free-roaming floats—to simulate the flow and properties of seawater. The researchers used it to virtually track 100,000 simulated particles of water backward in time over multiple decades to determine where they came from before winding up in Subantarctic Mode Waters.

The particle-tracking experiment confirmed that subtropical and Antarctic waters indeed meet and mix in all areas where Subantarctic Mode Waters form but offered more insight into the journeys and roles of the two water sources.

In the Indian Ocean, the simulations suggest, Subantarctic Mode Waters come mainly from warm, shallow, subtropical waters to the north. In contrast, in the Pacific Ocean, Subantarctic Mode Waters originate primarily from a water mass to the south known as Circumpolar Deep Water.

Along their southward flow to the subantarctic, subtropical waters release heat into the atmosphere and become denser, while ocean mixing reduces their salinity. Meanwhile, the cooler Circumpolar Deep Water absorbs heat and becomes fresher and lighter as it upwells and flows northward from the Antarctic region to the subantarctic.

These findings suggest that Subantarctic Mode Waters affect Earth’s climate differently depending on whether they form in the Indian or Pacific Ocean—with potential implications for northward transport of carbon and nutrients. Further observations could help confirm and deepen understanding of these intricacies. (AGU Advances, https://doi.org/10.1029/2024AV001449, 2025)

—Sarah Stanley, Science Writer

Citation: Stanley, S. (2025), On the origins of Subantarctic Mode Waters, Eos, 106, https://doi.org/10.1029/2025EO250207. Published on 2 June 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.

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

On 30 May 2025, a rock slope major failure occurred at a quarry at Gunung Kuda, which is located on the edge of Cipanas Village in Dukupuntang District, Cirebon Regency, West Java, Indonesia. At the time of writing, it has been confirmed that 19 people were killed in the accident, with a further six people remaining missing. Four people were injured.

The location of the failure is [-6.7754, 108.4022]. This is the site in Google Earth:-

Google Earth image of the site of the 30 May 2025 landslide at Gunung Kuda mine.

Universitas Siber Asia has a good article about the event, in Indonesian but it translates well. There is also some Youtube footage of the site immediately after the failure:-

There are other videos circulating of a dramatic rock slope failure, but the ones that I have seen are not this event.

There is also some very clear drone footage of the site after the failure:-

This includes this view of the landslide:-

Drone footage of the site of the 30 May 2025 landslide at Gunung Kuda mine. Still from a video posted to Youtube by Andrea Ramadhan.

The geological structure of this quarry is very complex, with many joints being visible in the above image that would promote instability.

The Universitas Siber Asia article describes a site with a very poor history regarding instability:-

“The Geological Agency said the mine location was in a zone of high soil movement vulnerability, with a probability of landslide of more than 50%. The Head of the West Java Energy and Mineral Resources Office, Bambang Tirto Mulyono, stated that the main cause was the wrong mining method, namely digging from under the cliff, making the soil structure fragile. Repeated warnings from the Energy and Mineral Resources and police lines since February 2025 have been ignored by mine managers. As a result, the West Java Provincial Government revoked the mining permit that was supposed to be valid until October 2025 and closed the site permanently.”

Interestingly, the quarrying was licensed, albeit with substantial safety concerns. Detik Jabar describes the long term worries about the site:–

“…the Head of the West Java Energy and Mineral Resources Office, Bambang Tirto Mulyono, stated that the incident was caused by a faulty mining method carried out by the mine management. Warnings have been conveyed many times by the Energy and Mineral Resources department, and even preventive measures have been taken by the police.”

“We have repeatedly warned the mining authorities, even in a loud tone. The Cirebon Police have also installed a police line at the location since February because the mining methods carried out are not in accordance with safety standards. Mining should have been done from above, not from below,” said Bambang when met at the scene.“

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.

The textbook explanation of how hailstones grow goes something like this: Nuclei collect frozen layers as they are repeatedly lofted up and fall through clouds. But scientists have had hints that this up-down cycle doesn’t always reflect real hailstones’ journeys. Now researchers have revived an old technique to track dozens of hailstones. The new results, published in Advances in Atmospheric Sciences, suggest that many hailstones take simpler paths.

The idea that hailstones grow as they repeatedly rise and fall on repeat arose as a way to explain stones’ alternating layers of different transparencies, said Xiangyu Lin, an atmospheric scientist at Peking University in Beijing and an author on the new study. But scientists don’t have any direct observations of individual hailstones’ paths in clouds because the severe storms that produce hail are difficult, even dangerous, to observe.

“The vast majority of our understanding of how hail grows has come from numerical modeling,” said Matthew Kumjian, an atmospheric scientist at Pennsylvania State University who wasn’t part of the study. The new research is “a nice piece of experimental evidence” to validate those models, he said.

“Over the past 8 years, we have collected more than 3,000 hailstones.”

At a seminar at Peking University in 2018, Kumjian showed a simple arcing trajectory—rather than a yo-yoing one—for simulated hailstones. Seeing those results, one of Lin’s colleagues at Peking University, atmospheric scientist Qinghong Zhang, wondered whether she could find real hailstones that followed a similar path. That year she started collecting hailstones, using social media to ask the public to save the icy orbs. “Over the past 8 years, we have collected more than 3,000 hailstones,” she said.

To trace the hailstones’ trajectories, the team turned to stable isotopes. At lower altitude, the ice that forms on hailstones tends to have a greater concentration of heavier isotopes of hydrogen and oxygen than the ice that forms higher up. Researchers can measure the ratio of heavy and light isotopes in a layer, providing a postmark of sorts for the altitude at which the ice originated.

The scientists analyzed 27 hailstones from nine different storms spread across eastern China. They sliced each stone in half to reveal its layers. Then they cut the hailstones down layer by layer, so they could melt each layer and measure its isotopes. To find the link between isotope concentrations and height in a storm cloud, the team used temperature, humidity, and pressure data from weather balloons that floated through the atmosphere near each storm.

Hailing from Where?

The isotopes showed that of the hailstones they analyzed, only one had more than one upward flight segment. A few hailstones grew at a relatively constant altitude, and 16 either rose or fell steadily as they grew.

Eight hailstones ascended once before falling to the ground. These eight hailstones were significantly larger than the other stones, Lin said. Hailstones primarily grew between −10°C and −30°C, the team found. With their up-and-down path, these eight stones seem to have spent more time in that zone, causing them to grow larger than others.

Many hailstones are not perfect spheres. Credit: Xiangyu Lin

Scientists used stable isotope analysis on hailstones some 50 years ago, but the technique fell out of favor, Kumjian said. Many of those early studies analyzed a small number of stones from few storms or sometimes a single storm. The new study is “bringing back this old type of analysis with modern methods,” he said.

But the analysis required assumptions that might cloud results. For instance, updrafts can mix air from different altitudes, Kumjian said. That can affect the isotopes in a hailstone’s layers.

Scientists are still exploring questions about hail across a range of scales from stone to storm. Though researchers know what sorts of storms can produce damaging hail, it’s hard to predict which will rain down baseball-sized stones or where exactly hail will fall. Meanwhile, the physics of hailstones’ growth is tricky. Researchers typically model stones as perfect spheres—a far cry from the bumpy lumps that fall from the sky. But those shapes affect how fast hail falls and the damage it can produce, Kumjian said.

“It’s a very exciting time in the hail world. We’re going to learn a lot in the coming years.”

Researchers are using modeling, radar observations, and isotope studies such as this one to improve forecasts. Hail can knock out crops, damage structures, and shatter solar panels. Even 10 minutes of warning is enough for people to move cars and prevent damage, Zhang said.

Kumjian is part of a team that is launching instrumented Styrofoam spheres into clouds that could provide insights on actual paths taken by stones. Zhang’s team is continuing to study isotopes in layers, now looking at larger stones that formed in storms over Italy. “It’s a very exciting time in the hail world,” Kumjian said. “We’re going to learn a lot in the coming years.”

—Carolyn Wilke (@CarolynMWilke), Science Writer

Citation: Wilke, C. (2025), Isotopes map hailstones’ paths through clouds, Eos, 106, https://doi.org/10.1029/2025EO250206. Published on 30 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.

Sixty-seven extreme heat events have occurred since May 2024. All of these events—including a deadly Mediterranean heat wave in July 2024, an unprecedented March 2025 heat wave in central Asia, and extreme heat in South Sudan in February 2025—broke temperature records, caused major harm to people or property, or did both.

According to a new analysis, each of these extreme events was made more likely by climate change. The number of days with extreme heat is now at least double what it would have been without climate change in 195 countries and territories. Climate change added at least an extra month of extreme heat in the past year for 4 billion people—half the world’s population. 

“The numbers are staggering.”

“There’s really no corner of the globe that has been untouched by climate-driven extreme heat,” said Kristina Dahl, a climate researcher at the climate change research and communication nonprofit Climate Central who was part of the report team. “Half the world’s population is experiencing an extra month of extreme heat. The numbers are staggering.”

The authors of the report say it serves as a stark reminder of the dangers of climate change and the urgent need for better early-warning systems, heat action plans, and long-term planning for heat events across the globe. 

The report was created by scientists at Climate Central; World Weather Attribution, a climate research group; and the Red Cross Climate Centre. 

More Frequent Heat

In the new report, scientists calculated the number of days between 1 May 2024 and 1 May 2025 in which temperatures in a country or territory were above 90% of the historical temperatures from 1991 to 2020. Then, they analyzed how many of these extreme heat days were made more likely by climate change using the climate shift index, a methodology developed by Climate Central that compares actual temperatures to a simulated world without human-caused climate change. 

The team found that climate change made extreme heat events more likely in every country.

Over all the countries and territories, climate change added the greatest number of extreme heat days to the Federated States of Micronesia (57 days), and Aruba had the most extreme heat days in total over the past year, 187 days. The report’s authors estimate that in a world without climate change, Aruba would have experienced just 45 days of extreme heat.

Other Caribbean and Oceanic islands were among the countries and territories most strongly affected by climate change. People in the United States experienced 46 days of extreme heat, 24 of which were added by climate change. 

The authors of the report calculated the number of extreme heat days added by climate change in the past year. Credit: World Weather Attribution, Climate Central, and Red Cross Red Crescent Climate Centre

Of the 67 extreme heat events that occurred in the past year, the one most influenced by climate change was a heat wave that struck Pacific islands in May 2024. Researchers estimated the event was made at least 69 times more likely by climate change. 

The findings are not a surprise to Nick Leach, a climate scientist at the University of Oxford who was not involved in the report. “We’ve understood the impact of climate change on temperature and extreme heat for quite some time…[including] how it’s increasing the frequency and intensity of extreme heat,” he said. Research has consistently shown that heat events on Earth are made more likely, more intense, and longer lasting as a result of climate change. 

“Only comprehensive mitigation, through phasing out fossil fuels, will limit the severity of future heat-related harms.”

Leach said the new report gives a good overview of how climate change is influencing heat waves worldwide. However, defining extreme heat as temperatures above the 1991–2020 90th percentile creates a relatively broad analysis, he said. Studies using a more extreme definition of extreme heat may be more relevant to the impacts of extreme heat, and studies estimating those impacts are typically more policy relevant, he said.

The report’s authors chose the 90% threshold because heat-related illness and mortality begin to increase at those temperatures, Dahl said. 

Taking Action on Heat Waves

For rising global temperatures, “the causes are well known,” the report’s authors wrote. Burning of fossil fuels such as coal, oil, and gas has released enough greenhouse gases to warm the planet by 1.3°C (2.34°F; calculated as a 5-year average); 2024 marked the first year with average global temperatures exceeding 1.5°C (2.7°F) above preindustrial temperatures.

“Only comprehensive mitigation, through phasing out fossil fuels, will limit the severity of future heat-related harms,” the authors wrote.

Extreme heat puts strain on the human body as it tries to cool itself. This strain can worsen chronic conditions such as cardiovascular problems, mental health problems, and diabetes and can cause heat exhaustion and heat stroke, which can be deadly. Extreme heat is particularly dangerous for already-vulnerable populations, including those with preexisting health conditions, low-income populations lacking access to cool shelter, and outdoor workers. 

Heat Action Day on 2 June, hosted by the International Federation of Red Cross and Red Crescent Societies, raises awareness of heat risks across the globe. This year, the day of action will focus on how to recognize signs of heat exhaustion and heat stroke. Dahl recommends using the Centers for Disease Control and Prevention tips on heat and health to stay safe. “Most heat-related illness and death is preventable,” she said.

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

Citation: van Deelen, G. (2025), Climate change made extreme heat days more likely, Eos, 106, https://doi.org/10.1029/2025EO250208. Published on 30 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.

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

The relationship between phytoplankton production and dissolved iron affects the net annual air-sea exchange of carbon dioxide and impacts the ability of the subantarctic Southern Ocean to act as a carbon sink.

Traill et al. [2025] combine 27 years of monitoring data from a time series site in the subantarctic Southern Ocean south of Australia with ship-based observations to develop a composite seasonal cycle of productivity and dissolved iron. The seasonal cycle shows three phases that are defined by controls on production by light and multiple iron sources (Phase 1), iron limitation (Phase 2), and biomass decline from a shift to net heterotrophy and recycled nutrients (Phase 3). The seasonal cycle of coupling between dissolved iron and productivity provides validation of ocean biogeochemical models and informs understanding of variability associated with changing Southern Ocean iron supply mechanisms. 

Citation: Traill, C. D., Rohr, T., Shadwick, E., Schallenberg, C., Ellwood, M., & Bowie, A. (2025). Coupling between the subantarctic seasonal iron cycle and productivity at the Southern Ocean Time Series (SOTS). AGU Advances, 6, e2024AV001599.  https://doi.org/10.1029/2024AV001599

—Eileen Hofmann, Editor, AGU Advances  

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.