Feed aggregator

A new paper (Faral et al. 2025) provides details of a seismically-triggered landslide cascade and tsunami that killed up to 12,000 people.

On 22 November 1815, a very significant landslide disaster occurred in Bali, in what is now Indonesia, killing between 10,000 and 12,000 people. A very interesting new paper (Faral et al. 2025) in the journal Geomorphology has sought to investigate and understand this catastrophe.

The event, which occurred on Buyan-Bratan caldera, was triggered by the Mw=7.3 1815 Bali earthquake offshore. This triggered a translational landslide near the peak of the caldera. To provide a context, this is a Google Earth image of the site:-

Google Earth image of the site of the 22 November 1815 Gejer Bali disaster.

This image shows the steep caldera at the top, with the steep slopes down to the sea.

Faral et al. (2025) have identified the location of the initial failure. This is surprisingly clear on the Google Earth imagery:-

Google Earth image of the source zone of the 22 November 1815 Gejer Bali disaster.

The failure was a translational landslide – the source zone is so clear because it was left steep slopes in the rear scar and the lateral scarps, which now have dense vegetation. These lateral scarps are 30 to 35 m tall, delineating a landslide with a source area of 2.38 km2 and a volume of about 64 million cubic metres. Descriptions of the event highlight that the area had been subject to “intense and prolonged rainfall”, which may have contributed to the instability.

Faral et al. (2025) have undertaken detailed work to understand the characteristics of the landslide as it travelled the c.17 km to the coast. I think the initial path is quite easy to spot on the imagery:-

Google Earth image of the source zone and possible initial track (my own interpretation) of the 22 November 1815 Gejer Bali disaster.

Lower down the path becomes much less clear, but Faral et al. (2025) have used a range of mapping and stratigraphic techniques to try to understand it. In an earlier paper (Faral et al. 2024), the team has also mapped the c.15 villages that historical records indicate were destroyed by the landslide. They conclude that the initial failure transitioned into a debris avalanche and then a debris flow, eroding saturated sediment from the lower slopes and spreading laterally. This huge landslide moved enormous blocks of rock – one of which is 9 metres long for example – that are now scattered on the lower slopes.

Eventually the landslide reached the sea, and there are reports of a local tsunami. Faral et al. (2025) psotualte that this was most likely generated by the landslide rather than by the original earthquake. They note that there are no known deposits from the tsunami, which supports the notion of a smaller, localised event.

The 22 November 1815 Gejer Bali disaster was a very significant event that warrants more attention – I thank the authors of the two papers from highlighting and investigating this most fascinating disaster. I’m left pondering how we could anticipate a similar event. The nature of this type of initial failure seems hard to determine in advance, given its seismic origin, and the behaviour of the flow is also very difficult to forecast. Thus, such events represent a massive challenge in managing landslide risk.

References

Faral, A., Lavigne, F., Sastrawan, W.J. et al. 2024. Deadliest natural disaster in Balinese history in November 1815 revealed by Western and Indonesian written sources. Natural Hazards 120, 12011–12041 (2024). https://doi.org/10.1007/s11069-024-06671-5.

Faral, A. et al. 2025. Field evidence of the greatest disaster in Balinese history: The 1815 Geger Bali multi-hazard event in Buleleng. Geomorphology, https://doi.org/10.1016/j.geomorph.2025.109903.

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.

AbstractSeismic inversion translates seismic data into subsurface elastic property models, enabling geophysicists to better understand underground rocks and fluids. Due to the inherently ill-posed nature of this inverse problem, accurately capturing the uncertainty associated with the solution is essential for reliable interpretations. Traditional Bayesian inversion methods, such as Markov Chain Monte Carlo (MCMC) and Laplace approximations, have been employed for this purpose but face significant limitations in terms of scalability and computational efficiency for large-scale problems. Combined with deep learning, Variational Inference (VI) has emerged as a promising alternative, striking a balance between computational efficiency and flexibility (i.e. the ability to approximate complex posterior distributions). However, selecting an appropriate proposal distribution remains a key challenge, as it directly influences the quality of the estimated posterior distribution. In this study, we extend IntraSeismic, an implicit neural representation (INR)-based framework for seismic inversion applications, to Bayesian inversion using VI with different parameterizations of the proposal distribution. We introduce two methods: B-IntraSeismic (BIS), which uses a mean-field Gaussian proposal, and B-IntraSeismic with Conditional Normalizing Flows (BIS-Flow), which utilizes a mean-field unparameterized proposal distribution to better capture deviations from Gaussianity in the posterior distribution. These methods are evaluated on a synthetic dataset (Marmousi) and two field data (Volve and Sleipner). Our results indicate that both BIS and BIS-Flow can accurately capture structural details and produce high-resolution mean models and standard deviation maps. BIS-Flow is also shown to be able to model complex posterior distributions, offering a more comprehensive characterization of uncertainty while maintaining computational feasibility.

AbstractThe phenomenon of elastic wave conversions, where acoustic, pressure (P-) waves are converted to elastic, shear (S-) waves, and vice-versa, is commonly disregarded in seismic imaging. This can lead to lower-quality images in regions with strong contrasts in elastic parameters. While a number of methods exist that do take wave conversions into account, they either deal with P- and S-waves separately, or are prohibitively computationally expensive, as is the case for elastic Full-Waveform Inversion. In this paper an alternative approach to taking converted waves into account is presented by extending Full Wavefield Migration (FWM) to account for wave conversions. FWM is a full-wavefield inversion method based on explicit, convolutional, one-way propagation and reflection operators in the space-frequency domain. By applying these operators recursively, multi-scattering data can be modelled. Using these operators, the FWM algorithm aims to reconstruct the reflection properties of the subsurface (i.e. the ‘image’). In this paper, the FWM method is extended by accounting for wave conversions due to angle-dependent reflections and transmissions using an extended version of Shuey’s approximation. The resulting algorithm is tested on two synthetic models to give a proof of concept. The results of these tests show that the proposed extension can model wave conversions accurately and yields better inversion results than applying conventional, acoustic FWM.

A new study delivers a stark warning that Central Asia has overshot its environmental safety limits concerning land footprint and biosphere integrity. The study, led by Prof. Duan Weili from the Xinjiang Institute of Ecology and Geography of the Chinese Academy of Sciences, provides a comprehensive sustainability assessment and identifies Kazakhstan and Uzbekistan as priority areas for environmental management.

In their paper published in Science of the Total Environment, researchers from IIASA and Lviv Polytechnic National University in Ukraine presented a novel approach to measure and understand human pressure on planet Earth. The researchers explored how carbon emissions can be translated into measures of "stress" and "strain" to derive new insights into how the planet is changing.

A powerful heat wave has been gripping large parts of southern Europe and North Africa, pushing air temperatures beyond seasonal norms and triggering widespread health and wildfire alerts. Among the hardest-hit countries are Spain, France, Italy, Greece, Cyprus, and Algeria.

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.

Today, President Trump is visiting a new immigration detention facility built on a disused airstrip in the Florida Everglades. On 27 June, environmental groups sued the U.S. Department of Homeland Security (DHS), U.S Immigration and Customs Enforcement (ICE), the Florida Division of Emergency Management, and Miami-Dade County, seeking a temporary restraining order to stop the construction and opening of the facility.

The lawsuit from Friends of the Everglades and the Center for Biological Diversity argues that the facility’s construction did not undergo environmental reviews legally required under the National Environmental Policy Act. The groups assert that constructing the facility, transporting and housing thousands of people on site, and then flying them directly from the facility to other locations, will undermine decades of work spent restoring and protecting the Everglades’ delicate ecosystem.

“The site is more than 96% wetlands, surrounded by Big Cypress National Preserve, and is habitat for the endangered Florida panther and other iconic species. This scheme is not only cruel, it threatens the Everglades ecosystem that state and federal taxpayers have spent billions to protect,” Eve Samples, executive director of Friends of the Everglades, said in a statement.

“The Miccosukee Tribe is opposed to the use of our ancestral lands in Big Cypress as a detention facility.”

A spokesperson for Florida Governor Ron DeSantis said that the facility “will have no impact on the surrounding environment” and that they will oppose the lawsuit in court.

DeSantis and other state officials have claimed emergency powers to commandeer Dade-Collier Training and Transition Airport and build the migrant facility in roughly a week. Given the nickname “Alligator Alcatraz,” the detention facility is made of tents, trailers, and other temporary buildings and is designed to hold up to 5,000 people detained by DHS and ICE.

Immigration and human rights activists have raised additional concerns about housing thousands of people in tents and trailers at the height of a hot and humid Florida summer and during what is likely to be an above-normal hurricane season. Others are concerned about the environmental impact of a crowded detention center near an aquifer that supplies drinking water to the surrounding area.

 
Related

•  The lawsuit
•  Environmental groups sue to block opening of “Alligator Alcatraz” migrant detention center in Florida Everglades
•  Native leaders blast construction of Florida’s ‘Alligator Alcatraz’ on land they call sacred
•  Get Involved: AGU Science Policy Action Center
 

Indigenous tribes also vehemently oppose the construction of the facility on the land, which is sacred to the Miccosukee Tribe of Indians of Florida and the Seminole Tribe of Florida. There are 19 traditional Miccosukee and Seminole villages in Big Cypress, as well as ceremonial and burial grounds and other gathering sites.

Talbert Cypress, Chairman of the Miccosukee Tribe of Indians of Florida, stated, “Rather than Miccosukee homelands being an uninhabited wasteland for alligators and pythons, as some have suggested, the Big Cypress is the Tribe’s traditional homelands….The Miccosukee Tribe is opposed to the use of our ancestral lands in Big Cypress as a detention facility.”

Groups of environmental, Indigenous, immigration, and human rights activists protested outside the facility on 28 June. More protests are expected today as the facility opens and the president visits.

—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.

New research has, for the first time, tracked ice shelf, sea ice and ocean swell wave conditions over multiple years in the lead-up to three large-scale iceberg "calving" events in Antarctica, revealing common patterns.

A patch of the Atlantic Ocean just south of Greenland is cooling while much of the world warms. The origin of this "cold blob" has been linked to weakening ocean currents that help regulate global climate—called the Atlantic Meridional Overturning Circulation (AMOC). A team of scientists led by Penn State has found a weakening AMOC impacts not just the ocean but also the atmosphere, and that these two factors may contribute equally to the cold anomaly.

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

Clouds are a major source of uncertainty in atmospheric predictability and simulating them accurately remains a challenge for large-scale models. Bogenschutz et al. [2025] evaluate a new high-resolution model called the Simple Cloud-Resolving E3SM Atmosphere Model (SCREAM) developed by the United States Department of Energy (DOE), which is designed to better capture cloud and storm processes. The authors use a fast, small-scale version of the model and compare its output to modern real-world observations from the DOE’s Atmospheric Radiation Measurement (ARM) program.

The model performed better at higher resolutions but still struggled with certain cloud types, especially mid-level “congestus” clouds that form between shallow and deep convection. SCREAM also tended to shift too abruptly from shallow clouds to intense storms, and its performance depended on how finely the vertical layers of the atmosphere were represented.

These results help pinpoint key weaknesses in the model’s treatment of clouds and turbulence. The new library of ARM cases added in this work will help guide future improvements to SCREAM and support more accurate simulations of cloud processes.

Citation: Bogenschutz, P. A., Zhang, Y., Zheng, X., Tian, Y., Zhang, M., Lin, L., et al. (2025). Exposing process-level biases in a global cloud permitting model with ARM observations. Journal of Geophysical Research: Atmospheres, 130, e2024JD043059. https://doi.org/10.1029/2024JD043059

—Yun Qian, Editor, JGR: Atmospheres

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.

On 16 September 2023, a low hum started swaying seismometers around the world. Unlike during the short and jagged frenzy of an earthquake, this signal wobbled every 92 seconds and continued for 9 days. About a month later, while seismologists were still puzzling over the incident, the hum started again and lasted roughly a week.

Researchers traced the confusing signals to East Greenland, where satellite imagery revealed the scars of recent rockslides in Dickson Fjord. They deduced that millions of cubic meters of rock and ice had suddenly fallen into the sea on 16 September, creating a 200-meter (650-foot) tsunami and a long-lasting wave called a seiche. Rather than ricochet out to sea, crooked topography kept the tsunami sloshing back and forth between the fjord’s parallel walls. The later hum was from a second, smaller rockslide and seiche.

The area is unpopulated, meaning no one was threatened by the initial wave but no one observed the event either.

Seiches typically need a continuous energy source such as a windstorm to persist, but the long-lasting waves in Dickson Fjord appeared to be self-sustaining. Two teams independently developed simulations showing Dickson Fjord could support a long-lasting seiche. A new study in Nature Communications builds on that work, using satellite data to provide the first direct observations of the seiche.

“To really robustly be able to say, ‘This is what was shaking the Earth at this time,’ we needed that observational evidence,” said Thomas Monahan, an oceanographer at the University of Oxford and first author of the new paper.

Before (left) and after images show the obvious collapse of a glacier in Greenland’s Dickson Fjord. Credit: Søren Rysgaard As Above, So Below

East Greenland is remote, and the seiche mostly dissipated before the Danish military arrived 3 days after the initial wave to investigate the collapsed mountain face in Dickson Fjord. By then, the amplitude of the wave was already too small to detect from the boat. However, the shift in sea surface was visible from space thanks to the international Surface Water and Ocean Topography (SWOT) satellite, launched in 2022.

“We’ve never had the capability to do things in these regions at this level before.”

SWOT uses two altimeters spaced 10 meters apart to triangulate small changes in water height. Prior to SWOT, satellites had one altimeter and could offer a one-dimensional footprint of the ocean. Now, Monahan said, researchers can obtain precise, high-resolution imagery of the sea surface, even between the deep walls of a distant fjord.

“We’ve never had the capability to do things in these regions at this level before,” he said.

The satellite passed over Dickson Fjord several times during the main event and the smaller rockslide that followed. Monahan and his colleagues examined SWOT data from four transits, tracking the sea surface slope along the same transect each time.

The water was sloshing back and forth between the fjord walls.

The researchers extended their search to rule out other causes. The timing of the waves did not match the timing of winds recorded by a weather station in the fjord or the pattern of tides recorded by SWOT over the next 13 months. The magnitude of the wave did, however, match the seismic signal, further suggesting the fjord’s geometry had trapped a wave.

A sloshing tsunami in Dickson Fjord shimmied seismometers for 9 days starting on 16 September 2023. This data visualization of the fjord on 17 September 2023 shows the sloshing water and adds direct observational evidence to earlier models. Credit: NASA Earth Observatory “Science at Its Best”

The study further confirmed the seiche but also showed the early utility of SWOT, which had finished calibrating just 2 months before the initial rockslide.

“They’re sort of perfect partners, satellite and seismic data.”

“It’s a nice surprise to see the result,” said Yao Yu, a physical oceanographer who works with SWOT data at the Scripps Institution of Oceanography. The satellite is built for oceans, rivers, and lakes, she said, but the new study shows it can also collect good data from high-latitude fjords in areas unreachable by prior satellites. “A lot of things we never expected SWOT can do, it’s actually working very well,” she said.

SWOT’s spatial resolution is especially important in the Arctic, where seismometers are sparse. The satellite provides only intermittent observations, but it can access remote locations. That fills a gap, said Stephen Hicks, a seismologist at University College London and coauthor on one of the original seiche papers.

“They’re sort of perfect partners, satellite and seismic data,” he said. The new study backs up and builds upon the original research, he added, and “that’s sort of science at its best.”

—J. Besl (@J_Besl), Science Writer

Citation: Besl, J. (2025), New satellite adds evidence of an Earth-shaking wave, Eos, 106, https://doi.org/10.1029/2025EO250236. Published on 1 July 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.

Eastern Australia is one of the most fire-prone regions in the world, with bushfires responsible for the deaths of about 800 people and millions of animals since about 1850.

A major slope failure killed many people, possibly over 300, in an area of unlicenced mining of the mineral Coltan.

On 19 June 2025, a very significant landslide occurred at the Rubaya mining site in Masisi territory, North Kivu, which is located in the eastern part of the Democratic Republic of the Congo (DRC). The landslide, which reportedly affected a place called Bibatama, killed at least 21 people, but in all probability many more people died. Local news site Mines.cd reports over 300 fatalities.

The Rubaya mining area is a large, unlicenced and unregulated shallow excavation for the extraction of Coltan (known industrially as tantalite), an ore from which niobium and tantalum are extracted. The primary use of tantalum is in mobile phones, but it is also used in computer hard drives and road vehicle electronics.

These types of disastrous mining landslides in less developed countries rarely attract much interest (imagine what would have happened if this event had occurred in Canada or Australia), so I decided to see whether I could find anything out about it. I must note that landslides at this site are common – for example, about 100 people were killed in a landslide in 2013.

The Rubaya mining area is well covered in Google Earth – this is an image from 2021. The marker gives the general location – we’ll come back to this spot below:-

Google Earth image from 2021 showing the Rubaya mining area in the DRC.

Zoom in and you find a landscape scarred by shallow workings and landslides:-

Google Earth image from 2021 showing a part of the Rubaya mining area in the DRC.

The Rubaya mining area has a very challenging history. In recent years, possession has alternated between the military and various militias, who have run the site as a protection racket. Since April 2024, the site has been controlled by the March 23 Movement (M23), a rebel group with a long history of human rights violations.

I have been trying to use Planet Labs images to try to identify the location of the 19 June 2025 landslide. I think the most likely location is in the mining area located at [-1.58203, 28.89378]:-

Google Earth image from 2021 showing the likely location of the 19 June 2025 landslide in the Rubaya mining area in the DRC.

This mining area has expanded rapidly in recent years. The 2021 Google Earth image shows that it has been subject to a number of landslides.

I have downloaded a Planet Labs image from 14 June 2025 – five days before the landslide, and I have draped onto the Google Earth DEM. Of course, the Planet Labs imagery has a lower spatial resolution than the Google Earth imagery:-

Planet Labs image of the likely site of the 19 June 2025 landslide in the Rubaya mining area. Image copyright Planet Labs, used with permission. Image dated 14 June 2025.

The image shows a higher level of mining activity than was the case in 2021, and possibly some further landslides. By comparison, the image below was captured on 25 June 2025, after the landslide:-

Planet Labs image of the aftermath of the 19 June 2025 landslide in the Rubaya mining area. Image copyright Planet Labs, used with permission. Image dated 25 June 2025.

And here is a slider to allow the images to be compared:-

Image copyright Planet Labs, using the Google Earth DEM.

I think the landslide is visible on the left side of the mining area. A series of shallow workings have been destroyed, and the track and runout zone of the landslidecan be seen. The feature that is probably the landslide is about 250 metres long.

These types of landslides in unlicenced and unregulated mining sites are a major contributor to global landslide fatalities, but they are rarely investigated.

Finally, in an interesting twist, the FT reported last week that an ally of Donald Trump, Gentry Beach, is seeking to “snap up” the Rubaya mine site.

Reference

Planet Team 2024. 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.

SummaryGlobal seismology mainly uses seismic waves propagating in the sagittal plane along the great circle path (GCP). However, heterogeneities in the mantle laterally deviate the path of seismic signals, which arrive out-of-plane (OOP) at arrays of sensors at teleseismic distances. Detection and back-projection of these signals have, in the past, provided independent evidence for the location of distant subducted slabs in the deep mantle, complementing global tomographic imaging. To infer physical properties of these subducted slabs, 3D waveform modeling of OOP waves for a finite-thickness slab is needed but still missing. In this study, we conduct a series of synthetic tests using a spectral element solver. We test the detectability of OOP signals and, by progressively adding complexities, we evaluate to which extent these signals can be used to infer physical properties of the modeled slab. We carry out three-component array analysis and investigate focal mechanism dependency. Our results show that the transverse component might be the best candidate for such studies, also for P-to-P OOP signals. Vertical and radial component recordings are usually dominated by P-SV energy arriving from the earthquakes along the GCP, which masks possible OOP signals. Contrary, the transverse component filters out any P-SV energy arriving directly from the source and, owing to its intrinsic directionality, allow for higher resolution measurement of P-to-P OOP signals. This is especially the case prior to the arrival of the S-wavefield. We pick a series of OOP arrivals which are back-projected using a multi-phase trial-and-error approach, that is considered successful only when different OOP seismic phases converge to the modeled (true) structure. We retrieve the location of the slab, its bottom and top edges, and its thickness in the lower mantle. These inferences are tested against varying topography, orientation and size of the modeled slab. The insights gained with modeling are confirmed with real data examples, supporting higher resolution mapping of 3D mantle structure based on OOP seismology.

SummaryWe investigate the lithospheric structure of the Southwest Iberian margin along an active seismic profile southwest of São Vicente Cape, ranging from the southern Tagus Abyssal Plain to the westernmost part of the Gulf of Cadiz. This profile, approximately 320 km long, intersects almost perpendicularly three major thrust faults: the Tagus Abyssal Plain, Marquês de Pombal and Horseshoe faults. The crustal structure, derived from spatially coincident wide-angle seismic (WAS) and multichannel seismic (MCS) data, was validated and constrained using gravimetric data. Joint travel-time inversion of refracted phases identified in WAS and reflected seismic phases from both WAS and MCS records were used to build a detailed two-dimensional P-wave velocity (Vp) structure. The resulting model reveals a Vp distribution with abrupt lateral velocity and structural variations, characterized by a rugged basement top and sharp changes in crustal thickness. Three main lithospheric domains consisting of continental, oceanic, and exhumed mantle affinity were identified from south to north. The travel-time inversion of the deepest reflected seismic phases reveals four major southeast-dipping reflectors, likely corresponding to major regional thrust faults with significant seismic and tsunamigenic potential. Integrating the modelled and interpreted seismic results with the locations of recent well-constrained earthquakes suggests that the Marquês de Pombal and Tagus Abyssal Plain extend deeper than previously thought, with fairly high seismic activity in the deep levels. This has significant implications for their seismogenic potential and should be taken into account for accurate assessment of seismic hazards in the region.

For years, sand nourishment has been an important way to protect the Dutch coast against erosion and rising sea levels. But we know surprisingly little about one type of nourishment, shoreface nourishment. A recent scientific review article published in Earth-Science Reviews by the University of Twente maps out existing knowledge in this area and underlines the need for further research.

Scientists have discovered that changes in climate and water levels are reducing the ability of some ecosystems in the Everglades to sequester carbon, while the environmental shifts are enhancing the potential for carbon uptake by scrub mangroves.

As glaciers retreat due to a rise in global temperatures, one study shows that detailed 3D elevation models could drastically improve predictions about how they react to Earth's warming climate.

Researchers have discovered a dramatic and unexpected shift in the Southern Ocean, with surface water salinity rising and sea ice in steep decline.

They're in the headlines every week—critical minerals like lithium, cobalt, nickel, graphite and the rare earth elements essential for high-technology and national security applications.