Feed aggregator

By 2050, offshore wind power capacity in the North Sea is set to increase more than tenfold. Researchers at the Helmholtz Center Hereon have analyzed the long-term overall impact of this large number of wind farms on the hydrodynamics of the North Sea for the first time. They found that the current pattern could change on a large scale. The study highlights approaches for minimizing potential risks to the environment at an early stage. The work was recently published in the journal Communications Earth & Environment.

Widespread 'enhanced rock weathering' could slow global warming

Phys.org: Earth science - Mon, 02/16/2026 - 16:40

It's one of the latest technologies for sequestering carbon: crush silicate rocks, add to crop soil, and let the rock dust naturally react with carbon dioxide. The reactions bind carbon into stable mineral forms that can persist for millennia, while also enriching the soil with nutrients, boosting crop yields and increasing farmer profits.

Amazon deforestation raises surface temperature by 3°C during dry season, satellite data show

Phys.org: Earth science - Mon, 02/16/2026 - 16:20

Deforestation in the Amazon is causing significant regional changes in climate compared to areas with forest cover above 80%. The loss of vegetation leads to an increase in surface temperature, a decrease in evapotranspiration, and a reduction in precipitation during the dry season and in the number of rainy days.

Antarctica sits above Earth's strongest 'gravity hole.' Now we know how it got that way

Phys.org: Earth science - Mon, 02/16/2026 - 15:40

Gravity feels reliable—stable and consistent enough to count on. But reality is far stranger than our intuition. In truth, the strength of gravity varies over Earth's surface. And it is weakest beneath the frozen continent of Antarctica after accounting for Earth's rotation.

Rocky Shore Erosion Shaped by Multi-Scale Tectonics

EOS - Mon, 02/16/2026 - 14:00

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

Coastal landscapes evolve under the combined influence of wave action, climatic variations, sea‑level change, and tectonic processes. Shoreline evolution is especially important along rocky coasts such as those of the western United States, where it shapes hazards to people and infrastructure and affects exposure to events like tsunamis. In this context, tectonically driven uplift plays a key role over both individual earthquake cycles and longer timescales associated with fault-system and topographic development.

Using a compilation of coastal change metrics and statistical analyses, Lopez and Masteller [2026] identify a tentative link between tectonics and shoreline change. On decadal timescales, uplift can slow coastline retreat, as might be expected. Over many earthquake cycles, however, higher long-term uplift associated with cumulative subduction-zone deformation appears to enhance shoreline retreat. These findings highlight some of the interactions between coastal and solid earth hazards. They also point toward future models that integrate similar constraints to improve our understanding of how earthquakes build topography and how sea level, coastal processes, and tectonics together modulate short‑ and long‑term coastal risk.

Citation: Lopez, C. G., & Masteller, C. C. (2026). Tectonics as a regulator of shoreline retreat and rocky coast evolution across timescales. AGU Advances, 7, e2025AV002065. https://doi.org/10.1029/2025AV002065

—Thorsten Becker, Editor, AGU Advances

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

The 16 June 2024 landslide cluster in Wuping County, Fujian Province, China

EOS - Mon, 02/16/2026 - 08:10

Six people were killed when intense rainfall triggered over 6,500 landslides

On 16 June 2024, an extreme rainfall event triggered a dense cluster of landslides and channelised debris flows in Wuping County, Fujian Province, China. This is one of many such events in recent years – anecdotally at least, these events are becoming more common and more severe.

Thus, I very much welcome a paper in the journal Landslides (Liao et al. 2026) that describes this event. The paper is not open access, but this link should allow you to read the full manuscript. The authors highlight the impact of the event – six people were killed (two of whom were never recovered) and hundreds of houses were damaged.

The cluster of landslides centres on the area around [24.94745, 116.29172]. This Planet Labs image, captured on 27 November 2024 after the event, shows some of the landslides triggered:-

Landslides triggered by the 16 June 2024 rainfall event in Wuping County, Fujian Province. Image copyright Planet Labs, used with permission, collected on 27 November 2024.

Note the presence of multiple shallow landslides that have combined to form channelised debris flows. In the centre of the image, by the marker, there is a small reservoir that has been almost entirely infilled by debris from the landslides.

In total, Liao et al. (2026) have mapped 6,526 landslide triggered by the rainfall event. The main initiating rainfall appears to have been a period between 14:00 and 18:00 on 16 June 2024, during which 161 mm was recorded, with a peak intensity of 55 mm per hour. Interestingly, though, the landslide density correlates with rainfall total prior to the main initiating event, rather than to the total rainfall. I wonder whether this indicates that the key parameter (the distribution of peak rainfall intensity, for example) is not being captured in the data?

Very helpfully, Liao et al. (2026) have investigated the mechanism of the landslides in some detail. They find that behaviour differed according to the bedrock lithology. In areas underlain by granite, failure occurred on the interface between the weathered and the unweathered materials, a common situation. In most cases, granitic landslides did not generate channelised debris flows.

On the other hand, in areas underlain by greywacke, failures also occurred in these interface areas, but channelised debris flows were more common. This may be related to the steeper local topography in the greywacke areas.

The paper by Liao et al. (2026) further helps us to understand these clusters of landslides and channelised debris flows, which are proving to be so very destructive. Expect more of these events in the coming months and beyond.

Reference and acknowledgement

Liao, Z., Wu, J., Ma, J. et al. 2026. Characteristics and initiation mechanism of clustered landslides triggered by an extreme rainfall in Wuping County, Fujian Province, China. Landslides. https://doi.org/10.1007/s10346-026-02712-1.

Many thanks to the wonderful people at Planet Labs for providing access to the satellite imagery.

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

Accelerating Bayesian Full Waveform Inversion using Reconstruction-Guided Diffusion Sampling

Geophysical Journal International - Mon, 02/16/2026 - 00:00

SummaryFull waveform inversion (FWI) is a powerful tool in seismic imaging, capable of producing high-resolution models of the subsurface. However, the method remains computationally intensive and sensitive to initial models due to its nonlinearity and ill-posed nature. To quantify uncertainty in FWI results, variational inference (VI) methods, such as Stein Variational Gradient Descent (SVGD), have been increasingly explored. These approaches approximate the posterior distribution by evolving a set of particles using gradient information from the log-posterior. Despite their promise, their effectiveness heavily depends on the quality of the prior used for initialization. In this work, we propose a hybrid framework that improves the efficiency and robustness of VI-based FWI by initializing SVGD with samples drawn from a reconstruction-guided diffusion model. Rather than replacing SVGD with a generative sampler, our approach preserves the theoretical foundations of VI while leveraging the expressive capacity of deep generative models. The diffusion model is trained to generate geologically plausible models conditioned on seismic images, thereby guiding the SVGD initialization toward regions of high posterior support. This initialization significantly reduces the number of required SVGD updates and improves convergence, while keeping the core VI formulation intact. Our results show enhanced posterior approximation and more geologically consistent solutions, with an order of magnitude lower computational cost compared to naïvely initialized SVGD. However, challenges remain, such as the computational demands of likelihood evaluations, the formation of a training set that encompasses all plausible realizations, and sensitivity to reconstruction-guidance weights during sampling. Overall, this method provides a principled and efficient approach to uncertainty-aware FWI, integrating physics-informed inference with data-driven generative modeling for practical applications in full waveform inversion.