Feed aggregator

Complex-valued neural networks for spectral induced polarization applications

Geophysical Journal International - Wed, 09/03/2025 - 00:00
SummarySpectral induced polarization (SIP) aims to characterize geological materials by measuring the dispersion of their complex conductivity in the frequency domain. Despite the complex-valued nature of SIP data, most machine learning models used for its analysis rely on real-valued representations that discard phase information and may limit performance. This study investigates the benefits of complex-valued neural networks (CVNN) for SIP applications by comparing their performance against real-valued neural networks (RVNN) across three tasks: mineral classification, Cole-Cole parameter estimation, and mechanistic modelling of ionic and electric potential perturbations around polarizable minerals. To ensure fair comparisons and emphasize the effect of complex-valued representations, we design CVNN and RVNN models with matched capacity, aspect ratio, and training duration. Our numerical experiments show that CVNNs consistently outperform RVNNs in the classification task, achieving lower validation loss and up to 5 percent higher classification metrics (p-value = 2.9 × 10−7). We test the Cole-Cole inversion networks on laboratory SIP measurements and validate the parameter estimation accuracy using synthetic data. Test results indicate that CVNNs produce curve fits that are ≈4 % more accurate for the imaginary part of resistivity (p-value = 3.1 × 10−4), and validation results show accuracy improvements of up to 2 percent for chargeability, relaxation time, and the Cole-Cole exponent (p-value = 1.7 × 10−7). CVNNs also yield more accurate approximations of mechanistic model variables, with error reductions of up to 1 percent for ionic concentrations (p-value = 1.6 × 10−4). Our experiments suggest that CVNNs provide modest but statistically significant benefits in SIP applications involving laboratory or synthetic data. While RVNNs may eventually reach comparable predictive accuracy if trained longer, we observe that CVNNs converge more rapidly under matched training conditions. This study provides a reproducible framework for benchmarking neural network architectures in SIP and supports the integration of CVNNs into geophysical workflows where phase responses encode physically meaningful information.

New dataset enhances understanding of atmosphere–surface interactions

Phys.org: Earth science - Tue, 09/02/2025 - 20:30
A research team led by Prof. Chen Cheng from the Hefei Institutes of Physical Science of the Chinese Academy of Sciences has retrieved global aerosol and surface properties using advanced polarization data from China's GF-5(02) satellite.

Once king of the seas, a giant iceberg is finally breaking up

Phys.org: Earth science - Tue, 09/02/2025 - 20:16
Nearly 40 years after breaking off Antarctica, a colossal iceberg ranked among the oldest and largest ever recorded is finally crumbling apart in warmer waters, and could disappear within weeks.

Deforestation reduces rainfall by 74% and increases temperatures by 16% in Amazon during dry season, study says

Phys.org: Earth science - Tue, 09/02/2025 - 19:22
Deforestation in the Brazilian Amazon is responsible for approximately 74.5% of the reduction in rainfall and 16.5% of the temperature increase in the biome during the dry season. For the first time, researchers have quantified the impact of vegetation loss and global climate change on the forest.

How Amazon trees use recent rainfall in the dry season and support the production of their own rain

Phys.org: Earth science - Tue, 09/02/2025 - 19:20
The Amazon is the world's largest tropical forest, home to unmatched biodiversity and one of the planet's longest rivers. Besides the Amazon River, the Amazon rainforest also features "flying rivers:" invisible streams of vapor that travel through the atmosphere, fueling rainfall both within the forest and far beyond its boundaries.

Cities face double trouble: Extreme heat and air pollution mean increasing compound weather events

Phys.org: Earth science - Tue, 09/02/2025 - 18:40
U.S. cities are facing a growing threat that goes beyond hot weather or hazy air. New research from the University of Oklahoma reveals that "compound events"—periods when heat wave conditions coincide with high air pollution levels—are becoming more frequent and intense in urban areas across the United States.

Scientists tune in to the surf's hidden signals for potential mapping data

Phys.org: Earth science - Tue, 09/02/2025 - 18:40
Along the coast, waves break with a familiar sound. The gentle swash of the surf on the seashore can lull us to sleep, while the pounding of storm surge warns us to seek shelter.

The ocean carbon sink is ailing: 10% drop in CO₂ absorption seen during record 2023 marine heat wave

Phys.org: Earth science - Tue, 09/02/2025 - 17:16
Measurements analyzed by an international research team led by ETH Zurich show that the global ocean absorbed significantly less CO₂ than anticipated during the unprecedented marine heat wave in 2023.

Scientists identify unique chemical regime for secondary organic aerosol formation in urban China

Phys.org: Earth science - Tue, 09/02/2025 - 16:24
Air pollution from secondary organic aerosols (SOA) has now become a greater problem in Chinese cities since 2013 because regulations have successfully reduced fine particulate matter (PM2.5) emitted directly from, for example, vehicles and industries, according to a study led by Prof. Huang Rujin at the Institute of Earth Environment of the Chinese Academy of Sciences and Prof. Marianne Glasius at Aarhus University.

Why was the Afghanistan earthquake so deadly? A disaster resilience expert explains

Phys.org: Earth science - Tue, 09/02/2025 - 13:43
The death toll following the recent earthquake in Afghanistan continues to rise. Taliban-led health authorities now say at least 800 people have been killed and 2,000 injured.

Inside a Georgia Beach’s High-Tech Fight Against Erosion

EOS - Tue, 09/02/2025 - 13:09

This story was originally published by Grist. Sign up for Grist’s weekly newsletter here. This coverage is made possible through a partnership between Grist and WABE, Atlanta’s NPR station.

At low tide on Tybee Island, Georgia, the beach stretches out as wide as it gets with the small waves breaking far away across the sand—you’ll have a long walk if you want to take a dip. But these conditions are perfect for a team of researchers from the University of Georgia’s Skidaway Institute of Oceanography.

Every three months, at low tide, they set out a miniature helipad near the foot of the dune and send up their drone equipped with lidar—technology that points a laser down at the sand and uses it to measure the elevation of the beach and dunes. The team flies it back and forth from the breakers to the far side of the dune and back until they have a complete, detailed map of the island’s 7-mile beach, about 400 acres.

“I see every flip-flop on the beach.”

“It’s high accuracy, it’s a high resolution,” explained research technician Claudia Venherm, who leads this project. “I see every flip-flop on the beach.”

That detailed information is crucial because Tybee is a barrier island, and rising seas are constantly eating away at the sandy beach and dunes that protect the island’s homes and businesses as well as a stretch of the Georgia mainland. Knowing exactly where the island is eroding and how the dunes are holding up to constant battering can help local leaders protect this piece of coastline.

“Tybee wants to retain its beach. It also wants to maintain, obviously, its dune. It’s a protection for them,” said Venherm. “We also give some of our data to the Corps of Engineers so they know what’s going on and when they have to renourish the beach.”

Since the 1970s the Army Corps of Engineers has helped maintain Tybee Island’s beaches with regular renourishment: Every seven years or so, the Corps dredges up sand from the ocean floor and deposits on the beach to replace sand that’s washed away. The data from the Skidaway team will only help the Corps do this work more effectively. Lidar isn’t new, and neither is aerial coastal mapping. Several federal agencies monitor coastlines with lidar, but those surveys are more typically several years apart for any one location, rather than a few months.

The last renourishment finished in January 2020, and Venherm and her team got to work a few months later. That means they have five years of high-resolution beach data, recorded every three months and after major storms like Hurricane Helene, creating a precise picture of how the beach is changing.

“I can compute what the elevation of the dune is, as well as how much volume has been lost or gained since a previous survey.”

“I can compute what the elevation of the dune is, as well as how much volume has been lost or gained since a previous survey,” said Venherm. “I can also compute how long it will take until the beach is completely gone, or how long will it take until water reaches the dune system.”

The Corps conducts regular renourishment projects on beaches all along the East Coast, and uses a template to inform that work, said Alan Robertson, a consultant who leads the city of Tybee’s resilience planning. But he hopes that such granular evidence of specific changes over time can shift where exactly the sand gets placed within the bounds of that template. An area near the island’s north end, for instance, is a clear hot spot for erosion, so the city may push for concentrating sand there, and north of that point so that it can travel south to fill in the erosion.

“We know exactly where the hotspots of erosion are. We know where there’s accretion,” he said, referring to areas where sand tends to build up. “[We] never had that before.”

The data can also inform the city’s own decision-making, because it provides a much clearer picture of what happens to the dunes and beach over time after the fresh sand is added. In the past, they’ve been able to see the most obvious erosion, but now they can compare how different methods of dune-building and even sources of sand hold up. The vegetation that’s critical to holding dunes together, for instance, takes root far better in sand dredged from the ocean compared to sand trucked in from the mainland, Robertson said.

“There’s an example of the research and the monitoring. I actually can make that statement,” he said. “I actually know where you should get your sand from if you can, and why. No one could have told you that eight years ago.”

That sort of proven information is key in resilience projects, which are often expensive and funded by grants from agencies that want confirmation their money is being spent well.

“Everything we do now on resiliency, measuring, and monitoring has become a priority,” said Robertson. “We’ve been able over these years through proof statements of ‘look at what this does for you’ to make it part of the project.”

—Emily Jones (@ejreports.bsky.social), Grist

This article originally appeared in Grist at https://grist.org/science/inside-a-georgia-beachs-high-tech-fight-against-erosion/.

Grist is a nonprofit, independent media organization dedicated to telling stories of climate solutions and a just future. Learn more at Grist.org

Subharmonic Shapiro steps in depinning dynamics of a two-dimensional solid dusty plasma modulated by one-dimensional nonlinear deformed periodic substrates

Physical Review E (Plasma physics) - Tue, 09/02/2025 - 10:00

Author(s): Zhaoye Wang, Nichen Yu, Ao Xu, Chen Liang, C. Reichhardt, C. J. O. Reichhardt, and Yan Feng

Langevin dynamical simulations are performed to investigate the depinning dynamics of a two-dimensional (2D) solid dusty plasma, which is modulated by one-dimensional (1D) nonlinear deformed periodic substrates, and also driven by the combination of the DC and AC forces. As the DC driving force incr…


[Phys. Rev. E 112, 035201] Published Tue Sep 02, 2025

The tipping of the last resilient glaciers: Filling in years of missing data from Tajikistan

Phys.org: Earth science - Tue, 09/02/2025 - 09:00
Too little snowfall is now also shaking the foundations of some of the world's most resilient "water towers," a new study led by the Pellicciotti group at the Institute of Science and Technology Austria (ISTA) shows. After establishing a monitoring network on a new benchmark glacier in central Tajikistan, the international team of researchers was able to model the entire catchment's behavior from 1999 to 2023.

For the first time in 40 Years, Panama's deep and cold ocean waters fail to emerge

Phys.org: Earth science - Mon, 09/01/2025 - 19:50
The natural phenomenon of upwelling, which occurs annually in the Gulf of Panama, failed for the first time on record in 2025. A study led by scientists from the Smithsonian Tropical Research Institute (STRI) indicates that the weakening of the trade winds was the cause of this event. This finding highlights the climate's impact on fundamental oceanic processes and the coastal communities that depend on them.

Corrigendum to “Spatio-temporal characteristics of ionospheric irregularities in low latitude regions during the peak of solar cycle 25” [Adv. Space Res. 76(1) (2025) 254–268]

Publication date: 1 September 2025

Source: Advances in Space Research, Volume 76, Issue 5

Author(s): Napat Tongkasem, Pornchai Supnithi, Phimmasone Thammavongsy, Michi Nishioka, Septi Perwitasari, Susumu Saito, Jeff Klenzing, Lin Min Min Myint

Petrography of scoriaceous and unmelted micrometeorites from the Maitri Station, Antarctica collection

Publication date: 1 September 2025

Source: Advances in Space Research, Volume 76, Issue 5

Author(s): D. Fernandes, N.G. Rudraswami, V.P. Singh

Expanding opportunities for landing to Venus: using gravity assist and resonant orbits for orbiter-lander missions

Publication date: 1 September 2025

Source: Advances in Space Research, Volume 76, Issue 5

Author(s): Vladislav A. Zubko, Natan A. Eismont, Konstantin S. Fedyaev

Precise orbit determination of Spire CubeSats constellation using the raw observation approach

Publication date: 1 September 2025

Source: Advances in Space Research, Volume 76, Issue 5

Author(s): Parisa Shafiei, Jean Bemtgen, Sajad Tabibi

Mission analysis of space-based small camera for space debris detection

Publication date: 1 September 2025

Source: Advances in Space Research, Volume 76, Issue 5

Author(s): J. Filho, P. Gordo, N. Peixinho, R. Melicio, P. Garcia, T. Flohrer

A study on an optimal analytical and an accurate algorithmic landing techniques

Publication date: 1 September 2025

Source: Advances in Space Research, Volume 76, Issue 5

Author(s): José David Gutierrez de Alba, Eva Tresaco, Daniele Mortari

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer