Sunday, October 05, 2025

 

Natural barriers disturb the ecosystems in northern Sweden’s rivers




Umea University
Illustration A 

image: 

Previous theories have assumed that water, sediment and plant seeds move freely downstream (top image). However, lakes and rapids can become natural barriers where water can be slowed down as well. The solid lines show high transport or dispersal, and the dotted lines show reduced transport due to natural barriers.

view more 

Credit: Marlene Lahti





Rivers in northern Sweden do not always become wider or richer in species further downstream. Natural barriers shape the flow and stop plants from spreading, new research from Umeå University shows.

“Our results suggest that many foundational assumptions in river science may not apply here,” says Lina Polvi Sjöberg, Associate Professor at the Department of Ecology, Environment and Geoscience at Umeå University.

The rivers in northern Sweden flow through terrain shaped by the last Ice Age. The landscape is dotted with lakes and covered in sediment brought by the ice: sand, gravel and plenty of boulders. A news study from Umeå University shows that this creates natural barriers that disrupt the flow of water. It also hinders the transport of sediment and the dispersal of plant seeds. All of this affects both the shape of the streams and the mix of plant types found along the shores.

“We found that these landscapes are naturally fragmented, and that local conditions – such as sediment type and proximity to lakes – play a much larger role than previously thought,” says Lina Polvi Sjöberg.

Together with researcher Lovisa Lind, she studied two catchments in norther Sweden, Bjurbäcken and Hjuksån, located above and below the highest coastline after the last glaciation. Using maps and field studies, they analysed the shape of the streams and the plant life along several miles of the shores.

Their findings challenge widely accepted theories that streams become wider and more biologically diverse downstream. The researchers found no clear correlations between the drainage area and channel width, and no consistent increase in plant species diversity downstream within these medium-sized catchments.

“In the catchment below the former highest coastline, we did see slightly stronger patterns, likely due to finer sediments from the sea. But overall, the presence of lakes and coarse glacial deposits breaks up the expected downstream trends,” says Lina Polvi Sjöberg.

The study shows that lakes stop transport of plant seeds by water. This leads to more diversity in plant life between nearby parts of the rivers. The researchers also saw an unexpected pattern: that species density (the number of species in a given area) was constant – or even decreased – downstream.

These results will have an impact on river restoration in areas shaped by the Ice Age. In these fragmented systems, passive recovery – where plants recolonise naturally – will probably not succeed. Instead, active interventions such as planting and physically reshaping the stream channels might be necessary.

“Restoration strategies need to be adapted to these local realities,” says Lina Polvi Sjöberg.

Streams in northern Sweden that were affected by the last Ice Age are fragmented, with three main types of reaches: rapids with large boulders, flats in sand or peat, and lakes.

Credit

Marlene Lahti


The terrain in northern Sweden was shaped by the last Ice Age.

Credit

Richard Mason



The terrain in northern Sweden was shaped by the last Ice Age.

Credit

Lina Polvi Sjöberg

AI optimizes evacuation, diagnosis, and treatment of wounded soldiers in Ukraine

Analyzing real-time data, AI enhanced evacuation efficiency, optimized routes, and prioritized patients by injury severity, according to study


American College of Surgeons





Key Takeaways 

  • In Ukraine, the medical role of AI has evolved from limited use to wide-ranging applications in evacuation, diagnosis, predictive analytics, and treatment of wounded soldiers. 

  • An analysis of 68 wounded soldiers showed that by analyzing data in real time from wearable medical devices, AI enhanced treatment by assisting medical personnel in delivering personalized care based on a soldier’s medical history, condition, and available resources. 

  • The researchers found that AI not only accelerated drug delivery, identified new treatments for injuries, and supported artificial limb fit and selection for soldiers with limb loss 


CHICAGO (October 3, 2025) — Analyzing data in real-time from artificial intelligence (AI)-powered wearable medical devices has enhanced the treatment of wounded Ukrainian soldiers by helping medical personnel deliver personalized care based on the soldier’s medical history, condition, and available resources, according to a new study. 

The research will be presented at the American College of Surgeons (ACS) Clinical Congress 2025 in Chicago, October 4-7. 

An analysis showed that AI-powered wearable devices such as electrocardiographs (ECGs), glucose and blood pressure monitors, multi-sensor vital patches, and advanced smart helmets tracked vital signs and alerted medical personnel to potential health risks in 68 patients with gunshot and mine-explosive injuries, the study said.  

“Although we don’t know the extent of improvement, there is clear, consistent, and robust evidence that AI led to faster identification of life-threatening injuries, faster arrival of supplies and drugs, and stronger rehabilitation outcomes,” said the study’s author Evgeni Kolesnikov, MD, PhD, FICS, FIMSA, of Shupyk National Healthcare University of Ukraine in Kyiv. 

Key Findings 

The study showed that AI: 

  • Accelerated drug delivery: AI-driven clinical decision support analyzed soldiers’ vital signs to recommend optimal drug dosing for conditions such as shock, pain, or infection, Dr. Kolesnikov said. In addition, AI triage algorithms prioritized limited supplies of drugs, such as morphine, tranexamic acid, and ketamine, for patients with the highest likelihood of survival. 

  • Identified new treatments for injuries: AI analyzed millions of chemical structures to predict which compounds could stop bleeding, prevent infection, or accelerate wound healing in combat-related injuries. AI also helped identify drugs (like existing anti-inflammatories) that could be repurposed to speed tissue regeneration after blast injuries and identified biomarkers of poor healing, infection risk, or organ failure.  

  • Supported artificial limb selection: AI is transforming how artificial limbs (prosthetics) are selected, fitted, and personalized for wounded soldiers with limb loss, improving their quality of life. For example, AI analyzes 3D imaging scans of the damaged limb to help design and custom-fit components of artificial limbs with millimeter-level precision, reducing skin breakdown and pain.  

The role of AI in providing medical care to wounded soldiers has evolved over the course of the war from limited practical use to a wide range of applications in evacuation, diagnosis, predictive analytics, and treatment of wounded soldiers with AI system-level coordination, Dr. Kolesnikov said. For example, using AI, Ukrainian troops can help evacuate wounded soldiers by assessing the terrain, avoiding obstacles, and locating the wounded. When sending medics poses an unacceptable risk, ground vehicles with autonomous navigation systems can be sent to evacuate casualties from the front lines, he said.  

Although there is no central intelligence that receives all of the relevant patient data from the various devices and makes recommendations, Ukraine’s military medical system relies on built-in AI modules, such as AI for triage (vital signs and injury assessment), AI for route optimization, and AI for image interpretation, among others, which are then integrated through command-and-control platforms. 

“AI does not replace doctors and surgeons, but expands their capabilities, reducing evacuation times, increasing the accuracy of diagnostics and surgical treatment, helping to save more lives with limited resources,” Dr. Kolesnikov said. 

Disclosures: The author has no relevant disclosures. 

Citation: Kolesnikov E. Artificial Intelligence in the Evacuation, Diagnosis, and Treatment of Wounded Soldiers During Wartime, Scientific Forum, American College of Surgeons (ACS) Clinical Congress 2025. 

Note: This research was presented as an abstract at the ACS Clinical Congress Scientific Forum. Research abstracts presented at the ACS Clinical Congress Scientific Forum are reviewed and selected by a program committee but are not yet peer reviewed. 

 

From palm waste to carbon catcher: Malaysian scientists turn agricultural leftovers into high-performance CO₂ sponge



Dr. Azam Taufik Mohd Din and team at Universiti Sains Malaysia pioneer a sustainable, low-cost adsorbent using oil palm ash—backed by machine learning predictions with near-perfect accuracy




Biochar Editorial Office, Shenyang Agricultural University

Enhanced CO2 capture using KOH-functionalized oil palm ash adsorbent: experimental and applied machine learning approach 

image: 

Enhanced CO2 capture using KOH-functionalized oil palm ash adsorbent: experimental and applied machine learning approach
 

view more 

Credit: Syamima Nasrin Mohamed Saleh, Fakhrony Sholahudin Rohman, Dinie Muhammad, Syafini Mohd Hussin, Bassim H. Hameed, Chew Thiam Leng & Azam Taufik Mohd Din





In Malaysia, one of the world’s top producers of palm oil, millions of tons of oil palm ash (OPA) are left behind as agricultural waste every year—a disposal challenge that could soon become a climate solution. Now, groundbreaking research from Universiti Sains Malaysia (USM) shows that this humble byproduct can be transformed into a powerful, eco-friendly material capable of capturing carbon dioxide from the air. Published on August 18, 2025, in Carbon Research as an open-access original article, this innovative study was led by Dr. Azam Taufik Mohd Din from the School of Chemical Engineering at Universiti Sains Malaysia’s Engineering Campus in Nibong Tebal, Penang. The team didn’t just repurpose waste—they engineered it. By treating raw oil palm ash with acid, then subjecting it to carbonization and chemical activation using potassium hydroxide (KOH), they created a new material dubbed OPA-KOH(1:2). The result? A tailor-made adsorbent with a highly optimized mesoporous structure—pores so precisely shaped that they allow CO₂ molecules to flow in easily and stick effectively. Despite having a modest surface area of 30.95 m²/g—far lower than many commercial activated carbons—the material achieved an impressive CO adsorption capacity of 2.9 mmol/g. That performance rivals or even exceeds more expensive materials with much higher surface areas, proving that pore architecture matters more than size alone. “This isn’t just recycling—it’s upcycling at the molecular level,” says Dr. Mohd Din. “We’re taking a waste product that often ends up in landfills and turning it into a high-performance tool for carbon capture.”

How It Works: Small Pores, Big Impact

The secret lies in the structure. With an average pore size of 72.71 Å, OPA-KOH(1:2) creates an ideal environment for CO₂ molecules to enter quickly and bind efficiently. Comprehensive analysis revealed that adsorption is exothermic and spontaneous, primarily driven by physisorption—a process where CO₂ sticks to the surface through weak physical forces—supported by a minor contribution from weak chemisorption, enhancing overall stability. This dual mechanism means the material can capture CO₂ effectively under realistic conditions, making it a promising candidate for real-world carbon capture, utilization, and storage (CCUS) systems.

Machine Learning Meets Materials Science

What sets this study apart is its fusion of experimental science with artificial intelligence. Recognizing that traditional modeling has limits, the team deployed machine learning (ML) algorithms to predict CO₂ adsorption behavior. Among several models tested, a bilayered neural network (NN) emerged as a star performer—achieving an astonishing R² value greater than 0.99, meaning it predicted experimental outcomes with near-perfect accuracy. “This shows ML isn’t just a trend—it’s becoming essential,” explains Dr. Mohd Din. “It allows us to simulate, optimize, and understand adsorption processes faster and more deeply than ever before.” The successful integration of ML opens doors for accelerating the design of next-generation adsorbents, reducing trial-and-error in the lab, and scaling up sustainable technologies more efficiently.

A Win for Sustainability and Industry

Malaysia produces over 20 million tons of palm oil annually—generating vast amounts of residue. By converting oil palm ash, a problematic waste, into a valuable carbon-capture material, this research offers a closed-loop, circular economy solution. It also presents a cost-effective alternative to synthetic adsorbents, which are often energy-intensive to produce and expensive to deploy at scale. “Sustainable doesn’t have to mean less effective,” says Dr. Mohd Din. “Our work proves that green materials can compete—and win—on performance.”

Spotlight on Universiti Sains Malaysia’s Leadership

This study highlights the growing strength of the School of Chemical Engineering at Universiti Sains Malaysia in advancing clean energy and environmental technologies. Located on the Engineering Campus in Nibong Tebal, Penang, the school is emerging as a regional hub for innovation in waste valorization, carbon management, and AI-driven chemical engineering.

Dr. Azam Taufik Mohd Din’s leadership exemplifies how local solutions—rooted in regional resources and global science—can contribute to planetary challenges.

The Road Ahead: Scaling Up the Solution

The success of OPA-KOH(1:2) paves the way for pilot-scale testing in flue gas treatment, biogas upgrading, and direct air capture systems. Future work will explore regeneration cycles, long-term stability, and integration into industrial processes. With climate targets tightening worldwide, affordable and scalable CCUS technologies are urgently needed. This research delivers both a material and a method: a low-cost, high-efficiency adsorbent born from biomass waste, guided by the predictive power of machine learning. So the next time you enjoy a product containing palm oil, remember: from that same industry’s waste, scientists in Malaysia are building a cleaner future—one CO₂ molecule at a time.

 

=== 

  • Title: Enhanced CO2 capture using KOH-functionalized oil palm ash adsorbent: experimental and applied machine learning approach
  • Keywords: Carbon dioxide adsorption; Oil palm ash-based adsorbent; KOH activation; Machine learning; Bilayered neural network model
  • Citation: Mohamed Saleh, S.N., Rohman, F.S., Muhammad, D. et al. Enhanced CO2 capture using KOH-functionalized oil palm ash adsorbent: experimental and applied machine learning approach. Carbon Res. 4, 60 (2025). https://doi.org/10.1007/s44246-025-00227-3 

 

=== 

About Carbon Research

The journal Carbon Research is an international multidisciplinary platform for communicating advances in fundamental and applied research on natural and engineered carbonaceous materials that are associated with ecological and environmental functions, energy generation, and global change. It is a fully Open Access (OA) journal and the Article Publishing Charges (APC) are waived until Dec 31, 2025. It is dedicated to serving as an innovative, efficient and professional platform for researchers in the field of carbon functions around the world to deliver findings from this rapidly expanding field of science. The journal is currently indexed by Scopus and Ei Compendex, and as of June 2025, the dynamic CiteScore value is 15.4.

Follow us on FacebookX, and Bluesky