It’s possible that I shall make an ass of myself. But in that case one can always get out of it with a little dialectic. I have, of course, so worded my proposition as to be right either way (K.Marx, Letter to F.Engels on the Indian Mutiny)
Wednesday, March 19, 2025
Satellite images reveal how Andalusia’s forests have changed over the past three decades
The first Landsat series satellite was launched by NASA in the summer of 1972 with the aim of tracking changes in the Earth's surface. It was followed by 8 more. The last of them, put into orbit in 2021, adds to the extensive list of these systems currently hurdling through space in search of key information for planetary management.
This satellite mission, which has already been surveilling the planet for more than half a century, has become the raw material of the Terrestrial Ecology research group at the University of Cordoba, which has processed the images offered by this satellite for the last 28 years (1994-2021) with the aim of analyzing how Andalusian forests have evolved in the face of aridity over the last three decades. Drawing on these images, capable of capturing information in bands of the electromagnetic spectrum not visible to the human eye, the team examined phenological changes in the main Mediterranean forest species, including the holm oak, cork oak, and different species of pine, eucalyptus, olive and chestnut.
A ‘greener’ Andalusia than 30 years ago
One of the work's main conclusions is that Andalusia’s forest landscapes, in general terms, are now greener than they were three decades ago. In other words, Andalusia now has more forest mass than it had at the beginning of the 90s, although this growth was especially marked during the first years (1994-2005).
The reasons for this increase are several and complex, explained Rafael Villar, principal investigator of the group that carried out the work. The abandonment of fields after the rural exodus, plants' own adaptation to adverse climatic conditions (such as Pinus halepensis and Quercus ilex), the effects of atmospheric CO2 fertilization, forest management, and changes in conservation policies could be some of them.
Despite this, as another of the participating researchers, Cristina Acosta, stressed, it is important to highlight that "this trend towards greening is the result of an average of all of Andalusia over time." In fact, this increase in greenery has not been as pronounced in the region's most arid areas, such as Almería, where Pinus sylvestris has shown a more moderate response, due to the scarcity of rainfall.
The work also points to the effect of aridity on the growing season of trees. Although this has been shortened in some of the species studied, such as the olive tree and eucalyptus, its effect is even more pronounced in the wild pine, a tree that is particularly sensitive to scant rainfall, and that reduces its growth period as an adaptation mechanism, depending on the start of the first autumn rains and summer's early arrival.
The importance of remote sensing
The study, in the words of Pablo Salazar, a researcher in the Department of Botany, Ecology and Plant Physiology, underscores the importance of remote sensing as a key tool in species management in a context of global change, especially to monitor large areas in a way complementing field work, which makes it a more cost-effective, efficient and a faster option when it comes to obtaining results on a large scale. In this way, the methodology makes it possible to continuously evaluate the evolution of forests, detect phenomena like decay early, observe forests’ capacities as carbon sinks, and analyze how climate change affects vegetation.
Aridity-induced phenological shifts and greening trends in Mediterranean forest species: Insights from 28 years of Landsat data in southern Spain
Quantum light source for eco-friendly production of biogas
At TU Wien (Vienna), methods are being developed to extract valuable substances from biomass – and quantum cascade lasers offer some very interesting new possibilities.
Much of our waste is far too valuable to simply be incinerated. If it is recycled in a carefully controlled way, not only can thermal energy be generated, but the resulting gas can also be used to produce valuable chemicals - from hydrogen to methane or methanol. However, the gas production process needs to be closely monitored and regulated.
Until now, a very common by-product of gasification – water vapor – has been a particular headache. To control gasification efficiently, it is important to know the water content of the product gas as accurately as possible. However, conventional methods make it difficult to measure the water content. In a collaboration between process engineering and photonics at TU Wien (Vienna), this problem has now been solved using a very special type of light source: terahertz radiation from a quantum cascade laser. State-of-the-art quantum technology now supports environmentally friendly biomass recycling.
Conventional measurements are not enough
"Many chemical components of the product gas can be detected using infrared light," explains Florian Müller, who is researching renewable carbon systems as part of the CO2Refinery PhD programme at the Institute of Chemical, Environmental and Biological Engineering at TU Wien. “Different molecules absorb different wavelengths of infrared light. By measuring which part of which wavelength is absorbed by a sample, it is possible to determine whether the sample contains a certain substance or not.”
However, this is hard to do with water vapor, a by-product which is particularly important for the gas production process. “When you convert biomass into gases, you end up with a complicated gas mixture that contains not only water vapor but also many different hydrocarbons," says Florian Müller. And some of them absorb infrared radiation at exactly the same frequencies as water. This means that it is not possible to say exactly which substance is responsible for the absorption, and therefore the water content in the product gas cannot be accurately determined. You can cool a sample of gas and then measure the amount of condensed water - but this takes time. It is not possible to react quickly to fluctuating water concentrations, and this makes efficient operation difficult.
TU Wien develops terahertz radiation sources
At the same time, however, Michael Jaidl was conducting research at the Institute of Photonics at TU Wien on laser beams in the terahertz range, i.e. radiation with a wavelength slightly longer than the infrared radiation commonly used today for spectroscopic measurements. Michael Jaidl and Florian Müller are old friends who have known each other since school days – and so they came up with the idea of combining their research areas.
Michael Jaidl was able to show that frequencies in the terahertz range can be found that are specifically absorbed only by water molecules and not by the many other substances that exist in significant concentrations in the product gas of a biomass gasification plant. The problem of detecting water vapor can therefore be solved by using terahertz radiation instead of the usual infrared radiation.
Terahertz radiation is difficult to produce. At TU Wien, tricks from quantum technology are being used to produce quantum cascade lasers – tiny semiconductors with a tailor-made geometric structure on the nanometre scale that ensures that only radiation of a very specific wavelength is emitted when an electrical voltage is applied. This quantum cascade laser requires its own cooling, but the two researchers have succeeded in developing a compact, portable device that can reliably measure the water content in hot product gases using a terahertz beam.
First successful tests
"A key advantage of our method is that it provides reliable results over a wide range of water vapor concentrations and temperatures," says Michael Jaidl. "This is because the terahertz radiation we use is particularly strongly absorbed by water vapor – this allows us to use a more compact setup. Another major advantage of the compact design is that the temperature in the measuring cell does not fluctuate as much, which reduces the risk of errors.”
The fact that the new method works perfectly was demonstrated in gas production experiments using waste wood at the Getreidemarkt campus at TU WIen. Now the two researchers and their teams want to improve their technology even further: firstly, to make it even more handy and user-friendly, and secondly, to investigate whether other components of the product gases can be reliably detected using terahertz technology.
Mycobacterium abscessus can cause dangerous lung infections.
Treatment usually requires a combination of antibiotics for more than a year.
Researchers in China report that curcumin, found in turmeric, can enhance treatment with bedaquiline, an antimycobacterial.
Animal studies showed that treatment with the combination led to a faster clearance of the infection.
Washington, D.C.—Mycobacterium abscessus is a fast-growing, pathogenic mycobacteria that can cause lung infections, and people who have respiratory conditions or are immunocompromised face a higher risk. It can also cause skin infections. The microbe is closely related to the one that causes tuberculosis and is naturally resistant to many antibiotics. Infections often require a year or more of a combination of drugs.
A study published this week in Microbiology Spectrum reports a potential way to improve treatment: Add a little spice. Researchers at Shanghai Jiao Tong University, in China, found that adding curcumin boosts the efficacy of bedaquiline, an antimycobacterial used to treat tuberculosis, in combating M. abscessus infections. Curcumin is the compound that gives turmeric its characteristic bright orange color.
“This low-toxicity natural product combined with existing drugs could pioneer new treatment pathways for resistant infections,” said microbiologist Zhe Wang, Ph.D, senior author on the study. “It’s particularly relevant in immunocompromised populations,” Wang added, who are more vulnerable to these infections.
Wang’s lab focuses on innovative approaches to treating infectious disease; those approaches include repurposing known drugs and finding ways to combine natural products with known treatments. They knew that treatment for M. abscessus often leads to poor outcomes—only about half of people who undergo treatment become non-infectious, according to previous studies. Bedaquiline is an antibiotic used to treat multidrug-resistant tuberculosis and has shown some promise in relieving symptoms of non-tuberculosis mycobacterial infections, including M. abscessus. However, the drug does not eliminate all the infectious microbes from a sample.
The researchers, searching for ways to boost the efficacy of bedaquiline, investigated curcumin, which has long been used in traditional Asian medicine to treat a wide variety of conditions. Previous pharmacological studies suggest that curcumin has protective effects against tuberculosis.
In lab studies, the researchers found that bedaquiline alone first inhibited the growth of M. abscessus, but the bacteria began to grow again after 2 weeks. The combination of the drug and curcumin, however, suppressed the growth and reproduction of the bacteria, suggesting that curcumin may act as an antibiotic resistance breaker. In mice, the researchers found that the drug combination slowed or stopped infection better than either compound alone, both in immunocompromised mice and those with a healthy immune system. “The combination demonstrates synergistic enhancement of antibacterial activity and improved infection clearance,” Wang said.
The researchers are now investigating the specific molecular targets that play a role in the mechanisms behind the effects of the combination therapy. They’re also evaluating the combination against other resistant mycobacterial strains and conducting safety assessment to prepare for clinical trials and, down the road, the development of new therapeutics. “This study highlights the innovative value of combining drug repurposing with natural products,” Wang said.
###
The American Society for Microbiology is one of the largest professional societies dedicated to the life sciences and is composed of over 37,000 scientists and health practitioners. ASM's mission is to promote and advance the microbial sciences.
ASM advances the microbial sciences through conferences, publications, certifications, educational opportunities and advocacy efforts. It enhances laboratory capacity around the globe through training and resources. It provides a network for scientists in academia, industry and clinical settings. Additionally, ASM promotes a deeper understanding of the microbial sciences to all audiences.
Journal
Microbiology Spectrum
Coastal guardians pioneer a new way to protect the Florida Keys’ shorelines
Researchers create new tool to identify most effective stabilization methods to prevent erosion and damage
About 8% of the Florida Keys’ coastline is suitable for nature-based or hybrid solutions, while 25.1% is unsuitable, and 67% is already vegetated or naturally protected.
By 2050, sea levels along the United States coast are expected to rise by 0.25 to 0.30 meters, increasing flooding in low-lying areas. Due to its unique geography and infrastructure network, the Florida Keys is particularly at risk of climate hazards such as sea level rise, hurricanes and flooding. Since 2015, the Florida Keys has experienced four hurricanes – Irma (2107), Ian (2022), Helene (2024) and Milton (2024).
Nature-based solutions, such as restoring mangroves and coastal strands, can help mitigate these risks by stabilizing shorelines, improving ecosystems and enhancing resilience to flooding and hurricanes. These solutions, alongside hybrid approaches and soft armoring, which uses natural materials like plants, sand dunes, or rocks to protect shorelines from erosion, offer effective, site-specific protection.
While living shorelines are beneficial, they require careful design and planning to optimize their effectiveness.
Researchers from Florida Atlantic University, in collaboration with The Nature Conservancy, created a new tool to identify the most effective shoreline stabilization methods to prevent erosion and protect the Florida Keys from damage caused by natural forces like waves, tides and storms. Maintaining the shape and integrity of the shoreline reduces the risk of further erosion while protecting ecosystems, properties and infrastructure.
The goal is to guide decisions on using vegetated shorelines or combining them with structures to reduce waves, prevent erosion and protect Florida Keys communities from storms.
Results of the study, published in the Journal of Marine Science Engineering, reveal that nearly 8% of the approximately 2,550 kilometers of shoreline in the Florida Keys is suitable for nature-based solutions – mangrove planting, oyster reefs and beach dune vegetation – or hybrid solutions – some combination of hard structures and vegetation. Conversely, roughly 25.1% of the Florida Keys shoreline was deemed unsuitable for nature-based approaches, and approximately 67% is already vegetated or represents some other type of natural shoreline.
For the study, researchers designed a GIS-based multi-criteria decision tool that facilitates coastal restoration and integrates nature-based solutions into conventional shoreline armoring. They combined spatial analysis tools with expert input to develop a weighted suitability score for various types of shoreline reinforcement where feasible. By integrating data on existing shoreline types – sourced from an updated version of NOAA’s NOS Environmental Sensitivity Index – along with wind and wave exposure and physical environmental factors, they generated a composite Shoreline Relative Exposure Index. Based on this assessment, broadly defined categories of project types were recommended for various combinations of shoreline features and flood risk conditions.
Experts who completed the survey covered coastal engineering, stormwater management, marine biology, habitat restoration, community resilience, urban planning and sustainability. The data was used to calculate scores, which were analyzed through a machine-learning model to identify the best stabilization options for different shoreline types, including developed, undeveloped and protected areas.
Findings indicate that while conventional seawall armoring is needed in some areas of the Florida Keys coastline, hybrid and living shorelines should be prioritized where possible to protect people, habitats and resources. This requires involvement from private stakeholders and coordination among public entities to strengthen coastal resilience.
“Implementing innovative shoreline stabilization methods is crucial as environmental shifts and population growth are expected to exacerbate flood management challenges, making it essential to adopt sustainable, nature-based solutions that enhance resilience and protect vulnerable communities,” said Diana Mitsova, Ph.D., senior author and chair and professor of the Department of Urban and Regional Planning within FAU’s Charles E. Schmidt College of Science.
South Florida’s coastal ecosystems, including mangrove swamps and coastal strands, have already been incorporated into various shoreline management practices that reduce erosion potential and create appropriate habitat conditions. Mangroves are essential for sustaining estuarine and marine ecosystems in South Florida, providing critical habitat, stabilizing shorelines and supporting biodiversity. They offer nesting spots for many species and help the marine food chain by being a main source of small bits of organic matter. Their complex root systems keep the soil in place, reduce water cloudiness and help collect debris and particles in the water.
“New improvements in geospatial technology now allow us to combine human-made impact data with local land and ocean environmental data across large areas,” said Chris Bergh, field program director at The Nature Conservancy. “This information helps coastal managers identify key areas that need protection or are important for commercial and recreational activities. By doing this, it can help avoid conflicts between different uses of the coast and create a more flexible, forward-thinking and sustainable way of managing the area.”
The data from this study can be accessed through The Nature Conservancy’s Coastal Resilience, an online tool that uses GIS technology to help users visualize proposed shoreline stabilization methods tailored to different areas of the Florida Keys. It also allows users to overlay local data, like projected sea level rise, coastal habitats and land use.
Study co-authors are Kevin Cresswell, Ph.D., an adjunct faculty in the FAU Department of Urban and Regional Planning; Melina Matos, Ph.D., an assistant professor in the FAU Department of Urban and Regional Planning; Stephanie Wakefield, Ph.D., an assistant professor in the FAU Department of Urban and Regional Planning; Kathleen Freeman, GIS specialist, The Nature Conservancy; and William Carlos Lima, Ph.D., an adjunct faculty in FAU Department of Urban and Regional Planning.
- FAU -
About Florida Atlantic University: Florida Atlantic University, established in 1961, officially opened its doors in 1964 as the fifth public university in Florida. Today, Florida Atlantic serves more than 30,000 undergraduate and graduate students across six campuses located along the Southeast Florida coast. In recent years, the University has doubled its research expenditures and outpaced its peers in student achievement rates. Through the coexistence of access and excellence, Florida Atlantic embodies an innovative model where traditional achievement gaps vanish. Florida Atlantic is designated as a Hispanic-serving institution, ranked as a top public university by U.S. News & World Report, and holds the designation of “R1: Very High Research Spending and Doctorate Production” by the Carnegie Classification of Institutions of Higher Education. Florida Atlantic shares this status with less than 5% of the nearly 4,000 universities in the United States. For more information, visit www.fau.edu.
Powering artificial intelligence comes with a massive energy bill attached. Professor Wolfgang Maaß and his research team at Saarland University and the German Research Center for Artificial Intelligence (DFKI) want to make AI up to 90% percent more energy efficient. To improve AI’s carbon footprint and to reduce costs, the Saarbrücken team is rethinking data centres, large language models and image analysis models – and their research is opening up access to powerful AI models for small and medium-sized companies. One of the methods that researchers Sabine Janzen (right) and Hannah Stein (left) are using is known as ‘knowledge distillation’ – a form of model compression that transfers knowledge from complex models to simpler ones. After all, we don't need to read an entire library to answer a specific question. Instead, we focus on those books that are relevant to the question at hand. From 31 March to 4 April, the researchers will be showcasing their work at this year’s Hannover Messe at the stand of the Federal Ministry for Economic Affairs and Climate Action (Hall 2, Stand A18).
Powering artificial intelligence comes with a massive energy bill attached. Professor Wolfgang Maaß and his research team at Saarland University and the German Research Center for Artificial Intelligence (DFKI) want to make AI up to 90% percent more energy efficient. To improve AI’s carbon footprint, the Saarbrücken team is rethinking data centres, large language models and image analysis models – and their research is opening up access to powerful AI models for small and medium-sized companies. From 31 March to 4 April, the researchers will be at this year’s Hannover Messe showcasing their work at the stand of the Federal Ministry for Economic Affairs and Climate Action (Hall 2, Stand A18).
Data centres consume vast quantities of energy. According to Bitkom, the leading industry association in Germany’s digital sector, the electricity requirements to power data centres have more than doubled over the past decade. And with digital transformation only just out of the starting blocks, this trend is really gathering pace. Storing, processing, transmitting and retrieving data takes energy. Artificial intelligence, in particular, is a huge energy guzzler. Globally, multiple terawatt hours are being used to train and run today’s massive AI models. (One terawatt hour is equal to one billion kilowatt hours of electrical energy). Using these models to generate images and texts also consumes vast amounts of energy. As a result, data centres are having to get bigger and bigger, which means they need more and more electricity to power and cool the huge numbers of processors involved, which in turn is causing a massive uptick in their carbon footprint. None of this is helping Europe achieve its goal of net-zero greenhouse gas emissions by 2050. Clearly something has to change.
‘AI can become far more energy efficient. With the right approach, we can make the data centres of the future much more sustainable,’ says Professor Wolfgang Maaß, who conducts research at Saarland University and the German Research Center for Artificial Intelligence (DFKI). In an effort to curb AI's hunger for energy and to conserve resources, his research team is developing leaner, customized AI models. They also want to identify ways that data centres can become more energy smart.
‘By making the models smaller and more efficient, we’re helping to drive sustainability,’ says Dr. Sabine Janzen, a senior research scientist in Wolfgang Maaß's team. ‘Our work is also opening up access to powerful AI models for small and medium-sized businesses, because these smaller, leaner AI models don’t need a large technical infrastructure. This will enable everyone – not just the big players – to leverage this new technology,’ says Janzen.
Today's AI chatbots such as ChatGPT and visual AI models use trillions of parameters and utilize vast datasets to perform their tasks. The amount of energy they consume is correspondingly huge. The researchers in Saarbrücken are developing ways to reduce this energy consumption, without compromising the quality of the output from these leaner AI models. ‘A central element of our work is a technique known as knowledge distillation. It’s a type of compression technique that enables us to make models that are smaller and therefore more energy efficient, but that perform just as well as the larger models,’ explains Sabine Janzen.
The approach used by the research team could be described as follows: When looking for the answer to a specific question, you don’t read an entire library; you focus instead only on those books that are relevant to your question. The researchers in Saarbrücken extract smaller, more focused and more energy-efficient ‘student’ models from larger ‘teacher’ models. By distilling the knowledge needed to perform tasks in a specific area and reducing it to the essentials, they can reduce the size of the data models by up to ninety percent. Model parameters that are not relevant to the area of interest are not touched. ‘In terms of inference speed, i.e. how quickly the model can process input data and produce results, these student models perform at a level comparable to that of the larger teacher models, but require 90% less energy to do so,’ explains Janzen.
By using another automated efficiency technique known as ‘neural architecture search’ (NAS), the team has also achieved some impressive results with visual AI models, i.e. models that process digital image data. ‘Our most recent results show that we can use the NAS method to reduce the size of the models by around ninety percent,’ says Sabine Janzen. In this work, the researchers focus on machine learning with artificial neural networks – a very energy-intensive AI method that can analyse large volumes of data. Artificial neural networks are designed to mimic the human brain. Our brains contain many billions of nerve cells, called neurons, that are connected to each other via trillions of synapses. A synapse is essentially the interface between two neurons across which the two nerve cells communicate with each other. When we learn something new, neurons send electrical signals to each other across synapses, as we continue learning, the same neurons keep firing together and the connections between them get stronger, whereas the connections between inactive neurons become weaker.
Learning processes in artificial neural networks are similar and by feeding these networks large amounts of data, they can be trained to recognize patterns in natural language or in images. But whereas the brain is a master of energy-efficient learning, training a large artificial neural network requires a lot of computing power and a lot of energy. Training an artificial neural network so that it can yield meaningful results also involves a significant amount of human input. Typically, these artificial networks are designed and configured manually, and the many parameters involved are adjusted and optimized by experts until they perform at the required level. This is where the Saarbrücken researchers bring ‘neural architecture search’ (NAS) into play. ‘Instead of designing the neural networks manually, we automate the design optimization process using NAS,’ explains Sabine Janzen. ‘NAS allows us to examine different network architectures and optimize them to create a model that delivers high performance, efficiency and reduced costs.’
To test these compacter AI models in practice, Wolfgang Maaß's team is working together with the steel company Stahl Holding Saar. The aim is to teach the artificial neural networks to sort steel scrap efficiently. In order to produce new steel from scrap steel, producers need scrap of the right quality. Only certain types of scrap can be recycled for the manufacture of high-quality steels. However, the steel scrap that gets delivered to the smelting plant is a mix of all types and has to be sorted. Scrap sorting can be automated, but so far, the AI model is too big to be practical. ‘We have compressed the visual AI sorting model, making it compacter and more energy efficient. In fact, on certain metrics, the smaller model even performs better, making the steel recycling process more efficient,’ says Janzen. Where previously a huge AI model would have required a lot of energy to operate, a small, customized, energy-efficient model is now able to perform the same task.
The researchers start by training their models with the full dataset that contains all the information. They then shrink the AI models using knowledge distillation and specially compiled neural networks so that the models only contain those parameters that are really necessary for the task at hand. In this particular case, the aim is to create an AI that has all the knowledge it needs to be able to analyse camera images to classify the scrap steel being delivered to the site.
The Saarbrücken research team is also working with partners to outline a concept and compile recommendations for sustainable data centres and energy-efficient AI. ‘Up until now it has been difficult to estimate just how much energy is needed to create and operate an AI model. That makes it harder for businesses to plan ahead,’ explains PhD student Hannah Stein who is conducting research into these energy-efficient AI models. ‘We’re currently developing a tool that provides reliable forecasts of the energy consumed by and the costs associated with the different AI models,’ says Stein. Data centres and AI users can then use this information to plan more effectively, identify inefficient processes and take corrective action as necessary – for example, scheduling heavy computational loads at times when the price of electricity is low.
The research being conducted by Professor Wolfgang Maaß and his team was selected for the Federal Ministry for Economic Affairs and Climate Action's stand at this year’s Hannover Messe.The team will be presenting the latest results from the federally funded ‘ESCADE’ project, which is based at the German Research Centre for Artificial Intelligence DFKI.
Background:
ESCADE (‘Energy-Efficient Large-Scale Artificial Intelligence for Sustainable Data Centers’) is a three-year project with a budget of around €5 million being financed by the Federal Ministry for Economic Affairs and Climate Action (BMWK).
The project will run until the end of April 2026. The ESCADE consortium is made up of the research team headed by Wolfgang Maaß (Saarland University and DFKI), NT Neue Technologie AG, Stahl-Holding-Saar GmbH & Co. KGaA, SEITEC GmbH, Dresden University of Technology, the University of Bielefeld and the Austrian applied research institute Salzburg Research. https://escade-project.de
