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)
Sunday, July 26, 2026
Regional drought prediction from Sentinel-2 time series using Random Forest, DNN, and 1D-CNN: a case study in Marchfeld, Austria
Total precipitation sums (in mm) from May 1st to October 31st across multiple years for the Marchfeld region, based on SPARTACUS v2.1 daily precipitation data
A new study published in Big Earth Data proposes an operational framework that integrates multiple machine learning and deep learning models with high-resolution Sentinel-2 data to estimate agricultural drought conditions in irrigated and non-irrigated maize fields in Eastern Austria. Results from 2018–2023 show that deep learning models, particularly the Deep Neural Network (DNN), achieved the best performance, while non-irrigated fields exhibited higher prediction accuracy due to clearer drought signals, demonstrating the value of Sentinel-2–based deep learning for operational drought monitoring.
Citation
Ghorbanzadeh, O., un Nisa, Z., Gholamnia, K., Ogutu, B., Dobrowolska, E., Volden, E., … Dash, J. (2026). Regional drought prediction from Sentinel-2 time series using Random Forest, DNN, and 1D-CNN: a case study in Marchfeld, Austria. Big Earth Data, 1–27. https://doi.org/10.1080/20964471.2026.2649428
Abstract
In recent years, the integration of machine learning (ML) with earth observation data has improved early warning systems and drought management strategies through advanced drought monitoring and prediction. This study proposes an operational framework that synergistically combines multiple ML models with high-resolution Sentinel-2 data to estimate agricultural drought conditions, while focusing on maize fields in Eastern Austria. The study area includes both irrigated and non-irrigated fields, allowing for comparative performance analysis under different water regimes. A comprehensive network of reference points was established using over 20 satellite-derived indices integrated with ground data including field capacity, precipitation, and soil composition. The integrity of the reference points was further tested with temporal analysis and expert validation. The study employed Random Forest (RF), and two deep learning models, Deep Neural Network (DNN) and One-Dimensional Convolutional Neural Network (1D-CNN), to generate pixel-level drought maps from pre-processed time-series sentinel-derived variables. The prediction accuracy is evaluated over multiple years (2018–2023). The results reveal distinct differences in model performance, with non-irrigated fields demonstrating higher prediction accuracy (on average 79%) and lower error metrics (on average 0.15), likely due to the clearer drought signals they present. In contrast, irrigated fields present more significant drought patterns, which increase the complexity of prediction, reflected in lower prediction accuracy (on average 77.4%) and error metrics (on average 0.16). Among all models, DNN demonstrated the best overall performance. The results highlight the advantage of deep learning with Sentinel-2 in operational drought monitoring for improved agricultural drought management.
Big Earth Data is an interdisciplinary Open Access journal which aims to provide an efficient and high-quality platform for promoting the sharing, processing and analyses of Earth-related big data, thereby revolutionizing the cognition of the Earth’s systems. The journal publishes a wide range of content, including Research Articles, Review Articles, Data Notes, Technical Notes, and Perspectives. It is now included in ESCI (IF=3.8, Q1), Scopus (CiteScore=9.0, Q1), Ei Compendex, GEOBASE, and Inspec. Starting from 2023, Big Earth Data has announced a new award series for authors: Best and Outstanding Paper Awards.
[Research Article] Regional drought prediction from Sentinel-2 time series using Random Forest, DNN, and 1D-CNN: a case study in Marchfeld, Austria
The lost golden age of language diversity
New modelling suggests that tens of thousands of languages were spoken between 1,000 and 3,000 years ago - and that their decline began well before European colonial expansion
Max Planck Institute for Evolutionary Anthropology
Contemporary languages and endangerment status. Made with a Natural Earth base map using data from Glottolog 5.3 and information adapted from GlottoScope.
Fewer languages before agriculture: The researchers estimate that between approximately 4,500 and 6,200 languages were spoken at the beginning of the Holocene, around 12,000 years ago - probably fewer than the roughly 7,500 languages spoken and signed today.
A linguistic “golden age”: As the global human population grew, language diversity also increased. The models suggest that tens of thousands of languages may have existed between 1,000 and 3,000 years ago.
A deep history of language loss: The decline in linguistic diversity appears to have begun with the expansion of large states and empires, long before modern European colonialism.
Today’s languages are survivors: The languages spoken today represent a small and historically biased sample of past linguistic diversity, with important implications for how researchers explain global patterns in language and culture.
Today, people speak or sign around 7,500 languages. Yet this figure provides only a snapshot of a constantly changing history. How many languages existed before written records? Did agriculture reduce linguistic diversity by allowing expanding farming populations to replace smaller hunter-gatherer groups? Or did population growth initially create more opportunities for new languages to emerge?
An international research team involving researchers from Spain, the USA and Germany have reconstructed plausible trajectories of global language diversity across the last 12,000 years. Their study, published in Science, suggests that language diversity increased for most of the Holocene, reached a remarkable peak between approximately 1,000 and 3,000 years ago, and then declined rapidly thereafter.
Reconstructing a history without written records
“This paper presents an innovative approach to understanding the dynamics of human cultural diversity in the past, addressing one of the central yet least accessible questions in the human sciences. Progress has been limited not by a lack of theoretical interest but by the exceptional difficulty of observing the relevant dynamics empirically”, says Marcus Hamilton of the University of Texas at San Antonio.
Direct evidence for the number of languages is almost entirely absent before the appearance of writing around 6,000 years ago, and remains fragmentary until quite recently. The researchers therefore combined ethnographic information, estimates of prehistoric population size, and statistical and social-computational modelling.
They began with ethnographic data from 171 hunter-gatherer and fisher societies whose traditional subsistence and mobility had not been profoundly transformed by contact with food-producing populations. These data allowed the researchers to estimate the likely distribution of ethnolinguistic group sizes near the beginning of the Holocene.
The team then combined these estimates with reconstructions suggesting that the global human population 12,000 years ago was between approximately 4.4 and seven million. Assuming that ethnolinguistic groups at that time generally corresponded to distinct languages, the models produced an early-Holocene estimate centred on roughly 4,500 to 6,200 languages, although broader plausible estimates ranged from around 3,300 to 7,800.
“One of the first surprises was that the world immediately before agriculture was probably not exceptionally rich in languages. There were most likely fewer languages than there are today. Linguistic diversity then grew alongside the human population for thousands of years”, says Russell Gray, Director of the Department of Linguistic and Cultural Evolution at the Max Planck Institute for Evolutionary Anthropology in Leipzig.
Tens of thousands of languages
Agriculture, technological innovation and more stable food production allowed the global population to expand dramatically. At the same time, however, societies became larger, meaning that progressively more people could share the same language.
To capture these opposing processes, the researchers modelled thousands of possible trajectories connecting estimates at the beginning of the Holocene with the present. The models allowed the number of people per language to increase over time, while incorporating independent reconstructions of global population growth.
Although the precise trajectory remains uncertain, the models consistently produced a striking general pattern. Language diversity increased over most of the Holocene and typically peaked between 1,000 and 3,000 years ago. Many trajectories imply an approximately tenfold increase from early-Holocene levels, placing the peak number of languages in the tens of thousands.
“The most unexpected result is how recent the peak appears to have been. The period of greatest linguistic diversity may have occurred not in the remote Palaeolithic, but during the era of expanding states and early empires”, says Claire Bowern of Yale University.
The authors describe this period as a linguistic “golden age”. It has largely escaped previous discussion because most surviving historical records were produced by the expanding societies whose languages eventually became dominant.
Language loss before modern colonialism
The linguistic golden age was followed by a rapid decline, particularly during the last two millennia. The results therefore challenge the view that the present language-endangerment crisis began only with European colonial expansion around 500 years ago.
European colonialism unquestionably intensified language replacement on an enormous scale. The new models suggest, however, that the process had deeper roots in the expansion of earlier multinational states and empires. Such expansions brought previously separated populations into sustained contact and spread dominant languages together with political institutions, technologies, cultural practices and pathogens.
The model does not identify which particular empires, regions or historical events caused the global decline. Rather, it shows that a very large reduction in linguistic diversity must have taken place before the present.
Today’s languages passed through a bottleneck
This loss means that the languages spoken today are not a representative sample of those that once existed. Languages associated with expanding populations were disproportionately likely to survive, while the languages of absorbed, displaced or declining populations disappeared.
This historical bottleneck may also have altered the global frequency of grammatical structures, speech sounds and other linguistic features. Researchers often explain common features by suggesting that they are especially easy to learn, efficient to communicate or well suited to human cognition. The new study suggests that demographic expansion and extinction must also be considered.
“The languages we see today are the survivors of a massive and highly selective historical bottleneck. A linguistic feature may be common not because it is inherently superior, but because it happened to be carried by populations that expanded. Extinction may have shaped linguistic diversity much more profoundly than we previously appreciated”, says Damian Blasi, lead author of the study and researcher at the Catalan Institution for Research and Advanced Studies (ICREA) and Harvard University.
Because languages are often transmitted together with religious practices, kinship systems, social institutions and other forms of cultural knowledge, the authors argue that this bottleneck extended beyond language. Many cultural traditions may also have vanished without leaving direct historical evidence.
A model, not a prehistoric census
The researchers emphasise that the study provides a coarse global reconstruction rather than a direct count of prehistoric languages. Its estimates depend on assumptions about early ethnolinguistic group sizes and the long-term relationship between population and language numbers. Regional histories undoubtedly differed, and short-term collapses caused by warfare, disease and environmental disruption cannot be individually reconstructed by the model.
Nevertheless, the central pattern persisted across a wide range of modelling conditions and alternative assumptions. The results suggest that the history of language diversity was not a simple, continuous decline from a highly diverse Palaeolithic world. Instead, linguistic diversity rose with human population growth, reached an extraordinary but short-lived peak, and then underwent a severe global contraction.
Persistent financial hardship accelerates age-related cognitive decline, according to a new study led by University College London (UCL) researchers.
The study, published in Innovation in Aging, looked at data from 2,759 people in the UK who filled in questionnaires throughout their lives as part of the MRC National Survey of Health and Development (also known as the 1946 British cohort study).
The research team found that people who experienced either persistent money struggles or persistent low income in early and middle adulthood performed less well in cognitive tests by the age of 53.
Looking at a subgroup of people who had brain scans, the team found that those who experienced persistent low income had worse brain health (including more brain shrinkage) in later life (ages 69 to 71).
These links remained even after accounting for factors that might have skewed the results such as childhood cognition, education level and childhood disadvantage.
Corresponding author Dr Jacques Wels (Unit for Lifelong Health & Ageing at UCL) said: “Most studies on cognitive ageing look at financial hardship at only a single point in time. Our study using several decades of data allows us to see that it is the accumulation of hardship over many years that is linked to the worst cognitive health outcomes, rather than occasional episodes of adversity.”
Senior author Professor Praveetha Patalay (Unit for Lifelong Health & Ageing and Centre for Longitudinal Studies, UCL) said: “Our findings suggest that supporting people facing financial hardship and reducing chronic poverty could also help prevent cognitive decline and dementia cases in the future.”
The research team found that the link between financial adversity and poorer brain health in later life was particularly strong for men, those who experienced childhood disadvantage, and those who carried a genetic variant, APOE-ε4, that raises Alzheimer’s risk.
Men who experienced persistent financial adversity also did less well in cognitive tests at aged 53 than female counterparts. Possible reasons for this difference might include that disadvantaged men may have worse health behaviours (such as smoking and alcohol misuse) than disadvantaged women and that men might experience the stress of financial adversity more, especially as they would have been the primary breadwinners in this cohort born in 1946.
The researchers mention different mechanisms that could link cognitive ageing and financial hardship including inflammation, which is known to accelerate brain ageing and which is caused by chronic stress.
Another factor is that frequently worrying about money may increase cognitive load, leaving people with less bandwidth for other cognitive tasks.
The researchers also said that, while those who experienced financial hardship or low income did less well on cognitive tests at 53, their performance in a memory test declined more slowly between 53 and 69, which was likely due to them already having suffered significant cognitive losses compared to counterparts who did not have persistent financial strain.
Study participants were asked about their household income at three points, at the ages of 26, 43 and 53, and classed as having persistent low income if they were in the bottom 20% of the group at least twice, which amounted to 16% (about one in six) of the participants.
Financial hardship was assessed using questions such as whether people found it hard to manage on their income and whether they had had trouble paying for bills. Participants were classed as having experienced persistent hardship if they scored above a threshold on the questionnaire at least twice between the ages of 36 and 53. This represented 12% (about one in eight) of the participants.
Cognitive tests assessed verbal memory and processing speed; magnetic resonance imaging (MRI) scans allowed researchers to measure factors such as brain atrophy (shrinkage) and ventricular expansion – i.e., the expansion of fluid-filled cavities in the brain – which is a sign of poor brain health.
The 1946 British cohort study, hosted by UCL, is the world’s longest continuously running birth cohort study. Participants (who were enrolled at birth) celebrated their 80th birthday earlier this year.
New research from the University of St Andrews and Université de Neuchâtel, has uncovered the mechanism which allows sperm whales to sleep in a unique vertical position, just under the ocean surface.
Sperm whales are the only whale known to rest vertically. It’s thought to enable them to sleep effectively by buffering them from surface wave action, while avoiding the energy needed to dive to deeper depths. However, until now, it’s been unclear how they maintained their position just below the sea surface.
In a paper published in the Journal of Experimental Biology, researchers discovered that while resting, sperm whales release gas bubbles to help them regulate their buoyancy and remain below the sea surface. This regulation is crucial as Sperm Whales are positively buoyant naturally due to the large amounts of spermaceti oil in their heads. Additionally, sperm whales are breath-hold divers, if they slowly drift up while resting, the gas in their lung will slowly expand, releasing bubbles enables them to counteract the expanding diving gas and achieve neutral buoyancy.
Researchers were able to collect data from placing small tags using suction cups on Sperm Whales off the Norwegian coast. The tags record sound and 3-D animal movements. Clear bubble sounds were recorded by the tags, and animal movement data was used to create a simulation based on tissue density, drag through the water and gas volumes in their body.
Results from the simulation confirmed that release of bubbles functioned to reduce the whales’ positive buoyancy, enabling them to remain submerged while resting. Additionally, the results indicate that sperm whales start resting dives with less diving gas volume than for deep foraging dives.
Professor Patrick Miller from the Sea Mammal Research Unit at the University of St Andrews said: "This study confirms that sperm whales exquisitely control their buoyancy by releasing gases to remain submerged with near-neutral buoyancy just below the sea surface. It is particularly fascinating they are able to make these fine-scale adjustments while they are thought to be asleep, a challenge totally alien to terrestrial mammals like us humans.”
The authors suspect that releasing bubbles could be related to the off-gassing of excess CO2 or N2 from tissues into the lungs, so this behaviour could also have implications for metabolic gas exchange.
(JULY 23, 2026) WASHINGTON, D.C. — More than half of the world's migratory bird species are declining according to a major new review led by researchers from Georgetown University, the Smithsonian’s National Zoo and Conservation Biology Institute, and a global collaboration of ornithologists.
Published in Nature Reviews Biodiversity, the paper provides the most comprehensive global assessment of migratory bird population status, threats, and conservation approaches, and presents a new framework for reversing population declines. The review synthesizes decades of research, international monitoring efforts, global biodiversity databases, and conservation agreements to identify what scientists know, where major knowledge gaps remain, and what actions are needed to reverse declines before more species become threatened with extinction.
The findings point to a broader challenge facing global biodiversity conservation. Because migratory animals, including birds, move across continents, oceans, and political boundaries throughout their annual cycles, they depend on interconnected networks of breeding grounds, migration stopover sites, and non-breeding habitats. As a result, they face a complex and often overlapping set of threats throughout their lives, making them among the most difficult groups of animals to conserve.
The authors estimate that approximately 51% of migratory bird species are declining globally across every major flyway on Earth. While nearly 300 migratory species are already considered globally threatened, many once-common and widespread birds are also experiencing substantial population losses, underscoring the scale of the biodiversity crisis.
"The scale of migratory bird declines is both alarming and sobering," said Peter P. Marra, Dean of Georgetown University's Earth Commons Institute and co-lead author of the study. "What we're seeing is the predictable consequence of decades of environmental degradation. Yet these declines are not inevitable. This review goes beyond documenting the crisis by identifying the major threats, highlighting the critical knowledge gaps, and providing a science-based framework for reversing declines before more species are pushed toward extinction."
Migratory birds are among the most visible and widespread components of global biodiversity. They help regulate insect populations, disperse seeds and pollinate plants, and support food webs across ecosystems and continents. They are also one of the primary ways people connect with nature, with millions of birdwatchers and community scientists around the world contributing observations to global databases that have become essential for understanding bird populations and informing conservation action.
Yet understanding exactly where and why populations are declining, and which of the world’s 11,000 bird species should be considered migratory, remains a challenge. Scientists are increasingly using advances in tracking technology, community science initiatives such as eBird, AI-enabled acoustic monitoring, and population modeling to identify migration routes, understand how populations are connected across seasons, and uncover the drivers of decline, including habitat loss, climate change, invasive species, collisions with buildings and infrastructure, disease, and pesticide use. The authors argue that reversing declines will require conservation strategies that address these threats across entire migratory lifecycles and international flyways rather than in isolation.
"Previous analyses have brought global attention to the fact that North American and European migratory birds are declining at alarming rates," said Nathan W. Cooper of Smithsonian’s National Zoo and Conservation Biology Institute and co-lead author of the study. “But, until now, we did not have a good assessment about how birds are doing across all of the global terrestrial and marine flyways.”
The review outlines a framework for reversing declines that includes expanding monitoring and scientific understanding, strengthening conservation social science, protecting critical habitats, reducing major threats, and increasing international cooperation. It calls for integrating ecological and social sciences to develop conservation solutions that are effective locally and scalable across countries, cultures, economies, and governance systems.
One example of this approach is Road to Recovery, a science-based initiative co-founded by Marra and now housed at Georgetown University. The initiative coordinates a hemispheric response to identify the factors limiting populations and implement targeted conservation actions before species become threatened by combining advances in science and technology with the human dimensions of species conservation.
The researchers stress that despite the concerning trends, species-focused conservation initiatives have a proven record of success. Past efforts have helped recover waterfowl, cranes, shorebirds, and raptors once on the brink of extinction. The challenge, they argue, is scaling those successes to match the geographic scope and complexity of migratory birds.
"These declines are not inevitable," Marra said. "We know conservation works. This review shows that we have the scientific knowledge needed to reverse many of these declines. The challenge now is bringing together the partnerships, resources, and political and public commitment to implement those solutions at the scale migratory birds require."
