New research reveals that ‘foundation models’ trained on vast, general time‑series data may be able to forecast river flows accurately, even in regions with little or no local hydrological records. The approach could improve flood warnings, drought planning and water-resource management in parts of the world where monitoring data is limited.

Swedish old-growth forests store 83 percent more carbon than managed forests, according to a new study from Lund University. The difference is substantially larger than previous estimates and is mainly due to large carbon stocks in the soil.

Following the first independent investigation in fifty years of Monte Verde – a landmark archaeological site in Chile – researchers report it may be much younger than previously believed. According to the study, Monte Verde dates from ~8000 to 4000 years old, not 14,500 years, as previously thought. The findings reshape the story of the continent’s first settlers (though they don’t rule out pre-Clovis human presence in South America, as supported by other sites); they also highlight the need for independent verification of old archeological sites. Monte Verde is one of the most important archaeological sites for understanding when humans first reached South America, the last continent colonized by humans. Excavations of the site’s Monte Verde II component uncovered stone tools and well-preserved organic materials, such as wooden artifacts, cordage, and fossil remains of extinct Pleistocene fauna. Earlier dating suggested that the site was occupied roughly 14,500 years ago, making it nearly 1,500 years older than the Clovis culture – the once-dominant benchmark for earliest human settlement in the Americas. Although widely accepted as key evidence of a pre-Clovis human presence in southern South America, the findings from Monte Verde have long been debated. Critics have questioned whether the artifacts, sediment layers, and radiocarbon dates are truly associated, raising possibilities such as redeposited ancient material or dating inaccuracies that could exaggerate the site’s age.

Now, Todd Surovell and colleagues show that the antiquity of Monte Verde may have been overestimated. Surovell et al. reexamined the age and geological context of Monte Verde II by describing, sampling, and dating nine sediment exposures along the banks of the nearby Chinchihuapi Creek. The analysis shows that the abandoned floodplain on which the site resides is far more complex than previously understood. According to the findings, the area contains layers of sand and gravel deposited by glacial meltwater between about 26,000 and 15,500 years ago, followed by deposits of ancient wood, marsh sediments, and a volcanic ash layer identified as the regionally widespread Lepúe Tephra, which is well-dated to roughly 11,000 years ago. The authors argue that because the floodplain deposit containing the archaeological site sits above this ash layer, it must be younger than 11,000 years. Further radiocarbon dating of wood and peat from the floodplain sediments produced ages between ~8200 and 4100 years, indicating that the deposit formed during the Middle Holocene. The authors suggest that earlier dates previously reported for the site were likely influenced by Late Pleistocene-age materials from older sediments that were redeposited into the site via erosion. The findings suggest that Monte Verde II is Middle Holocene in age or younger, challenging earlier interpretations that placed the site much earlier in the late Ice Age.

In a Perspective, Jason Rech discusses the study as well as the implications of the findings. “Although Monte Verde grounded chronologies for early colonization of the Americas for decades, the landscape is different now, with more sites that appear to be older than the Clovis culture,” Rech writes. “Yet as Surovell et al. conclude, their findings highlight the need for independent verification of old archeological sites.”

The planet’s undisturbed old-growth boreal forests may be far more important in the fight against climate change than previously realized, according to a new study, which finds that primary forests in Sweden store over 70% more carbon than managed secondary forests. Boreal forests, the world’s largest forest biome, play a crucial role in absorbing roughly 30% of human-caused carbon dioxide (CO₂) emissions. Yet they are increasingly exploited to supply timber and bioenergy. In European boreal forests, intensive management practices, such as clear-cutting, thinning, planting fast-growing trees, fertilization, drainage, and soil preparation, aim to maximize harvestable wood and timber quality, but their long-term impacts on carbon storage remain uncertain. Limited observations from undisturbed primary forests make it difficult to quantify how converting these forests into managed secondary forests affects carbon sequestration. With models projecting increased reliance on northern forests for bioenergy, understanding the effects of transforming primary boreal forests and their impact on carbon storage is essential.

Didac Pascual and colleagues combined data from the Swedish National Forest Inventory (NFI), the Swedish National Forest Soil Inventory, and targeted field surveys to estimate carbon storage in Sweden’s primary forests and to quantify how it differs from the region’s managed secondary forests. They measured carbon across multiple components, including vegetation, dead wood, soils, and harvested wood products, and applied multiple analytical methods to estimate total carbon storage. Pascual et al. found that primary forests stored about 72% more carbon than managed secondary forests when considering all carbon pools combined. Soils contained the largest carbon store and accounted for much of the difference between the forest types. Overall, across Sweden, primary forests store 9.9 kg carbon per square meter more than managed secondary forests – 2.7 to 8 times higher than earlier estimates. This sugegsts that the climate impact of converting primary boreal forests to managed forests may be much greater than previously thought.