1371 papers
This study demonstrates that Indigenous and local oral histories generally complement instrumental climate records regarding temperature trends across three North American biosphere reserves, though discrepancies often arise from differences in spatial scale, livelihood-specific sensitivities, and data limitations.
This study demonstrates that a major sudden stratospheric warming in March 2025 influenced record-breaking surface wind extremes in Eastern Asia during April 2025 by descending easterlies that reflected planetary waves to intensify the East Asian trough, thereby enhancing short-range forecast skill when stratospheric data is incorporated into deep learning models.
This study presents a national assessment framework for Canada's blue carbon ecosystems, revealing significant carbon stocks and sequestration potential while demonstrating that expanding their protection to 30% by 2030 could yield substantial cumulative climate mitigation benefits by 2050.
This study integrates cloud-based remote sensing and a Cellular Automata-Artificial Neural Network (CA-ANN) model to analyze historical and project future Land Use and Land Cover (LULC) changes in Ethiopia's Gilgel Gibe I Watershed, revealing a significant decline in forest and grassland areas alongside rapid expansion of agriculture and built-up zones that threaten ecosystem services and environmental sustainability.
This study introduces a high-resolution modeling framework to demonstrate that future urban expansion will disproportionately intensify heat hazards in already hot, densely populated urban cores across North America, revealing a critical, localized vulnerability driven by the compounding effects of climate change, demographic shifts, and spatial development.
This study reveals that edible bivalves (*Pilsbryoconcha exilis*) from Makassar's Tallo coast are contaminated with microplastics, primarily blue fibers of PE, PET, and PVC, posing significant dietary health risks—particularly to children—as evidenced by risk quotients and excess cancer risk estimates exceeding safety thresholds.
LakeRINO is a novel forecasting framework that leverages a Continuous-Time Neural Operator and diverse static and temporal features to convert irregular satellite altimetry data into accurate short-horizon lake-level predictions across China, with a demonstrated ability to specifically improve forecast precision during critical low-water anomalies.
This study of forest-dependent households and officials in Vietnam reveals a significant gap between high willingness to participate in forest-carbon pilots and the limited practical capacity to implement them, highlighting the critical need for targeted training, improved data infrastructure, and staged, performance-tested pilots before scaling up such initiatives.
This study of Indigenous Carib communities in St. Vincent and the Grenadines reveals that social infrastructure, including kinship ties and community organizations, is a more critical determinant of climate resilience and vulnerability than geographic remoteness, highlighting the need to strengthen social networks and local knowledge for effective disaster risk reduction.
This study uses a neural network trained on climate simulations to demonstrate that while only a small fraction of the world's land currently experiences temperatures permanently exceeding historical records at 1.5°C of warming, this area expands rapidly with further warming, committing a majority of the global population and most aboveground biomass to unprecedented local temperatures under current policy trajectories by 3°C.