1371 papers
SimCast-S2S is a novel latent-diffusion framework that leverages transfer learning from climate simulations to efficiently generate probabilistic subseasonal precipitation forecasts, outperforming deep learning baselines and competing with state-of-the-art operational systems while operating in a compact latent space.
Using a quasi-experimental design across 11 sub-Saharan African countries, this study finds that providing mothers with access to free primary education significantly reduces the risk of child wasting caused by climate-related disasters, suggesting that education reforms can serve as a vital strategy for enhancing climate resilience.
This study utilizes macro-micro experimental techniques to demonstrate that dry-wet cycles, particularly under water saturation, exponentially degrade coal's mechanical properties by altering pore connectivity and inducing mineral dissolution, thereby offering critical insights for preventing dynamic mining disasters.
This study utilizes multi-regional input-output modeling to analyze the spatiotemporal characteristics and driving mechanisms of embodied carbon emissions in China's municipal construction industry, revealing that gross fixed capital formation and resident consumption are primary drivers while emissions intensity and input-output structure serve as key factors for emission increases and reductions, respectively.
This study analyzes 1970–2019 data to reveal significant regional variations in extreme rainfall frequency and intensity across Indian states, demonstrating a clear link between these events and human fatalities, particularly in river-basin-dominated and densely populated regions like West Bengal, Bihar, Uttar Pradesh, and Maharashtra, thereby highlighting the urgent need for climate-resilient adaptation and improved disaster management.
This study evaluates groundwater recharge potential in Banda Aceh-Indrapuri, Indonesia, by integrating GIS, remote sensing, and magnetotelluric data to demonstrate a strong negative correlation between recharge indices and subsurface resistivity, thereby establishing a robust framework for regional water management in complex tropical settings.
This study introduces an Ensemble Machine Learning framework to generate high-resolution global projections of Arable Equipped for Irrigation (AEI) under Shared Socioeconomic Pathways (SSPs), revealing that future irrigation capacity is likely to decline in most scenarios due to economic and infrastructural constraints rather than theoretical food demand, thereby providing critical data for Earth system modeling and food security assessments.
This study reveals that urban waste mismanagement, rather than industrial activity, is the primary source of widespread microplastic contamination in Ethiopian irrigated farmlands, highlighting an urgent need for improved municipal waste collection, stricter plastic bans, and farmer education to mitigate this emerging environmental threat.
This study proposes a citizen science-based framework for monitoring coipo (Myocastor coypus) in Chilean urban wetlands, evaluating various methods to recommend structured sign-based approaches supported by public observations and expert validation as the most feasible and scalable solution for long-term biodiversity monitoring.
This study utilizes satellite data to demonstrate that the 2015–2016 El Niño caused significant sea surface warming and a 54% decline in chlorophyll-a concentrations in southern Bali waters, driven by intensified vertical stratification that suppressed nutrient upwelling and phytoplankton productivity.