Abstract Volcanic ash is a significant hazard to aviation due to its potential to cause physical impacts and disrupt aviation networks through airspace and airport closures. Until now, ashfall impacts on airport operations have received much less attention than airborne ash impacts on aircraft and aviation. The time and resources taken to remove ashfall from runways and airport infrastructure can cause extended disruption to the aviation network beyond the presence of ash in the atmosphere. Using publicly available information, we have compiled a global dataset of 334 volcano-related airport impact events between 1944 and 2024. The dataset includes events at 141 unique airports in 44 different countries and territories, caused by eruptions from 73 different volcanoes. The countries with the most events are Indonesia (52), Mexico (27), and New Zealand (27), and the most impacted airports are Catania-Fontanarossa International Airport (20), Puebla International Airport (15), and La Nubia Airport (12). The volcanoes responsible for the most airport impact events (besides Eyjafjallajökull) are Etna (24), Soufriere Hills (22), and Popocatépetl (22). We used 74 data-rich, ashfall-induced events from this dataset to categorise ashfall impacts into five states of closure: <1 day, 1–2 days, > 2–7 days, > 1 week (finite), and permanent, and present fragility curves for each state. Data points are concentrated between 0.1- and 100-mm thickness such that fragility curve uncertainties are greatest at the smaller (< 0.1 mm) and larger (> 100 mm) thicknesses. The curves show that there is an > 80% probability of airport closure of any duration even for trace (defined here as 0.1 mm) amounts of ashfall. Closures of more than one day under trace ashfall are much more probable than closures lasting more than two days (30% vs. 6%, respectively), and once ashfall of 10 mm thickness is reached closures of more than two days are highly probable (> 85%). The development of continuous fragility curves from empirical airport closure data represents an advance that will prove useful for hazard management and long-term forecasting closure at airports. To illustrate the potential impacts of varying closure durations on flights and passengers, we provide a case study of the simulated closure of Ninoy Aquino International Airport in the Philippines. In addition to ashfall thickness, closure duration is affected by airport or national emergency management policies, resources available (personnel and equipment), and environmental conditions. This dataset and the derived curves provide a starting point and global evidence base for better understanding the impact of volcanic ashfall on airport operations.
BACKGROUND:Enlicitide decanoate, an oral proprotein convertase subtilisin-kexin type 9 (PCSK9) inhibitor, was shown to reduce low-density lipoprotein (LDL) cholesterol levels in a phase 2 trial; longer-term data are needed. METHODS:In this multinational, double-blind, randomized, placebo-controlled trial, we enrolled adults with a history of a major atherosclerotic cardiovascular disease event with an LDL cholesterol level of 55 mg per deciliter or higher and those who were at risk for a first atherosclerotic cardiovascular disease event with an LDL cholesterol level of 70 mg per deciliter or higher. Participants were assigned in a 2:1 ratio to receive enlicitide at a dose of 20 mg or placebo daily for 52 weeks. The primary end point was the mean percent change in LDL cholesterol level from baseline to week 24. Key secondary end points were the mean percent change in LDL cholesterol level at week 52 and the mean percent change in levels of non-high-density lipoprotein (non-HDL) cholesterol and apolipoprotein B and the percent change in lipoprotein(a) level at week 24. RESULTS:Of the 2909 participants in the intention-to-treat population, 1935 received enlicitide and 969 received placebo (5 did not receive enlicitide or placebo). The mean age of the participants was 63 years, and 39.3% were women. The mean (±SD) LDL cholesterol level at baseline was 96.1±38.9 mg per deciliter. The mean percent change in LDL cholesterol levels at week 24 was -57.1% (95% confidence interval [CI], -61.8 to -52.5) with enlicitide and 3.0% (95% CI, 0.9 to 5.1) with placebo, representing an adjusted between-group difference of -55.8 percentage points (95% CI, -60.9 to -50.7; P<0.001). The mean percent change in LDL cholesterol level at week 52, the mean percent changes in non-HDL cholesterol and apolipoprotein B levels at week 24, and the percent change in lipoprotein(a) levels at week 24 were significantly greater with enlicitide than with placebo (P<0.001 for all comparisons). The incidence of adverse events did not appear to differ between the groups. CONCLUSIONS:Among participants who had a history of or were at risk for a first atherosclerotic cardiovascular disease event, treatment with the oral PCSK9 inhibitor enlicitide resulted in significantly lower LDL cholesterol levels than placebo at 24 weeks. (Funded by MSD [Rahway, NJ]; CORALreef Lipids ClinicalTrials.gov number, NCT05952856.).
As the global population ages, the prevalence of cognitive impairment continues to rise, highlighting the urgent need for accessible and low-burden cognitive assessment. Current assessments based on clinical scales are often hindered by subjectivity, significant user burden, and practice effects, limiting their applicability. We observe that passive visual stimuli can engage multiple cognitive domains while minimizing the need for active participation. Considering the lack of related datasets, we establish EasyCog, the first large-scale multimodal dataset for low-burden cognitive assessments. EasyCog collects synchronized forehead/ear EEG and contactless eye tracking data while participants passively view a short, cognitively structured video followed by an eyes-closed rest. The dataset includes 101 participants spanning healthy controls and patients with PD, AD, and VaD, with clinician-administered MoCA/MMSE scores collected in daily settings. We provide detailed collection procedures, quality validation, implementation, and benchmark baselines. Results indicate assessment feasibility while highlighting generalization challenges. By integrating passive visual stimuli with affordable sensing, EasyCog provides a foundation for future research in accessible and scalable cognitive monitoring in both clinical and community settings.
We survey Small Business Investment Companies (SBICs) to perform a novel analysis of their performance. SBIC funds outperform comparable non-SBIC peers by around 2% to 3% as measured by internal rate of return and about 0.3x to 0.7x as measured by multiple on invested capital, depending on benchmark deployed. To mitigate sample selection bias, we also examine SBICs in the MSCI Private Capital Universe data, which shows similar, but smaller, outperformance. We analyze SBIC funds by fund strategy and amount of leverage utilized to provide a granular view of risk-adjusted performance. We believe this to be the first large-sample analysis of SBIC returns.
Large Language Models (LLMs) struggle with the complex, multi-modal, and network-native data underlying financial risk. Standard Retrieval-Augmented Generation (RAG) oversimplifies relationships, while specialist models are costly and static. We address this gap with an LLM-centric agent framework for supply chain risk analysis. Our core contribution is to exploit the inherent duality between networks and knowledge graphs (KG). We treat the supply chain network as a KG, allowing us to use structural network science principles for retrieval. A graph traverser, guided by network centrality scores, efficiently extracts the most economically salient risk paths. An agentic architecture orchestrates this graph retrieval alongside data from numerical factor tables and news streams. Crucially, it employs novel ``context shells'' -- descriptive templates that embed raw figures in natural language -- to make quantitative data fully intelligible to the LLM. This lightweight approach enables the model to generate concise, explainable, and context-rich risk narratives in real-time without costly fine-tuning or a dedicated graph database.
