Scientific Data最新文献

After a family murder in rural Germany, authorities failed to locate the suspect in a vast forest despite a massive search. To aid the search, a research aircraft captured high-resolution aerial imagery. Due to dense vegetation obscuring small clues, automated analysis was ineffective, prompting a crowd-sourcing initiative. This effort produced a unique dataset of labeled, hard-to-detect anomalies under occluded, real-world conditions. It can serve as a benchmark for improving anomaly detection approaches in complex forest environments, supporting manhunts and rescue operations. Initial benchmark tests showed existing methods performed poorly, highlighting the need for context-aware approaches. The dataset is openly accessible for offline processing. An additional interactive web interface supports online viewing and dynamic growth by allowing users to annotate and submit new findings.

The Quantitative Neuroimaging Laboratory Dataset provides magnetic resonance imaging (MRI) modalities and two resting-state and twelve task-based functional MRI (fMRI) tapping into four cognitive domains (episodic memory, fluid reasoning, processing speed, and crystallized memory). It also includes three positron emission tomography (PET) scans ([18 F]Fluorodeoxyglucose (FDG), Florbetaben, and MK-6240), plus neuropsychological assessments, and vital signs. Currently, 356 participants consented (97 young: 20 ~ 40 years; and 259 elderly: 60 ~ 80 years), while 259 completed at least one scan. We uploaded 4688 MRI/fMRI and 719 PET scans (232 Florbetaben, 251 FDG, and 236 MK-6240). 189 participants completed all scan modalities. All imaging underwent an in-house, pre-processing pipeline developed for each modality. This dataset aims to characterize the spatial and temporal properties of the brain's hemodynamic response in the opposite direction (i.e., brain deactivation), its task dependency, and its interaction with the brain's large-scale functional connectivity networks. Ultimately, this will enable the translation of neuroimaging findings into personalized medicine approaches that better characterize and predict individual pathologies in neuropsychiatric diseases.

Stable isotope analysis has long been a key approach for reconstructing ancient human and animal diets and mobility in archaeology. Over the past four decades, and particularly during the past decade, isotopic datasets in China have expanded rapidly, creating a clear need for timely updates to comprehensive compilations. Existing compilations primarily focus on human bone carbon and nitrogen isotope data, while omitting animal and plant multi-isotope datasets, as well as strontium isotope records. In this study, we present the Isotope Dataset for Archaeological Biological Remains in China, the most comprehensive dataset to date. It compiles nearly 20,700 isotope measurements (1984-2026) spanning the Palaeolithic to historical periods across major cultural regions. The dataset integrates multi-isotope data (δ¹³C, δ¹⁵N, δ¹⁸O, δ³⁴S, and ⁸⁷Sr/⁸⁶Sr) from human and animal tissues (e.g., bone, dentine, enamel, hair) and plant remains, alongside archaeological, chronological, and geographic data. Chronological control is provided by radiocarbon dates, cultural chronology, and stratigraphic information; quality indicators (collagen yield, %C, %N, %S, C/N ratios, C/S ratios, N/S ratios) are recorded where available. Hosted on the open-access Zenodo platform, the dataset provides an up-to-date resource for cross-study and cross-regional comparisons of archaeological isotope data in China, supporting future sampling strategies.

Accurate information on crop types and irrigation systems is a fundamental input for improving our understanding of land-atmosphere interactions and monitoring agricultural dynamics in semi-arid environments. This paper presents a comprehensive, open-access ground-truth dataset of 10000 geolocated agricultural parcels collected across the five major agricultural regions of Morocco, El Gharb, Tadla, Doukkala, El Haouz, and Souss, between December 2023 and January 2025. The dataset encompasses 45 distinct crop types, including main seasonal crops such as wheat, corn, and alfalfa, as well as perennial crops including olives, citrus, and argan. In addition, six different irrigation systems were observed, from the traditional flood irrigation to advanced pivot systems, reflecting the diversity of Moroccan agro-ecosystems. Field data were collected using the GIS-based mobile application QField, ensuring high positional accuracy and detailed attribute information for each parcel. To evaluate dataset consistency, Normalized Difference Vegetation Index (NDVI) values were analyzed for all surveyed plots throughout 2024, revealing coherent seasonal dynamics in agricultural activity. Each record is accompanied by a geo-referenced photograph, providing an additional visual reference for validation and contextual understanding. The freely available dataset offers a high-quality reference for calibrating and validating Earth Observation (EO)-based crop type and irrigation mapping products. By representing the diversity of cropping systems and irrigation practices across Morocco's climatic zones, this dataset provides a valuable foundation for advancing EO-driven agricultural monitoring, supporting policy assessment, and promoting sustainable land and water management in Morocco and the wider North African region.

Carbon stock and carbon accumulation rate data are vital to multiple aspects of tidal wetland conservation and restoration policy. In California, USA tidal soil data are rare outside of the San Francisco Bay and Sacramento Delta regions, despite the differing conditions experienced by the outer coastline. Here we provide carbon stocks and decadal-to-centennial-scale carbon accumulation rate calculations. This dataset presents 83 soil depth profiles from 15 sites, with 58 cores from 12 tidal wetland sites analyzed for carbon stock, mostly from the outer coastline of California. Mean organic matter content was 11%, and stocks estimated to 1 meter depth ranged from 15.4 to 44.7 kgC m^-2. Organic matter content generally declined asymptotically with depth. Carbon accumulation rates ranged from 39.2 to 130.0 gC m^-2 yr^-1. Neither carbon stock nor carbon accumulation rates were notably different from global average values. Data at this level of reporting are vital for establishing restoration baselines, informing greenhouse gas mitigation planning, and projecting future ecosystem response to sea-level rise.

Mental health issues such as stress and anxiety are highly prevalent among university students, often affecting their academic performance and overall well-being. Understanding these conditions through objective, real-world data is essential for developing effective monitoring and intervention strategies. We present a multimodal dataset that captures students' daily stress and anxiety levels through self-reports and wearable sensor data. The dataset was collected during one academic semester (February-July 2025) from undergraduate volunteers at two Mexican universities. Participants provided daily ratings of stress and anxiety using a mobile application, while Fitbit Inspire 3 devices continuously recorded physiological and behavioral data including heart rate variability, sleep quality, oxygen saturation, stress score, physical activity, and step count. The dataset features over 80% questionnaire compliance and validated Fitbit measurements. This dataset addresses the scarcity of public, ecologically valid datasets on student mental health and enables reproducible research and analyses in affective computing, wearable sensing, and machine learning for stress and anxiety monitoring.

The mid-Norwegian margin is one of the best studied volcanic rifted margins on Earth. Geophysical investigations have demonstrated the presence of well-developed inner and outer Seaward Dipping Reflectors (SDRs), landward flows, lava deltas, marginal highs, volcanic centers, ash layers, and sill complexes. These features have been proven to consist of magmatic rocks through the international Deep Sea Drilling Program (DSDP Leg 38, 1974), Ocean Drilling Program (ODP Leg 104, 1985), International Ocean Discovery Program (IODP Expedition 396, 2021), and commercial drilling. A total of fifteen drill cores penetrated magmatic rocks that formed between 57 and 50 million years ago (Ma). Here we provide (i) new (n = 224) major and trace element compositions obtained by X-ray fluorescence (XRF), inductively-coupled plasma mass spectrometry (ICP-MS), and inductively-coupled optical emission spectrometry (ICP-OES) on whole rock powders of magmatic rocks for IODP Exp. 396 (n = 119), ODP Exp. 104 (n = 79), DSDP Exp. 38 (n = 24); and (ii) a compilation of all new and published data for magmatic rocks in the fifteen drill cores (n = 563). Portable X-ray fluorescence (pXRF) data (n = 381) for the IODP Exp. 396 cores are also reported. These datasets provide a resource for examining the origin of magmatism associated with continental breakup and rifted margin formation, particularly the formation of excess magmatism compared to normal mid-oceanic spreading ridges, mantle-crust interaction, and the linkage of magmatism to global hyperthermal events on Earth's surface.

Historic data are essential for quantifying changes in biodiversity and their drivers. Much legacy data remains undigitized in archives, such as natural history museums, representing untapped resources for assessing past biodiversity patterns. In 1904, the first systematic survey of species distributions of the entire Danish vascular flora was initiated. Field surveys were essentially conducted in the early decades of the 20^th century. Results were published as black and white printed maps (1931-1976). We used automated georeferencing and image classification to extract data from all published maps. We present 1340 historic distribution maps of vascular plant species, subspecies and varieties in Denmark at 10 × 10 km resolution. These maps represent the Danish flora of native and archaeophyte species known at the time, except a dozen ubiquitous native species. By comparison with modern distribution data, long-term increases and declines can be accurately quantified, contributing to large-scale analysis of national, European or global plant diversity changes in the modern era.

Today, an increasing number of applications in domains such as cultural heritage, healthcare, education, entertainment, and fashion require high-fidelity 3D avatars. However, generating avatars that faithfully reproduce users' bodies through modeling or acquisition techniques remains challenging and time-consuming, particularly in applications where the accurate quantitative reproduction of body shape and precise anthropometric measurements is required. Thus, attention is shifting towards machine learning-based approaches, in particular those able to fit a parametric model representing the avatar to the intended body shape. Among these models, the Sparse Unified Part-Based Human Representation (SUPR) has been proven to offer superior performance compared to other representations. However, its adoption is primarily hindered by the lack of datasets built upon it. This paper addresses this gap by proposing BOdy shape parameter and 3D meshes of Individuals basEd on SUPR (BODIES), a dataset containing 84,000 synthetic-generated subjects described using the SUPR model with different numbers of parameters. The paper also presents the results of three experimental studies aimed at assessing the improvements brought by the SUPR model over the state-of-the-art when used to feed an existing framework for generating 3D avatar meshes.

Surface roughness is a critical parameter in machining, as it affects the corrosion resistance and fatigue properties of the finished workpiece. For difficult-to-machine materials such as Inconel-625, ensuring desired levels of surface roughness is even more challenging. Accurate surface roughness estimation for such expensive alloys can help in reducing material wastage and enhancing machining efficiency. In this paper, machining data collected during the turning of a particularly difficult-to-machine alloy, Inconel-625, is presented. Inconel is widely used in aerospace applications and is difficult to machine, as unlike other materials, Inconel does not get softer with increasing temperature. This dataset comprises twenty-seven sets of vibration data collected using a triaxial accelerometer and corresponding force and moment data collected using a dynamometer, resulting in 382,189,197 samples in total, acquired during the dry turning of Inconel-625 on a Kirloskar Turnmaster 40 Lathe. A Mitutoyo Surface Roughness Tester was used to measure the surface roughness after each turning operation. This publicly available dataset will be of help to the scientific community in developing machine learning/deep learning based on-line surface roughness estimation models for turning processes.