A study proposes an operational framework combining machine learning models with Sentinel-2 data to estimate agricultural drought conditions in irrigated and non-irrigated maize fields. Deep Neural Network (DNN) achieved the best performance, showing higher prediction accuracy and lower error metrics for non-irrigated fields.
A new study proposes a switching-based pervasive augmented reality framework that integrates location-based AR, deep learning, and context-awareness to improve landmark recognition. The framework significantly enhanced detection accuracy compared with conventional LBAR systems, while demonstrating high user satisfaction.
A new catalog centralizes publicly available soil moisture content datasets, streamlining data discovery and comparison. The UB-SMDC portal provides a unified platform for researchers to efficiently access and analyze global soil moisture data.
Digital twins are expanding rapidly, using real-time data to simulate and analyze systems before applying them in the real world. The study emphasizes the need for interoperable architectures, machine-readable metadata, and standardized trust frameworks to address challenges such as privacy, cybersecurity, and uncertainty.
The study systematically evaluates popular Uncertainty Quantification (UQ) methods and metrics for AI/ML-based geospatial applications. It demonstrates that Deep Ensembles and Bayesian Neural Networks achieved the best performance, while highlighting framework-specific differences between TensorFlow and PyTorch.
A new study introduces a free, open-source Web Crop Phenology Metrics Service (WCPMS) for analyzing large Earth observation satellite time series, enabling efficient crop monitoring. The tool calculates phenological metrics from data cubes and validates its effectiveness using soybean sowing dates in Brazil.
The CA-MTransU-Net architecture achieves superior mIoU of 87.00% and faster inference speeds compared to benchmark algorithms, providing a scalable solution for post-fire damage assessment in cloud-prone landscapes.
The Big Earth Data journal is launching a special issue to reflect on its decade-long impact and showcase cutting-edge advancements in big data research. The journal focuses on Earth-related big data, emerging as a flagship platform at the intersection of Earth science, space science, information science, and sustainability science.
A new axis-based data model resolves long-standing issues in Earth data grids, enabling more accurate, flexible, and interoperable data across science, policy, and industry. The framework clarifies grid structure, coordinate handling, and value interpretation, allowing for efficient querying of massive multidimensional datasets.
A new study evaluates the accuracy of ten satellite-based and reanalysis precipitation datasets using in-situ rain gauge measurements across Czechia from 2001 to 2021. The GSMaP dataset showed superior performance for rainy days, while ERA5-Land overestimated annual precipitation by 15–35%.
The study presents a comprehensive roadmap for building an integrated GGBEO to meet the United Nations' Sustainable Development Goals and advance climate science. The system would integrate regional and global ground-based in situ and remote sensing systems, marine, and airborne observational data.