earth-observation

Allen AI 发布 OlmoEarth v1.3 卫星嵌入模型，可为全球每个 10 米像素生成 128 维向量，并同时公开全球预计算嵌入数据与浏览器端嵌入探索工具，在 27 项评测中 20 项取得最佳成绩。

Google DeepMind introduces ForTy, a system that classifies every 10x10m patch of land on Earth into distinct forest and tree-crop types, highlighting how untouched natural forests store far more carbon and support more biodiversity than single-species timber plantations.

Google DeepMind, Google Research, WRI and IIASA released Forest Typology (ForTy) 2020, a 10-meter global map classifying every land patch into six forest and land-use types, plus a 2017–2025 time series showing what replaced natural forests, supporting the EU Deforestation Regulation.

ESA and MistralAI have signed a letter of intent to strengthen cooperation on artificial intelligence, focusing on developing secure and trusted European AI solutions for Earth observation, engineering, and next-generation space activities.

Albania has become the 73rd country to sign the Artemis Accords, committing to peaceful and transparent space exploration in partnership with NASA and other nations.

This paper proposes a machine learning API specification for openEO to integrate ML workflows into Earth Observation data cubes, enhancing reproducibility and cross-platform interoperability. It demonstrates feasibility with prototype implementations in R and Python and use cases in crop type mapping and deep learning.

This research paper examines worst-case scenarios for glacial lake outburst floods in transboundary Himalayan basins, assessing current and future hazards through integrative approaches and multiple scientific studies.

This paper proposes a task-driven three-layer distributed scheduling method for emergency Earth observation in large low-Earth-orbit constellations, achieving higher emergency coverage and lower routine-plan disruption compared to existing methods.

OlmoEarth Studio introduces support for computing and exporting custom embedding vectors from OlmoEarth foundation models, enabling downstream tasks like similarity search, segmentation, and change detection. The embeddings are available as Cloud-Optimized GeoTIFFs through the Studio UI or API.

GeoForge is a training-free, self-evolving framework for Earth-observation reasoning that structures completed trajectories into nonparametric memories to improve LLM agent planning and tool-use without updating the backbone model.

Europe's Copernicus Browser now includes wildfire tracking as a default visualization, making higher-resolution satellite imagery easier to access for monitoring fires and smoke.

Introduces Obshazard-bench, a real-time, observation-driven benchmark for evaluating multimodal foundation models on disaster intelligence from raw Earth observation streams, spanning 8 disaster categories across 60+ countries.

Introduces OVEarth-Bench, a benchmark for open-vocabulary Earth observation that broadens category coverage and query diversity, revealing that current methods remain limited and MLLM-based approaches perform best.

Array Labs raised a $21M strategic round led by Mitsubishi Electric to scale its cluster of small, mass-manufacturable radar satellites that provide real-time tracking of ships, aircraft, and missiles, with existing contracts from multiple U.S. defense agencies.

This paper presents a systematic comparison of two geospatial foundation models, TerraMind and THOR, developed under ESA's φ-lab, analyzing how architectural choices like patch size and decoder type affect performance across ten use cases in Earth observation tasks.

The first three operational FireSat satellites, backed by Google and Bezos Earth Fund, launched to detect wildfires as small as 5x5 meters, aiming to provide hourly global coverage by 2029.

This paper surveys the emerging paradigm of Geospatial Foundation Models (GeoFMs), which are pre-trained on massive geospatial datasets to enable rapid fine-tuning and zero-shot analysis of satellite and aerial imagery. It covers the paradigm shift, model adaptation strategies, and a forward-looking vision of Agentic Geospatial Reasoning using LLMs as orchestrators.

This chapter reviews design principles and current landscape of foundation models for Earth observation, highlighting the need for domain-specific adaptation, physically plausible representations, and consistent evaluation benchmarks. It includes case studies on harmful algal bloom prediction and adaptive monitoring station selection.

The paper presents the largest controlled scaling study for Earth-observation foundation models, showing that pretraining loss poorly predicts downstream performance and providing an optimal compute allocation rule. It trains scaled pixel-wise models (0.5B and 1B parameters) and distills them into compact student models that outperform larger open and proprietary models.

EO-Agents presents a three-agent LLM pipeline for generating Earth observation hypotheses, leveraging a NASA knowledge graph and graph neural network to rank candidate dataset pairings, with LLM agents filtering, generating, and evaluating structured research hypotheses.