geospatial

The author has built Global Fail Map, an open-source interactive tool documenting historical failures in science, technology, and business to serve as a research layer for AI agents and developers.

The blog post describes a step-by-step process to create aerial maps in under 30 minutes using a DJI Lito X1 drone and Zeitgeist Survey, highlighting automation and efficiency in flight planning and image processing.

This tutorial guides users on making their first edit to OpenStreetMap by adding a website tag using the JOSM editor in under 15 minutes.

Planet Labs has announced an open satellite feed, providing public access to its satellite imagery data.

Reactor Atlas is an interactive map and database that catalogs all nuclear reactors on Earth since 1954, featuring technical records, animated explanations, and real-time monitoring of outages and events.

This content shares a satellite real-time 3D map tool that allows users to view detailed roof and geographical information.

The State of the Map 2026 is the annual conference for OpenStreetMap users, to be held in Paris and online from August 28-30, 2026, featuring talks, workshops, and social events.

JupyterGIS 0.16 introduces enhanced geospatial visualization with collaborative Story Maps, support for new formats like openEO, and improved integration with the Python scientific ecosystem.

City2Graph is a new Python library that converts geospatial data into heterogeneous graphs for spatial analysis and Graph Neural Networks, with a peer-reviewed paper published in Computers, Environment and Urban Systems. It supports morphological, transport, mobility, and proximity graph construction, integrating with PyTorch Geometric and other graph tools.

Presents Click2Poly, a human-in-the-loop Vision Language Model based on Florence-2 that speeds up manual editing of building and wall vector layers, implemented as a QGIS plugin.

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.

This paper measures the implicit assumptions language models make about 'a city' by scoring anonymized urban profiles across 40 indicators, finding a shared preference for larger, faster-growing, and more infrastructure-rich cities. It uses open-weight checkpoints and replication data to make the default portrait of cities in LLMs empirically traceable.

Introduces GEOID-Flood, a large-scale multi-modal benchmark dataset for flood segmentation with over 14,000 tiles from 219 events across 65 countries, evaluating foundation models against conventional encoders across single-image, multi-temporal, and multi-modal protocols.

A blog post describes computing the Pole of Inaccessibility in the San Gabriel Mountains using a Voronoi diagram approach with OpenStreetMap road and trail data, and shares the resulting code in a GitHub repository.

The 2026 Cesium Developer Conference session recordings are now available, featuring talks on geospatial 3D, AI, 3D Tiles 2.0, and a keynote by SQLite creator D. Richard Hipp.

Ai2 introduces the OlmoEarth Platform, an infrastructure for running geospatial AI inference at continental scale, enabling applications like deforestation monitoring and wildfire risk using their family of Earth observation foundation models.

This paper presents a Graph Attention Network (GAT) approach to model spatial dependencies in soil samples for predicting microplastics and organic matter, achieving high R² values but limited cross-validation generalization due to small sample size.

Delineate Anything v2 is a globally scalable foundation model for agricultural field boundary mapping, outperforming state-of-the-art by 103.3% relative gain in [email protected], using a 73-million-instance multi-resolution dataset spanning 61 countries and a manually curated 100-country evaluation benchmark.

The New York City Office of Technology and Innovation publishes a weekly updated dataset of building footprints for all 1.08 million buildings in NYC, providing a detailed and current geospatial foundation for city data.