Open Buildings 2.5D Temporal: Annual Building Presence, Count and Height Rasters from Sentinel-2, 2016-2023 (Google)
- Provider
- Google Research (Open Buildings)
- License
- CC-BY-4.0 (source)
- Format
- TIFF (in sampled files)
- Stored at
- Google Cloud Storage, gs://open-buildings-temporal-data/v1 (us)
- Access
- Free and public. No account with the data provider is needed.
- Category
- Earth Observation & Geospatial
- Source
- https://arxiv.org/abs/2310.11622
About this dataset
Annual per-pixel maps of buildings across roughly 58 million square kilometers of Africa, South Asia, South-East Asia, and Latin America and the Caribbean, for each year from 2016 to 2023. A deep learning model reads stacks of 32 cloud-free Sentinel-2 images centered on June 30 of each year and predicts three channels: building presence (an uncalibrated model confidence), fractional building count, and building height above ground (capped at 100 m; mean absolute error 1.5 m where evaluated in North America, Europe and Japan). Effective resolution is about 4 m although rasters are written at 50 cm. Because the input is free Sentinel-2 imagery rather than commercial imagery, the series can track construction and urban growth year by year. Measured 2026-11-17: v1/geotiffs/ holds one folder per tile and date (for example 00824_2016_06_30/) of GeoTIFF rasters, and v1/manifests/ holds per-UTM-zone JSON manifests per year (e.g. 01_EPSG_32723_2016_06_30.json). Read the GeoTIFFs with rasterio or GDAL, or use the Earth Engine collection GOOGLE/Research/open-buildings-temporal/v1. Sensitive areas, including conflict zones, were left out. Released by Google under CC BY 4.0 or ODbL 1.0, at the user's choice; cite Sirko et al. 2023, High-Resolution Building and Road Detection from Sentinel-2 (arXiv:2310.11622).
How to access this dataset in Python
import gcsfs
fs = gcsfs.GCSFileSystem(token='anon')
fs.ls('open-buildings-temporal-data/v1/', detail=False)[:10]Lists the files with anonymous access. Runs on any machine, with no cloud account.
Data structure
File types in sampled files: TIFF
Files
Example file: v1/geotiffs/00824_2016_06_30/tile_3FYdP6L109o.tif
Analyze this dataset without downloading it
On Scigantic, Open Buildings 2.5D Temporal: Annual Building Presence, Count and Height Rasters from Sentinel-2, 2016-2023 (Google) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.