Author

Imago Team

Maintainer

Imago RSE Team

License

CC-BY-NC-4.0

Version

1.0.0

Type

Dataset

Created

January 25, 2026 at 12:29 PM UTC

Modified

January 29, 2026 at 3:15 PM UTC

Resources

4

Google Satellite Embedding V1 - London LSOAs (2020, 2024)

Beta version

Resources

uk_lsoa_london_embeds_2020.gpkg

uk_lsoa_london_embeds_2024.geojson

uk_lsoa_london_embeds_2024.gpkg

uk_lsoa_london_embeds_2020.geojson

Metadata

Constraints None
Content This dataset provides Google Satellite Embedding values (64-dimensional, L2-normalised vectors, with values ranging between −1 and +1) across the 64 original dimensions, aggregated to small-area geographies — Lower Layer Super Output Areas (LSOAs) for Greater London. The source data are Google Satellite Embedding V1, processed at 10-metre spatial resolution from the annual products (2020 and 2024). Aggregation was carried out using a geospatial pipeline that exported the tiled embedding data to Google Cloud Storage and then aggregated them to 2021 LSOA boundaries by computing the binary mean of all pixels within each LSOA in Google Colab. Two timeframes were extracted for the training purposes at Imago-2026 Geospatial Embeddings Workshop to illustrate temporal dynamics.
CRS EPSG:27700, OSGB36/British National Grid
Data quality Embedding values are derived from Google Satellite Embedding dataset V1 which provides global coverage; this product focuses on London. The original data was extracted as GeoTIFF and reprojected to EPSG:27700. The extracted GeoTIFFs were rescaled by a factor of 32767 to store values as Int16. This introduces only minimal quantisation error (mean absolute error ≈ 10⁻⁶). The further details on the methodology will be given in the next data product documentation for the entire UK. It is important to note that the underlying input data has a spatial resolution of 10-metres, and embedding values are aggregated to small area boundaries, which vary substantially in geographic area. Users may consult the Google Embedding introduction dataset (https://medium.com/google-earth/ai-powered-pixels-introducing-googles-satellite-embedding-dataset-31744c1f4650) for further information on applications and limitations in the source imagery. **IMPORTANT** - This version is released as beta. This means that, although we have applied our quality standard in its production, we are still assessing its overall quality and identifying minor issues that need resolution before a more stable version can be released. As soon as we are ready, we will issue a new release. Please check https://imago.ac.uk for news and updates.
Lineage The underlying methods and source information used to construct the dataset are documented in the upcoming technical report. LSOA-aggregated Embeddings are derived from the annual AlphaEarth Foundations Satellite Embedding dataset, which is produced by Google and Google DeepMind. The dataset was accessed via Google Earth Engine API (https://developers.google.com/earth-engine/datasets/catalog/GOOGLE_SATELLITE_EMBEDDING_V1_ANNUAL). It covers two periods: from 1 January 2020 to 31 December 2020 and from 1 January 2024 to 31 December 2024. Data is processed at 10-metre spatial resolution, reprojected from UTM to EPSG:27700 and stored on Google Cloud Storage. The extracted tiled GeoTIFF was processed through the internal Google Colab pipeline with zonal statistics for each LSOA. Dataset for 2024 year also features ‘population’ column, describing population count in LSOAs from ONS . All processing was conducted by the Imago Team.
Source Imago: Data Service for Imagery
Spatial coverage Greater London, United Kingdom
Spatial resolution Small area (LSOA / Data Zone / Small Area)
Temporal coverage 01-01-2020:31-12-2020, 01-01-2024:31-12-2024
Temporal resolution annual

Activity stream

Modified: January 29, 2026 at 3:15 PM UTC

Modified: January 29, 2026 at 3:14 PM UTC

Modified: January 26, 2026 at 11:11 PM UTC

Modified: January 26, 2026 at 3:30 PM UTC

Modified: January 26, 2026 at 3:30 PM UTC

Modified: January 25, 2026 at 1:00 PM UTC

Modified: January 25, 2026 at 12:59 PM UTC

Modified: January 25, 2026 at 12:59 PM UTC

Modified: January 25, 2026 at 12:59 PM UTC

Modified: January 25, 2026 at 12:59 PM UTC

Modified: January 25, 2026 at 12:59 PM UTC

Modified: January 25, 2026 at 12:58 PM UTC

Modified: January 25, 2026 at 12:58 PM UTC

Modified: January 25, 2026 at 12:52 PM UTC

Modified: January 25, 2026 at 12:52 PM UTC

Modified: January 25, 2026 at 12:52 PM UTC

Modified: January 25, 2026 at 12:52 PM UTC

Modified: January 25, 2026 at 12:51 PM UTC

Modified: January 25, 2026 at 12:50 PM UTC

Modified: January 25, 2026 at 12:49 PM UTC

Modified: January 25, 2026 at 12:49 PM UTC

Modified: January 25, 2026 at 12:49 PM UTC

Modified: January 25, 2026 at 12:45 PM UTC

Modified: January 25, 2026 at 12:29 PM UTC

Delete dataset

Are you sure you want to delete this dataset?

Create dataset