Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Fabric Maps can visualize spatial data stored in a Microsoft Fabric lakehouse. By connecting a map to a lakehouse, you can create map layers directly from files in OneLake—without copying data or deploying external GIS services. This approach lets you keep geospatial data centrally governed in the lakehouse while using Fabric Maps for interactive visualization and analysis.
Fabric Maps supports connecting to lakehouses directly from the map explorer and browsing items through the OneLake catalog.
How lakehouse data is used in maps
A lakehouse stores data files in OneLake. When you add a lakehouse file to a map, Fabric Maps reads the data directly from OneLake and renders it as a map layer. Lakehouse‑based layers are file‑backed and are optimized for reference, boundary, and imagery data rather than high‑frequency streaming updates. The lakehouse remains the system of record, and changes to the underlying data are reflected in the map based on refresh behavior.
Supported lakehouse data types
Fabric Maps supports creating layers from the following lakehouse‑based geospatial data types:
Vector data
- GeoJSON files – Feature collections containing point, line, or polygon geometries with associated properties.
- PMTiles – Cloud‑optimized, single‑file vector tile archives designed for efficient map rendering at scale. For more information on PMTiles, see PMTiles.
Vector data is rendered as interactive layers that support styling, labeling, filtering, and clustering, depending on the data type and layer configuration.
Raster data
A Cloud Optimized GeoTIFF (COG) is a GeoTIFF format optimized for cloud access, allowing clients to retrieve only the portions of an image needed for rendering. For more information, see cogeo.org.
Important
Fabric Maps currently supports Cloud Optimized GeoTIFF (COG) files that meet the following requirements:
- Projection: Only files using the EPSG:3857 (Web Mercator) projection system are supported.
- Bands: Supported formats include 3-band RGB (Red, Green, Blue) and 4-band RGBA (Red, Green, Blue, Alpha for transparency). Single-band and other multi-band configurations aren't supported.
Ensure imagery is correctly formatted before uploading. Files that don't meet these requirements won't render and may trigger an error.
Raster data is rendered as imagery layers and can be combined with basemaps and vector layers to provide geographic context.
Why use a lakehouse for map layers
Using a lakehouse as a data source for Fabric Maps provides several benefits:
- Centralized storage and governance in OneLake.
- Compatibility with other Fabric engines such as Notebook, Warehouse, and Power BI.
- No data duplication between analytics and visualization workloads.
- Support for both vector and raster geospatial formats.
Relationship to other data sources
Fabric Maps also supports other data sources, such as Eventhouses (KQL databases). Lakehouse‑based layers are file‑backed, while Eventhouse‑based layers can be created from KQL data in Kusto tables, functions, and materialized views. For more information on layers created from KQL databases, see Kusto integration in Fabric Maps. Both can be combined in the same map to enrich spatial analysis.