Inside the SDF — from data to dashboard
This walkthrough follows the full lifecycle of data through the Spatial Data Framework: registering sources, defining boundaries, processing products, building dashboards, and opening access via API. Each step is a short screen recording from the SDF admin console.
Four concepts, one framework
The SDF organises all data around four core concepts, via a strict data model. Every processing run is recorded with full provenance — so any output can be traced back to the exact sources, boundaries, and calculations that produced it.
| Concept | What it is | Example |
|---|---|---|
| Dataset | The source data to be analysed — satellite imagery, field surveys, national statistics, or research outputs. | Global Mangrove Watch habitat extents |
| Geometry | The spatial boundaries used to segment and summarise datasets — national borders, EEZ boundaries, or custom regions. | Exclusive Economic Zones (EEZ) per country |
| Product | A dataset processed against a geometry. Each product run is immutable, versioned, and fully traceable. | Mangrove extent per EEZ — one value per country |
| Indicator | The variables calculated for each product. Derived indicators can compose from base indicators for full flexibility. | Total mangrove area (ha), percentage change year-on-year |
Add a dataset
Register a new dataset in the admin console — whether satellite imagery, national statistics, or field survey data. The SDF supports raster and vector formats including STAC catalogues, Cloud Optimised GeoTIFF, GeoParquet, and zipped Shapefiles.
Add a geometry
Upload the spatial boundaries that will be used to segment and summarise your datasets. Geometries can range from national Exclusive Economic Zone boundaries to custom sub-national administrative regions or protected areas.
Create a product
Combine any dataset with any geometry to create a product. For example, pair Global Mangrove Watch data with EEZ boundaries to produce mangrove extent per country — or combine coral reef coverage with custom marine protected area polygons.
Define indicators
Browse and define the variables calculated for each product. Derived indicators can compose from base indicators — giving users full flexibility to create custom metrics like percentage coverage, condition scores, or year-on-year change rates without modifying the underlying processing pipeline.
Under the hood
Behind the scenes, the SDF processes data through an orchestration framework like Argo Workflows. Datasets and geometries are combined to produce products with immutable, versioned outputs — every run is traceable and reproducible, and each processing step records its provenance automatically.
Inspect a product
Drill into any product to see exactly which datasets, geometries, and indicators contributed to it. Every link is explicit and immutable — the full provenance chain is preserved so that any output value can be independently verified back to its source data.
Build a dashboard
Start assembling a dashboard from your data and products. Choose from templates for common reporting frameworks — ocean accounts, SDG reporting, national ecosystem accounting — or build from scratch combining maps, charts, and tables in a single view.
Dashboard provenance
Every element in a dashboard retains its full chain of provenance back to the underlying source data. Click any value to trace its lineage from raw input through processing to final output — the same auditability that applies to individual products carries through to every dashboard view.
Generate API keys
Create API keys to provide programmatic access to your data and products — enabling integration with external tools, scripts, and reporting systems. Keys are scoped and managed through the admin console with full audit trails.
Direct API access to data
The SDF exposes a full API for direct programmatic access to all data and products. This means any tool that can make an API request can pull live data from the framework — notebooks, scripts, pipelines, or custom applications. Here, a Python notebook queries the API to explore and visualise data with pandas, geopandas, and matplotlib.
From API to any dashboard
Use the SDF’s API to feed data into any external reporting tool or dashboard. If the built-in templates or open source examples don’t fit your needs, the API gives you complete flexibility — connect to Power BI, Tableau, custom web applications, or any system that can consume a REST API.
Explore the code
The Spatial Data Framework is open source. Browse the repositories, run it locally, or get in touch to discuss adoption and collaboration.