Data Hosting
Atlas Amazon has to do two things that pull in opposite directions: accept continuous contribution and validation, and provide stable datasets that analysis and reporting can be built on. It resolves this by running a live system alongside versioned data releases.
Two Systems
Live system
The operational database, built on Supabase (PostgreSQL). It handles real-time updates to metadata, governance, assessments, and validation workflows, and is the continuously evolving version of the Atlas.
Published datasets
Versioned datasets released through platforms such as Source Cooperative. Each snapshot captures the state of the Atlas at a point in time and is intended for analysis, sharing, and long-term reference.
Versioned Releases
Each yearly release is a complete snapshot of the Atlas, covering:
- mapped areas (geometry and taxonomy)
- metadata (governance, designation, context)
- assessment results (PAME, CASE)
- indicators (stewardship, resilience, degradation, pressure)
- source datasets used in indicator generation
Once published, a release is fixed and does not change.
Why this Matters
Fixed releases are what make reproducibility and temporal analysis possible. Users can reproduce indicator results, compare changes between releases, trace shifts in governance, designation, and conditions, and verify findings cited in reports or decisions. Each release works as a stable reference dataset that will still mean the same thing in a year's time.
Data Structure
Published datasets mirror the structure of the live system, keeping spatial and structured data separate: spatial data holds geometries and raster layers, while structured data holds metadata, assessments, and indicators. The two are joined through consistent identifiers, so they can be used together or independently.
Publication Workflow
Data moves from the live system into a published release through four steps.
Contribute
Data is added and updated in the platform.
Validate
Entries are reviewed and confirmed.
Snapshot
A version of the dataset is created.
Publish
The dataset is packaged and released.
Only validated and appropriately governed data makes it into a published release.
Governance and Access
Publication follows the same governance and permission rules as the rest of the platform: some data is openly available, some is shared conditionally, and some remains restricted. Sensitive data may be excluded, generalized, or delayed — particularly for Collective Lands and stewardship information. See Open Data for how those levels are defined.
In Practice
Pairing a live system with versioned releases means the Atlas serves as a collaborative platform for day-to-day data management and as a long-term archive of conservation and stewardship data at the same time. That combination is what makes it possible to study how governance, stewardship, and ecological conditions change, rather than only what they look like today.