Documentation

unseea

A UN SEEA-EA decision planning tool. Draw any area on Earth, get its five UN ecosystem accounts computed live from open data — then propose a land-use change and watch the accounts recompute into the standard's own vocabulary: managed expansion, managed reduction, degradation, enhancement, revaluation.

Status: scoping. No application code yet. These pages explain the method the tool will implement; the design documents in the repository carry the detail.

Explainers

Written for someone who needs a working mental model of ecosystem accounting before reading the design documents.

01

SEEALand, end to end

Published

The standard's own worked toy example — 250 hectares, six ecosystem types, one 2-hectare conversion — traced through all five accounts. What each step does, what it consumes, and what it produces. Our Phase 1 replication target.

02

The h3 accounting primitives

Drafting

How each account reduces to a single grouped aggregation over hex cells, why fractional coverage beats dominant-class for extent, and the measured latency budget per account.

03

Condition, and the reference-level problem

Planned

Why rescaling a variable to [0,1] is the hardest unsolved step in the standard, and why condition can never be averaged across ecosystem types.

04

The scenario engine

Planned

What happens when a user repaints a hex: which accounts recompute, which assumptions become visible controls, and how the result stays inside the standard's vocabulary.

In the repository

The design record lives in Markdown next to the code, and is the authoritative source where these pages and the repository disagree.

DESIGN.md — what SEEA EA is, how ARIES for SEEA compares, the hard problems, the phased plan
DATA.md — data acquisition plan, per-account source selection, licence analysis
PROVENANCE.md — the planned layer list, and why each layer is in the app
research/seea-ea-reference.md — distilled controlled vocabularies and table structures from the standard
Issue #20 — the roadmap epic
Berkeley Schmidt Center for Data Science & Environment Built on the geo-agent runtime and the GLEN h3 catalog Source