Provisional name — proposal under discussion

Making economic geography not just visible, but scientifically intelligible

GEONIA — Geospatial, Estimation, Optimisation, Numerical methods, Artificial Intelligence — is the data platform of the proposed Montréal research centre for geospatial AI and digital-twin science. It extends GVCdtLab's digital twin of the St. Lawrence – Great Lakes corridor.

What the platform does

A map as the front door

Every economic question opens on the map: zooming reveals successive layers — trade flows, ports, mines, indicators — and time is as manipulable as space.

Every answer carries provenance

Source, licence, timestamp, version and fingerprint for every datum; caveats on association, prediction and causation.

Web and API, as equals

Any question asked of the map can be asked from code (JSON, CSV, SQL), with examples on every page.

Five research directions

Direction	Live questions	Examples
Economic indicators & global value chains	15 / 15	Where is the world economy growing, inflating or opening up? How has this economy evolved, and what does the IMF project next? Are countries getting richer and healthier together?
Critical resources	4 / 4	Where are Canada's producing mines, smelters and next projects? Where could Canadian critical minerals replace a disrupted supplier? Which economies depend most on mineral wealth?
Blue economy	8 / 8	Which ports and regions move the most cargo? Where is seaborne trade growing or shrinking? Which ports are export platforms and which are import gateways?
Transport	1 / 3	Where does a province's or state's freight go, by which mode, and what's in it? Which ports are the critical nodes of the Great Lakes network? How do water levels constrain Great Lakes shipping?
Logistics	0 / 5	How do provinces trade with each other, and what do models predict? How would a tariff shock propagate through supply chains? Which Canada–U.S. industries transmit price shocks to others?

Beyond visualisation

Inspired by Gapminder, Google Earth Engine and living maps such as Offentligdata, the platform aims further: backing each layer with a calibrated state-transition model, uncertainty intervals and counterfactuals validated out of sample, in space and in time. A convincing visualisation remains a hypothesis until it has been tested.

Status: proposal. No institutional designation, affiliation or endorsement (UNU, universities) is claimed.