Mapping the living world
Paper · published 14 May 2026 · Naturecode Project
Most institutions that study the planet study a slice of it. One group watches forests. Another watches coral. Another watches air quality, or fish stocks, or soil carbon. The work is good. It is also incomplete in a particular way: the question what is happening to the living world cannot be answered from any one slice.
This paper sketches a research agenda for asking that question more honestly — by working across the places life actually lives, with the instruments that can actually see them, and in partnership with the people closest to the ground.
The argument
Ecosystems do not respect the boundaries we draw between disciplines. A coastal mangrove is a forest, a fishery, a carbon sink, a sediment trap, a human livelihood, and a satellite target — all at once. Studying it from one angle gives an answer that is technically correct and practically misleading.
The alternative is not a single grand model. It is comparable layers — open, time-stamped, and built so a researcher in one field can read the work of another without translation costs.
We argue for five things at once:
- Many places, one practice. Oceans, rivers and lakes, forests, soils, atmosphere, and orbit are not separate research domains for us. They are parts of a single living system that needs a single open practice for studying it.
- Many instruments, one record. Satellite imagery, acoustics, environmental DNA, in-situ sensors, and field observation each see something the others miss. Fusing them honestly is harder than fusing them quickly.
- AI as instrument, not oracle. AI helps when used as a careful method. It harms when used to manufacture confidence. We make this distinction the design principle, not the disclaimer.
- Open by obligation. Methods, datasets, and code are published because that is how research becomes trustworthy — not as marketing.
- Ground truth from the ground. Rangers, fishers, farmers, pastoralists, and Indigenous and local communities see what satellites cannot. The work is incomplete without them.
What this means in practice
The agenda translates into the five research themes that anchor our work:
- Mapping nature — assembling the open layers
- Monitoring ecosystems — reading them over time
- Analysis and prediction with AI — methods, carefully
- Conservation and food security — where evidence becomes decision
- Open data and open collaboration — how the work is shared and stewarded
Each theme is open to collaboration. Each is published in the open. None of them ends.