Platform outputs trace to scientific literature and knowledge traditions, name their uncertainties, and add to a shared evidence base
The reasoning layer draws on formal ecological datasets and published literature spanning plant functional ecology, climate science and downscaled projection modelling, soil science and biogeochemistry, landscape ecology, restoration ecology, assisted migration science, and environmental psychology.
It draws equally on the experiential knowledge of practitioners, farmers, gardeners, ecologists, and long-term land stewards. Every source is held to the same epistemic rules. No single tradition claims final authority.
PerenOS draws on formal scientific data
280,000+ taxa
TRY Plant Trait Database
Functional trait data across 280,000+ plant taxa, used to characterise species roles within modelled communities.
Distribution modelling
GBIF
Global species occurrence records used to ground distribution modelling under current and projected climate conditions.
ERA5 · CMIP6
Climate projections
ERA5 current conditions and CMIP6 scenario trajectories used to index community viability across each site's projected climate horizon.
National · global
Soil and site data
National and global soil datasets integrated with observed site conditions to anchor community selection in physical substrate.
1
Observation
What happens on the ground.
2
It is logged
A dated field record.
3
It enters the dataset
Tagged by zone, soil, community.
4
Patterns emerge
Across sites with similarities.
5
Guidance returns
Grounded in what actually happened.
Epistemic refusal
Where the evidence available is insufficient to produce a defensible recommendation, the platform states that explicitly and does not generate an output. All outputs are provisional.
Governance
Every output is traceable and uncertainty is named explicitly. The authority to decide always stays with the practitioner.