Advanced machine learning is moving to the heart of how the UK tracks the health and carbon impact of its forests – with an Edinburgh-based AI consultancy now applying foundation‑model techniques to woodland monitoring backed by the UK Space Agency.
Through the agency’s Climate Services Call, New Gradient is developing a digital Monitoring, Reporting and Verification (dMRV) platform that uses earth observation data and geospatial machine learning to automate analysis of woodland projects at scale, as outlined on the UK Space Agency blog. The goal is to replace slow, manual field surveys with continuous, AI‑driven monitoring that can support climate and nature markets.
According to reporting by Consultancy.uk, New Gradient’s system builds on a self‑supervised “foundation model” trained on around 1.8 million pairs of UK aerial images, which can be fine‑tuned to recognise different woodland features and conditions. The firm has also developed specialist models for tree counting, crown segmentation, species classification, canopy height estimation and a biomass and carbon pipeline, turning raw imagery into structured data on forest structure and carbon stocks.
By automating these processes, the platform is designed to give land managers, verifiers and investors much more frequent and granular insight into woodland carbon delivery and biodiversity outcomes. That capability is particularly relevant to frameworks such as the Woodland Carbon Code, where trusted data can speed up verification cycles and strengthen confidence in the resulting carbon units.
For the UK Space Agency, New Gradient sits within a broader effort to grow the UK’s climate services sector by combining space‑derived data with advanced analytics. Through its Climate Services Call, the agency is backing a cohort of projects that use earth observation and AI to support net zero, positioning space technology and machine learning as core infrastructure for the next generation of climate and nature markets.
