Biomass

Biomass is terraPulse's global dataset of actual and potential aboveground biomass, reported either as carbon mass (Mg C) or as biomass density (Mg C/ha). The product's validation study describes high-resolution monitoring of aboveground biomass as essential for carbon accounting and ecosystem management.

The estimates are produced by artificial intelligence trained on spaceborne lidar and other reference measurements, correlated with terraPulse's own satellite-derived land cover, disturbance history and vegetation index layers and other environmental variables. The product page describes resolution of up to 30 metres at annual time steps; the standard Actual and Potential Aboveground Biomass products are listed annually from 2000 to the present, with a latency of one year. Certainty is reported as root mean square error (RMSE).

Two levels are offered: a globally calibrated dataset, or an enterprise dataset calibrated to a client's own reference measurements. The layers can be explored in the terraView map dashboard and are part of the terraPulse data cube served through the terraServe API.

A study published in Science of Remote Sensing (2026), co-authored by terraPulse co-founder Joseph Sexton, compared the dataset with US Forest Service Forest Inventory and Analysis estimates across Utah, Nevada and Washington. It reported R² = 0.88 and RMSE = 26.68 Mg/ha at the 64,000-hectare hexagon scale, with overestimates in non-forest areas and lower values in high-biomass forests attributed to saturation of the optical covariates.