Effectiveness of remote-sensing-based models for predicting NSVB aboveground biomass

Determine whether empirical models that predict forest aboveground biomass from remotely sensed data, including passive optical satellite imagery such as Landsat, can accurately predict the National Scale Volume and Biomass Estimators (NSVB) aboveground biomass estimates produced by the U.S. Forest Inventory and Analysis program, given NSVB’s emphasis on top and limb biomass components that are difficult to observe with passive sensors.

Background

Forest aboveground biomass (AGB) in the United States has commonly been mapped by fitting empirical models to remotely sensed data using Forest Inventory and Analysis (FIA) plot estimates as reference. Historically, those FIA AGB estimates were computed with the Component Ratio Method (CRM).

In late 2023, FIA introduced the National Scale Volume and Biomass Estimators (NSVB), which materially change tree-level AGB estimates—particularly by improving and increasing estimates of top and limb biomass. Because passive satellite data (e.g., Landsat) often saturate in dense canopies and cannot directly sense canopy interiors, it is uncertain whether existing model-based approaches can accurately predict NSVB-based AGB.