Monitoring Changes in Aboveground Biomass Across US Forestlands Using Satellite Observations and Forest Inventory Data

Detecting changes of forest aboveground biomass from disturbance and recovery is critical for quantifying the dynamics of forests as well as sinks and sources of carbon at global scale. Here, we use the Forest Inventory and Analysis (FIA) data from the US Forest Service collected annually across US forestlands, and a suite of remote sensing observations from radar (ALOS PALSAR and SRTM) and optical (Landsat) satellites, to map the forest aboveground biomass (AGB) at 1-ha spatial grid cell over the conterminous United States (CONUS) forests. We use non-parametric models based on Maximum Entropy estimations to map AGB for different years (2005, 2010, 2015, 2016, and 2017) to allow quantification of changes at both 5-year and annual intervals along with uncertainty. Loss and Gain data from FIA plots in various forest types are used to constrain the changes observed in remote sensing mapping and provide regional estimates of carbon dynamics for more than a decade at state and county scales.