Regional rates of young US forest growth estimated from annual Landsat disturbance history and IKONOS stereo imagery

Forests of the Contiguous United States (CONUS) have been found to be a large contributor to the global atmospheric carbon sink. The magnitude and nature of this sink is still uncertain and recent studies have sought to define the dynamics that control its strength and longevity. The Landsat series of satellites has been a vital resource to understand the long-term changes in land cover that can impact ecosystem function and terrestrial carbon stock. We combine annual Landsat forest disturbance history from 1985 to 2011 with single date IKONOS stereo imagery to estimate the change in young forest canopy height and above ground live dry biomass accumulation for selected sites in the CONUS. Our approach follows an approximately linear growth rate following clearing over short intervals and does not estimate the distinct non-linear growth rate over longer intervals. We produced canopy height models by differencing digital surface models estimated from IKONOS stereo pairs with national elevation data (NED). Correlations between height and biomass were established independently using airborne LiDAR, and then applied to the IKONOS-estimated canopy height models. Graphing current biomass against time since disturbance provided biomass accumulation rates. For 20 study sites distributed across five regions of the CONUS, 19 showed statistically significant recovery trends (p < 0.001) with canopy growth from 0.26 m yr− 1 to 0.73 m yr− 1. Aboveground live dry biomass (AGB) density accumulation ranged from 1.31 t/ha yr− 1 to 12.47 t/ha yr− 1. Mean forest AGB accumulation was 6.31 t/ha yr− 1 among all sites with significant growth trends. We evaluated the accuracy of our estimates by comparing to field estimated site index curves of growth, airborne LiDAR data, and independent model predictions of C accumulation. Growth estimates found with this approach are consistent with site index curves and total biomass estimates fall within the range of field estimates. This is a viable approach to estimate forest biomass accumulation in regions with clear-cut harvest disturbances.