Regional forest canopy height and structure models, integrating GEDI spaceborne-lidar and Sentinel-1, -2 data
On this page:
The success of sustainable development and climate initiatives rests on our capacity to remotely monitor ecosystems in a way that is accurate, precise, and recurring. In particular, estimates of forest canopy height and structure underpin efforts to quantify forest carbon emissions and ensure the success of burgeoning carbon markets. Onboard the International Space Station, the Global Ecosystem Dynamics Investigation (GEDI) lidar instrument began collecting measurements of forest canopy height, canopy vertical structure, and surface elevation at an unprecedented scale in 2019. Due to the active nature of the sensor and the short timeframe of the two-year mission, achieving consistent, broad-scale monitoring of forest ecosystems requires extrapolating beyond the orbits of the GEDI instrument. Here, I train models to predict canopy height and structure based on synthetic aperture radar (SAR) and spectral data collected by the Sentinel-1 and -2 missions. The predicted images retain the 10-meter resolution of the Sentinel imagery. With regional airborne-lidar data, I calibrate the metrics collected by the GEDI sensor and determine the accuracy of the models’ predictions. Predicting out-of-sample with past Sentinel data yields an approach for assessing long-term historic trends as well as forward monitoring of forest dynamics.