Authors: Wenli Huang*,
Topics: Natural Resources, Remote Sensing, China
Keywords: Forest Height, Hunan, China, Lidar, Landsat, Radar
Session Type: Guided Poster
Start / End Time: 3:05 PM / 4:45 PM
Room: Roosevelt 3.5, Marriott, Exhibition Level
Presentation File: No File Uploaded
Accurate quantifying forest height at a fine spatial resolution over large areas is essential for accurate carbon accounting. Yet, current regional to national scale assessment of forest height rely primarily on statistical or coarse-scale model estimates, thus lack of spatial details for decision making at local scales. Recent advances in remote sensing technology provide great opportunities to fill this gap. Here, we present a suite of algorithms designed to combined forestry inventory data and remote sensing dataset to map forest height from local to regional scales. We utilized plot data from airborne lidar data for model calibration, and independent measurements collected from 2017 to 2018 for model calibration and validation. Specifically, we first modeled the relationship between plot-level weighted mean height and footprint-level lidar metrics and applied it to all footprints. Then, we calibrated machine-learning models by linking lidar-derived height and corresponding optical, radar and optical metrics. Lastly, we applied these models to the regional scale and evaluated results using independent plot data. Primary results indicate that there is no bias in height estimates when comparing to statistics summarized at the regional scale, while uncertainties at local scales arose from the mismatches between field measurements and remote sensing dataset. This study provides information toward establishing regional- to national- scale forest carbon evaluation system.