A Methodological Framework for Mapping Canopy Cover Using ICESat-2 in the Southern USA

NASA’s Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) provides exceptional opportunities for characterizing the structure of ecosystems through acquisition along-track, three-dimensional observations. Focusing on canopy cover as a fundamental parameter assessing forest conditions, overall goal this study was to establish framework generating gridded 30 m product with ICESat-2. Specifically, our objectives were (1) Determine compare ICESat-2-derived airborne lidar-derived 2016 National Land Cover Database (NLCD) estimates, (2) Evaluate methodology wall-to-wall mapping cover. Using two Southern US sites, Sam Houston Forest (SHNF) in south-east Texas Solon Dixon Forestry Education Center (SDFEC) southern Alabama, four measures estimated ICESat-2′s Land-Vegetation Along-Track Product, or ATL08, evaluated at pixel scale. Comparisons made using spatially coinciding NLCD pixels reference A suite Landsat Landsat-derived parameters then used predictors model map each measure Random Forests (RF), their accuracies assessed compared. Correlations (r) between lidar scale ranged from 0.57 0.78, R2 up 0.81 produced RF models developed extrapolating estimate yielded values 0.50 0.61 (RMSEs 16% 20%) when With demonstrated capability ICESat-2 vegetation biophysical parameters, findings serve support comprehensive other attributes, especially aboveground biomass, contribute development an up-to-date product.