Evaluating Statewide NAIP Photogrammetric Point Clouds for Operational Improvement of National Forest Inventory Estimates in Mixed Hardwood Forests of the Southeastern U.S.

The U.S. Forest Service, Inventory and Analysis (FIA) program is tasked with making reporting estimates of various forest attributes using a design-based network permanent sampling plots. To make its more precise, FIA uses technique known as post-stratification to group plots into homogenous classes, which helps lower variance when deriving population means. Currently nationally available map tree canopy cover for post-stratification, tends work well area but less so structural like volume. Here we explore the use new statewide digital aerial photogrammetric (DAP) point clouds developed from stereo imagery collected by National Agricultural Imagery Program (NAIP) improve these in southeastern mixed hardwood forests Tennessee Virginia, United States (U.S.). Our objectives are 1. evaluate relative quality NAIP DAP airborne LiDAR height measurements, 2. assess ability models (DHMs) operational inventory above gains already achieved FIA’s current approach. results show were moderately strongly correlated field measured maximum heights (average Pearson’s r = 0.74) slight negative bias (−1.56 m) an RMSE error ~4.0 m. cloud also accurate softwoods (R2s 0.60–0.79) than hardwoods 0.33–0.50) structure that was consistent across multiple years measurements. Several factors served degrade relationship between data, including lack 3D points areas advanced senescence, spurious values deep shadows imprecision plot locations (which estimated be off true +/− 8 m). Using strata maps yielded volume 31% precise on average stratified data. Combining DHMs type information national products helped stratification performance, especially softwoods. monetary value post-stratify survey unit total USD 1.8 million vs. costs installing achieve similar precision gains. Overall, our benefit growing feasibility estimates.