A comparison of single-photon lidar, conventional lidar, and digital aerial photogrammetry for area-based forest inventory

Abstract

Abstract More than 20 years have passed since forest management inventories were first implemented using discrete-return linear-mode lidar (LML) and the area-based approach (ABA). Among recent sensor innovations, single-photon lidar (SPL) stands out as a particularly promising technology for improving the cost efficiency of ABA. The main objective of this study was to assess the cost efficiency of ABA when reducing SPL point density by increasing flight altitude, thereby enabling larger area coverage. Three SPL datasets (10.1, 24.2, and 59.3 first returns/m2) were compared with two LML datasets (2.6 and 136.9 first returns/m2) and digital aerial photogrammetry (DAP) (61.3 points/m2). The analysis used 249 systematically distributed ground plots across various boreal forest types, and nonlinear models were constructed for basal area, stem density, volume, Lorey’s mean height, and dominant height. Model predictions were further validated using 530 independent plots aggregated to 47 stands. The results showed that SPL achieved model accuracies comparable to LML and consistently better than DAP. When averaging relative root mean square error (RMSE) across all biophysical attributes and forest types, SPL low-density data yielded smaller errors than low-density LML data and DAP in both cross-validation and independent stand validation. For example, in mature, highly productive forest—where more than half of the validation stands were located—the RMSE values for volume were 9.3, 8.7, 15.6, 6.4, 6.8, and 6.2% for LML low density, LML high density, DAP, SPL low density, SPL medium density, and SPL high density, respectively. Furthermore, SPL data collected at multiple flight altitudes indicate that operating above commonly reported in the literature and manufacturer-recommended heights can allow for up to a 30% reduction in flight distance while maintaining comparable model accuracy, although such gains may not translate into proportional cost reductions due to technical and atmospheric constraints limiting suitable flight conditions.

Plant phenotyping relevance

森林プロットの生物物理形質を推定する複数のLiDAR・写真測量法を比較し、独立データで精度検証しているため、センサーに基づく植物形質計測法の検証が中心である。

Code and data availability

公開論文であることは確認できましたが、現在の公式API・許可済み取得経路では本文を自動取得できませんでした。

No evidence-backed public reproduction asset is currently recorded.