Integrating personal laser scanning with uncrewed aerial vehicle-based photogrammetry or laser scanning for accurate stem volume estimation in temperate coniferous forests

ABSTRACT

An extensive number of trees are surveyed for logging purposes, necessitating efficient and precise logging survey methods. Integrating over- and under-canopy remote sensing, such as personal laser scanning (PLS) with uncrewed aerial vehicle-based laser scanning (ULS) or digital aerial photogrammetry (DAP), is a promising method. In this study, we aimed to estimate stem volume in a 16.89-ha coniferous forest using these techniques at the individual, plot, and stand levels. The root mean square error (RMSE) and relative RMSE (rRMSE) when compared to conventional field surveys of PLS-based volume estimations were 0.14 m³ (21.6%) before and 0.08 m³ (12.3%) after bias correction. When integrating PLS with ULS and DAP, the RMSE and rRMSE values were 0.16 m³ (26.7%) and 0.16 m³ (27.0%), respectively. The F-score for the individual tree detection analysis by ULS and DAP were 0.94 and 0.96, respectively. The RMSE of the estimated volume in the plots was 0.08 m³ for both ULS and DAP. The estimated volumes of timber produced in the stand using the utilization rate overestimated the actual volume by 30.6%, 13.0%, and 17.0%, for PLS, ULS, and DAP, respectively. While the individual stem volume estimation accuracy of ULS and DAP was lower than that of the field survey, at the plot level the results were comparable, with ULS and DAP showing results significantly closer to the actual results than those of PLS at the stand level. This study highlights the potential for integrating PLS with uncrewed aerial vehicle methods for estimating stem and log volumes at the plot and stand levels.

KEYWORDS:

• LiDAR
• photogrammetry
• backpack laser scanning
• individual tree detection
• Japanese cedar
• Japanese cypress

Acknowledgements

We thank the Fuji City Forestry Department for allowing us to study in the city-owned forests. We thank Bido Tsuchiya for supplying the conventional field survey data. We thank Masakazu Oguri and Hiroko Muramatsu for their assistance with the PLS surveys. We would like to thank Editage (www.editage.com) for English language editing.

Disclosure statement

No potential conflict of interest was reported by the author(s).