Stand volume estimation by combining low laser-sampling density LiDAR data with QuickBird panchromatic imagery in closed-canopy Japanese cedar (Cryptomeria japonica) plantations

摘要

This study proposes a simple method for stand volume estimation by combining low laser-sampling density Light Detection and Ranging (LiDAR) data (i.e. 1 hit per 4 m~2) with high-resolution optical imagery (i.e. 0.6 m) in coniferous plantations. The study area was in closed-canopy, mountainous, Japanese cedar (Cryptomeria japonica) plantations on undulating terrain at an elevation of 135-391 m above sea level. A total of 25 circular plots (0.04 ha) were established and stand volumes within plots were investigated in the field. The field-measured, plot-level stand volume ranged from 262.8 to 984.0 m~3 ha~(-1) and the average value was 555.7 m~3 ha~(-1). We used the measurements as validation data to evaluate estimates derived from an empirical regression model that was constructed on the basis of the allometry between crown diameter and diameter at breast height (DBH) from previous research. As a result, stand volume at various stand conditions could be estimated precisely regardless of different laser footprint sizes of 0.16-0.47 m when combining low-density LiDAR data with QuickBird panchromatic imagery. The maximum random error and root mean square error (RMSE) in stand volume estimates by data combination were 10% and 39% of the average stand volume, respectively. Thus, this method based on allometry and using low-density LiDAR data and high-resolution optical imagery could be capable of offering precise stand volume estimates in coniferous forests in different localities.