Estimating the overstory and understory vertical extents and their leaf area index in intensively managed loblolly pine (Pinus taeda L.) plantations using airborne laser scanning

Data from four discrete-return airborne laser scanning (ALS) acquisitions and three different sensor types across seven experimentally varied loblolly pine (Pinus taeda L.) plantations were used to test published and novel methodologies in quantifying forest structural attributes within stands, including height to live crown (HTLC; i.e. the lowest vertical canopy extent) of the canopy and the contributions to total plot-level leaf area from understory and overstory canopy vegetation. These ALS data were compared to in situ field measurements to develop ALS-based predictive models of these attributes. The correlation between field- and ALS-modeled HTLC data was strong, with an R-2 of 0.79 (p < 0.001). We assessed the ability of eight lidar light penetration indices to estimate effective leaf area index (eLAI) in the field. The best predictor of total (sum of understory and overstory) eLAI produced an R-2 of 0.88 (p < 0.001). The independent contributions of overstory and understory components could also be accurately predicted by ALS-derived canopy-only eLAI metrics (R-2 = 0.71; p < 0.001) and understory-only metrics (R-2 = 0.49; p < 0.001). Two new indices, calculated as the sum of return intensity for each foliar layer and correcting for transmission losses, were developed specifically for the vertical strata related to the understory (BLunder) or overstory (BLover). The estimates from BLover were equivalent to the best-performing indices for predicting canopy-only eLAI and the corresponding BLunder was superior to other indices for understory eLAI. The broad spatial and temporal extents of the data, as well as the inclusion of pine plantations with differing stand ages, planting densities, understory control, and thinning treatments, suggest the relationships generated from these methods are robust to site and seasonal variability. The results produced from the analysis of multiple acquisitions implies that the methods presented here are transferable across location, time and sensor design, without implementation-specific calibration, at least for structurally similar loblolly pine plantations.

Automated summary

This study utilized ALS data from different sensor types to quantify forest structural attributes in loblolly pine plantations, with strong correlations between ALS and field measurements for canopy height and leaf area. New lidar indices were developed to predict overstory and understory leaf area more accurately, suggesting transferability of the methods across location, time, and sensor design.