Using Tree Detection Algorithms to Predict Stand Sapwood Area, Basal Area and Stocking Density in Eucalyptus regnans Forest

Abstract

Managers of forested water supply catchments require efficient and accurate methods to quantify changes in forest water use due to changes in forest structure and density after disturbance. Using Light Detection and Ranging (LiDAR) data with as few as 0.9 pulses m −2 , we applied a local maximum filtering (LMF) method and normalised cut (NCut) algorithm to predict stocking density (SDen) of a 69-year-old Eucalyptus regnans forest comprising 251 plots with resolution of the order of 0.04 ha. Using the NCut method we predicted basal area (BAHa) per hectare and sapwood area (SAHa) per hectare, a well-established proxy for transpiration. Sapwood area was also indirectly estimated with allometric relationships dependent on LiDAR derived SDen and BAHa using a computationally efficient procedure. The individual tree detection (ITD) rates for the LMF and NCut methods respectively had 72% and 68% of stems correctly identified, 25% and 20% of stems missed, and 2% and 12% of stems over-segmented. The significantly higher computational requirement of the NCut algorithm makes the LMF method more suitable for predicting SDen across large forested areas. Using NCut derived ITD segments, observed versus predicted stand BAHa had R 2 ranging from 0.70 to 0.98 across six catchments, whereas a generalised parsimonious model applied to all sites used the portion

Use of a forest sapwood area index to explain long‐term variability in mean annual evapotranspiration and streamflow in moist eucalypt forests

“…Current research aims to use LiDAR data to map SA across larger catchments supplying water to the city of Melbourne, Australia. Progress has been made in deriving reasonably accurate estimates of FD and B 1.3 from LiDAR data [ Jaskierniak et al ., ]. The research presented here indicates that maps of forest SA derived using LiDAR will provide valuable information, for example by enabling comparison of SA and therefore L across different catchments, or by enabling forest protection or management to focus on parts of catchments identified as having low SA and therefore relatively high water yield.…”

Predicting long‐term streamflow variability in moist eucalypt forests using forest growth models and a sapwood area index

“…For ash forests, we use forest growth models to quantify decadal trends in post‐disturbance forest water use. Below, we outline how LiDAR data were first used to predict N and BA across the catchments with the LMF and NCut method respectively [ Jaskierniak et al ., ], and second, how the N and BA maps were used with PGP to generate forest growth models across the catchments.…”

“…Using Light Detection and Ranging (LiDAR) data, Jaskierniak et al . [] applied individual tree detection (ITD) algorithms to predict N , BA , and SA across approximately 10,000 ha. To estimate N , they applied the local maximum filtering (LMF) method and correctly identified 72% of stems, with over‐segmentation occurring for larger stems within low density forest (2%) and omission occurring in smaller stems within high stocking density forest (25%).…”

“…Using the Normalised Cut (NCut) algorithm, Jaskierniak et al . [] developed a generalized model that used LiDAR to explain 68% of BA across a broad range of E. regnans forest conditions. Benyon et al .…”

Tree Density and Forest Productivity in a Heterogeneous Alpine Environment: Insights from Airborne Laser Scanning and Imaging Spectroscopy

“…Yet, other studies evaluating the TD estimation accuracy report R 2 values ranging from 0.40 to 0.85 [5,25,30,41,75,86]. Jaskierniak et al [87] indicate TD estimation accuracies ranging between 40-70% considering several studies carried out in Scandinavian and European forests.…”