Influence of micro-topography and crown characteristics on tree height estimations in tropical forests based on LiDAR canopy height models

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Documents

DOI

  • Cici Alexander
  • Amanda H. Korstjens, Bournemouth University
  • ,
  • Ross A. Hill, Bournemouth University

Tree or canopy height is an important attribute for carbon stock estimation, forest management and habitat quality assessment. Airborne Laser Scanning (ALS) based on Light Detection and Ranging (LiDAR) has advantages over other remote sensing techniques for describing the structure of forests. However, sloped terrain can be challenging for accurate estimation of tree locations and heights based on a Canopy Height Model (CHM) generated from ALS data; a CHM is a height-normalised Digital Surface Model (DSM) obtained by subtracting a Digital Terrain Model (DTM) from a DSM. On sloped terrain, points at the same elevation on a tree crown appear to increase in height in the downhill direction, based on the ground elevations at these points. A point will be incorrectly identified as the treetop by individual tree crown (ITC) recognition algorithms if its height is greater than that of the actual treetop in the CHM, which will be recorded as the tree height. In this study, the influence of terrain slope and crown characteristics on the detection of treetops and estimation of tree heights is assessed using ALS data in a tropical forest with complex terrain (i.e. micro-topography) and tree crown characteristics. Locations and heights of 11,442 trees based on a DSM are compared with those based on a CHM. The horizontal (DH) and vertical displacements (DV) increase with terrain slope (r = 0.47 and r = 0.54 respectively, p < 0.001). The overestimations in tree height are up to 16.6 m on slopes greater than 50° in our study area in Sumatra. The errors in locations (DH) and tree heights (DV) are modelled for trees with conical and spherical tree crowns. For a spherical tree crown, DH can be modelled as R sin θ and DV as R (sec θ – 1). In this study, a model is developed for an idealised conical tree crown, DV = R (tan θ – tan ψ), where R is the crown radius, and θ and ψ are terrain and crown angles respectively. It is shown that errors occur only when terrain angle exceeds the crown angle, with the horizontal displacement equal to the crown radius. Errors in location are seen to be greater for spherical than conical trees on slopes where crown angles of conical trees are less than the terrain angle. The results are especially relevant for biomass and carbon stock estimations in tropical forests where there are trees with large crown radii on slopes.

Original languageEnglish
JournalInternational Journal of Applied Earth Observation and Geoinformation
Volume65
IssueMarch
Pages (from-to)105-113
Number of pages9
ISSN1569-8432
DOIs
Publication statusPublished - 2018

    Research areas

  • Airborne laser scanning, Forestry, ITC, REDD, Tree delineation, CARBON, LASER-SCANNING DATA, RAIN-FOREST, VARIABLE WINDOW SIZE, INDIVIDUAL TREES, EXTRACTION, POINT CLOUDS, TERRAIN MODELS, AIRBORNE LIDAR, VEGETATION

See relations at Aarhus University Citationformats

ID: 121253168