Premium
Feasibility of terrestrial laser scanning for quantification of vegetation structure parameters of restored sandy land in the southern Qinghai–Tibeten Plateau
Author(s) -
Tian Jiarong,
Li Haidong,
Li Yingkui,
Liao Chengrui,
Ma Weibo,
Xu Yannan
Publication year - 2020
Publication title -
land degradation and development
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.403
H-Index - 81
eISSN - 1099-145X
pISSN - 1085-3278
DOI - 10.1002/ldr.3784
Subject(s) - vegetation (pathology) , quadrat , plateau (mathematics) , crown (dentistry) , environmental science , scale (ratio) , physical geography , lidar , remote sensing , geology , hydrology (agriculture) , soil science , geography , cartography , mathematics , transect , medicine , mathematical analysis , oceanography , dentistry , pathology , geotechnical engineering
Abstract The efficiency of ecosystem restoration of degraded land is typically assessed using vegetation structure parameters (VSP). However, the field measurement of VSP, particularly in the Qinghai–Tibet Plateau (QTP), which is a problematic area for ecological restoration because of its harsh environment, is generally time‐consuming and labour intensive. Therefore, new methods for VSP measurement should be developed. In this study, VSP were quantified at the individual scale, including their number, height, and crown width. Furthermore, the quadrat scale, such as vegetation coverage (VC) and leaf area index (LAI) on a restored sandy plot (100 m × 100 m) in the middle reaches of the Yarlung Zangbo River basin, is quantified using terrestrial laser scanning. Thereafter, the TLS‐derived VSP are validated by field measurements. To extract the VSP of low‐height vegetation (shrubs) at the individual scale, three approaches are tested: the manual measurement of VSP from TLS data, individual vegetation segmentation (IVS) without specific root locations, and IVS with specific root locations. The results indicate that the VSP of shrubs can be manually derived from the TLS data and are consistent with field measurements. By specifying the root locations before performing the IVS, the identification rate of the number of shrubs can be improved significantly from 40.3% to 100%. The height extracted by the IVS corresponds better with the manual measurement ( R 2 > 0.99) than with the crown width ( R 2 < 0.65). At the quadrat scale, the optimal resolution is analyzed to extract the VC. The analysis suggests that the optimal resolution for extracting the VC based on the canopy height model is in the 1‐2 cm range. In extracting the LAI using the voxel‐based method, it is discovered that the appropriate voxel size is linearly dependent on the mean point spacing ( y = 2.143x, R 2 = 0.826). This study provides useful guidance for monitoring the effectiveness of ecological conservation and restoration on the QTP.