Estimating above-ground biomass of tropical rainforest of different degradation levels in Northern Borneo using airborne LiDAR

Published Date
Forest Ecology and Management
15 September 2014, Vol.328:335–341, doi:10.1016/j.foreco.2014.06.003

• Author 

• Keiko Ioki ^a,,
• Satoshi Tsuyuki ^a,
• Yasumasa Hirata ^b,
• Mui-How Phua ^c,
• Wilson Vun Chiong Wong ^a,c,
• Zia-Yiing Ling ^c,
• Hideki Saito ^b,
• Gen Takao ^b,

• aGraduate School of Agricultural and Life Sciences, University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8567, Japan
• bForest and Forest Products Research Institute, Matsunosato 1, Tsukuba, Ibaraki 305-8687, Japan
• cSchool of International Tropical Forestry, Universiti Malaysia Sabah, Jalan UMS, 88400 Kota Kinabalu, Sabah, Malaysia

Received 31 January 2014. Revised 28 May 2014. Accepted 2 June 2014. Available online 3 July 2014. 

Highlights

• •
  AGB was estimated using small-footprint LiDAR in a mountainous tropical rainforest.
• •
  We determined forest degradation levels through community composition and AGB.
• •
  Laser penetration rate contributed to improve the model accuracy for AGB estimation.
• •
  LiDAR data can capture AGB variability attributed to fine scale human disturbances.

Abstract

Deforestation and degradation of forests have severely depleted carbon storage in tropical countries, whose forests have the most carbon-rich ecosystems in the world. Estimating above-ground biomass (AGB) with high accuracy is critical to quantifying carbon stocks in the tropics. We propose a model to estimate AGB in the tropical montane forests of northern Borneo with different disturbance histories using airborne LiDAR data. The level of forest degradation was determined from species composition and field-observed AGB. Of 50 sample plots established in forests with various levels of degradation, we categorized 20 as highly degraded (AGB: 52.18–229.11 Mg/ha), 16 as moderately degraded (AGB: 136.00–382.59 Mg/ha), and 14 as old-growth forest (AGB: 280.31–622.79 Mg/ha). Height metrics and laser penetration rate (LP) at specific heights from the ground were derived from vertical point profiles of LiDAR data. After testing the performance of single variables, we used stepwise multiple regressions to select variables to include in the model for AGB estimation. The best model with a single variable used the mean height from the laser returns (R²= 0.78, RMSE = 65.54 Mg/ha). All LP variables were sensitive to AGB (R² > 0.60). The final model from stepwise analysis included the mean height of the canopy height model and LP at 7 m height (adjusted R² = 0.81, RMSE = 61.26 Mg/ha). The results confirm the suitability of LP variables for estimating AGB. We suggest that airborne LiDAR data can capture AGB variability at fine spatial scales, which correspond to deforestation and forest degradation caused by human activities and natural disturbances.

Keywords

Tropical rainforest

Above-ground biomass

LiDAR

Forest degradation

Borneo

REDD+

---

Fig. 1.

Fig. 2.

 Table 1

Table 1.

 Table 2

Table 2.

Fig. 3.

Fig. 4.

• ⁎ 
  Corresponding author. Tel.: +81 3 5841 8247; fax: +81 3 5841 5235.
  
  

For further details log on website :
http://www.sciencedirect.com/science/article/pii/S0378112714003648