Mt. Apo Natural Park in the Southern Philippines Using Terrestrial LiDAR System

  • Ligaya C. Rubas-Leal Texas A&M University
  • Robert A. Washington-Allen Texas A&M University
  • Sorin C. Popescu Texas A&M University
  • J. Richard Conner Texas A&M University
Keywords: above-ground biomass, height bins, Mt. Apo Natural Park, terrestrial LiDAR system


Extraction of plot-level field measurements entails a rigid and time-consuming task. Fine resolution remote sensing technology offers an objective and consistent method for estimation of forest vertical structures. We explored the development of algorithms for estimating above ground biomass (AGB) at the plot level using terrestrial LiDAR system (TLS). This research follows IPCC Tier 2 approach, by combining field and remote sensing data, in estimating forest carbon stocks. Permanent plots (30 × 30 m diameter) were established inside Mt. Apo Natural Park. Forest inventory was conducted in July 2013, recording tree heights and stem diameters for all hardwood species with diameter at breast height (DBH) ≥ 5 cm in three management zones: multiple use, strict protection, and restoration. Quadratic mean stem diameter was employed for large DBH intervals for deriving midpoint biomass. Three tropical allometric equations were used to derive referenced biomass values. Regressions results showed satisfactory modeling fit in relating plot-level AGB to DBH class size: 80%–89%. Mean tree heights from field and TLS data were related showing R2 = 88%. TLS variables derived include percentile heights and normalized height bins at 5-m intervals. The generalized linear model is a more robust model for percentile heights, while stepwise regression showed a better regression performance for normalized height bins. Strict protection zone contained the highest carbon storage. This study demonstrated the significant TLS-derived metrics to assess plot-level biomass. TLS scanning is also the first work to be done in this ASEAN Natural Heritage Park, which is constrained with local insurgency problems. Biomass in plot-level can be used to extrapolate to landscape-level using available multispectral or radar imagery.

Research Articles