Assessment of Biomass and Carbon Content in a Mediterranean Aleppo Pine Forest Using ALS Data

Antonio Luis Montealegre,Alberto García-Martín,Francisco Escribano,María Teresa Lamelas,Juan De La Riva

doi:10.3390/ecrs-1-d004

Antonio Luis Montealegre, Alberto García-Martín + Show 3 more

Open Access

https://doi.org/10.3390/ecrs-1-d004

Copy DOI

Abstract

Tree biomass estimate is essential for carbon accounting, bioenergy feasibility studies, and forest sustainable management. This fact, added to the availability of airborne laser scanning (ALS) information, provided by the Spanish National Plan for Aerial Orthophotography (PNOA), and the existence of little research focusing on the use of ALS technology in Mediterranean Aleppo pine (Pinus halepensis Mill.) forest, determined the main objective of this research. Thus, this study aims to test the suitability of the low point density (0.5 points/m2), discrete, multiple-return, PNOA-ALS data, to estimate and map the total biomass (TB) and its carbon content in Pinus halepensis Mill. forest stands, located in Aragon (north-eastern Spain). TB was calculated in 45 field plots, using allometric equations, and related through a multivariate linear regression analysis with a collection of independent variables extracted from the ALS data. The predictive model was validated using a leave-one-out cross-validation (LOOCV) technique. Then, a regular grid with cell size 25 x 25 m corresponding to the sample plot size was generated by means of GIS, in order to compute TB at stand level and convert biomass to carbon by using the 0.5 conversion factor. The maximum height, kurtosis and the percentage of returns above 1 meter, were the ALS metrics included in the fitted model, which presented a R2 value of 0.89. The implementation of the model in a GIS showed an average of 68633 kg/ha of TB and 34247.95 kg/ha of carbon fixed. The results indicate that despite the low point density of the ALS data, the final model is accurate enough to be used in forestry applications.

Highlights

In the last decades, operational collection of information on relevant characteristics of forest ecosystems by means of pure ground-based field inventories has been revolutionized by the development of remote sensing sensors [1]
Structural variables related to Aleppo pine forest show a remarkable variability
This implies a subsequent variability in biomass estimations from these variables, as can be seen in standard deviation values of aboveground biomass (AGB), Wr and total biomass (TB)

Summary

Introduction

Operational collection of information on relevant characteristics of forest ecosystems by means of pure ground-based field inventories has been revolutionized by the development of remote sensing sensors [1]. Optical and radar remote sensing have been widely used to map forest structural attributes and biophysical parameters [2,3,4,5]. Most commercial LiDAR systems are small-footprint, discrete-return airborne lasers, referred to as Airborne Laser Scanning (ALS) These systems are able to accurately characterize the threedimensional structure of the forest canopy due to its capacity to record the different reflections of each emitted pulse, corresponding to both, vegetation and terrain beneath it [6,8]. In order to generate a precise Digital Elevation Model (DEM), it is necessary to filter the ALS data and to interpolate the points classified as ground These DEMs are needed to normalize the heights of the ALS point cloud with respect to the terrain heights [9,10,11]

Methods

Results

Conclusion