Drones as a Tool for Monoculture Plantation Assessment in the Steepland Tropics

Ethan Miller,Matteo Detto,Jonathan Dandois,Jefferson Hall

doi:10.3390/f8050168

Ethan Miller, Matteo Detto + Show 2 more

Open Access

https://doi.org/10.3390/f8050168

Copy DOI

Journal: Forests	Publication Date: May 12, 2017
Citations: 16	License type: CC BY 4.0

Affiliation: Smithsonian Tropical Research Institute, Yale University

Abstract

Smallholder tree plantations are expanding in the steepland tropics due to demand for timber and interest in ecosystem services, such as carbon storage. Financial mechanisms are developing to compensate vegetation carbon stores. However, measuring biomass—necessary for accessing carbon funds—at small scales is costly and time-intensive. Therefore, we test whether low-cost drones can accurately estimate height and biomass in monoculture plantations in the tropics. We used Ecosynth, a drone-based structure from motion technique, to build 3D vegetation models from drone photographs. These data were filtered to create a digital terrain model (DTM) and digital surface model (DSM). Two different canopy height models (CHMs) from the Ecosynth DSM were obtained by subtracting terrain elevations from the Ecosynth DTM and a LIDAR DTM. We compared height and biomass derived from these CHMs to field data. Both CHMs accurately predicted the height of all species combined; however, the CHM from the LiDAR DTM predicted heights and biomass on a per-species basis more accurately. Height and biomass estimates were strong for evergreen single-stemmed trees, and unreliable for small leaf-off species during the dry season. This study demonstrates that drones can estimate plantation biomass for select species when used with an accurate DTM.

Highlights

Tropical forest plantations play a major role in global timber markets
For the drone terrain model, When linear models were created to predict aboveground biomass (AGB) and total biomass (TB) for individual species, we found that the optimal height metric depended on the species and terrain model
The results demonstrate that the Ecosynth canopy height models (CHMs) and the LiDAR CHM could predict TB with nearly identical regression coefficients to AGB models

Summary

Introduction

Tropical forest plantations play a major role in global timber markets. In 2012, 231 million m3 (41%of the global total) of industrial roundwood from plantations was produced in tropical countries [1].In addition to timber, national and international institutions are looking to plantations for ecosystem services such as carbon sequestration, water filtration, and soil conservation [2,3,4,5]. Tropical forest plantations play a major role in global timber markets. Of the global total) of industrial roundwood from plantations was produced in tropical countries [1]. National and international institutions are looking to plantations for ecosystem services such as carbon sequestration, water filtration, and soil conservation [2,3,4,5]. Countries and multilateral organizations such as the United Nations are implementing vegetation carbon accounting programs to integrate carbon stores, including those in plantations, into national and global budgets. These mechanisms are increasingly making funding available to landowners and organizations that can track changes in biomass and quantify carbon stored. Assessing height directly is often cost and time-prohibitive [8]

Methods

Results

Discussion

Conclusion