Retrieving Mangrove Biophysical Properties using Drone-Based
Structure-from-Motion (SfM) Photogrammetry
Abstract
Determination of biophysical properties of mangroves, e.g. tree height
and diameter at breast height, is necessary for assessing mangrove
ecosystem dynamics and growth. However, traditional surveying methods
(e.g. diameter at breast height, location with GPS, and tree’s height
with hypsometer) are time consuming and expensive. In this research we
aim to assess mangrove properties by Unmanned Aerial Vehicles (or:
UAV/drone) based photogrammetry. The additional benefit of this
methodology is that mangrove environments which are often difficult to
access can be reached. We focus on Porong Delta, Indonesia. In this area
rapid delta progradation has taken place since 2006 due to an extreme
mud volcano eruption. The regional climate conditions and added nutrient
flux to the new delta have provided a suitable environment for growth of
different mangrove species, such as Avicennia spp., Rhizophora spp., and
Sonneratia spp. We used an off-the-shelve drone to generate a
high-resolution spatial canopy height model (CHM) for the mangroves by
using structure-from-motion (SfM) photogrammetry with an achieved DEM
resolution of 5.2cm/pixel. This approach is used to determine the
structural characteristics of mangrove stands and is validated with
ground-truthing on two deltas. The dense point cloud derived from SfM
photogrammetry is processed in LAStools and bare earth extraction with a
Cloth Simulation Filter method. The CHM is generated from the processed
point clouds. The structural information of mangroves is retrieved by
using the lidR package in R. Our analysis lead to a median tree height
on north and south delta of 4.2m and 3.5m, respectively. Analysis of
photogrammetry data shows that drone-based observations of the mangrove
canopy height are a useful tool to provide trustworthy data of mangroves
characteristics.