The application of Unmanned Aerial Vehicles (UAVs) to estimate above-ground biomass of mangrove ecosystems
Introduction
Mangrove ecosystems provide unparalleled economically and ecologically critical services to coastal areas (Alongi, 2008), including: 1) provisioning (Hemminga and Duarte, 2000); 2) coastal protection (Badola and Hussain, 2005; Das and Vincent, 2009; Koch et al., 2009); 3) recreational and aesthetic uses (Bergstrom et al., 1990); and 4) soil formation and carbon sequestration (Atwood et al., 2017; Donato et al., 2011; Mcleod et al., 2011). Despite their importance they are one of the most threatened and vulnerable ecosystems worldwide (Hamilton and Casey, 2016; Thomas et al., 2017).
As they occur across the land-sea interface, mangrove forests are subject to both terrestrial and marine pressures. A large percentage of mangrove ecosystems have been altered, destroyed or degraded worldwide as a result of anthropogenic impacts, particularly near populated areas (Clark and Johnston, 2016; Richards and Friess, 2016). Consequently, mangrove ecosystems are the focus for many conservation and rehabilitation efforts (Duncan et al., 2016; Nam et al., 2016; Ren et al., 2010), which require regular monitoring of the forest structure characteristics for better management strategies. This is particularly relevant for any projects seeking to gain carbon credits for mangrove rehabilitation under the impacts of climate change related issues (sea level rise and changes in precipitation and temperature; Ward et al., 2016).
Mangrove forest monitoring and management has traditionally relied on regular on-ground surveys for collecting forest inventory data (Nam et al., 2016; Ren et al., 2010). This data collection is important because it provides a better understanding of species composition and carbon biomass of the survey areas. However, these surveys usually cover relatively small spatial scales (<0.5 ha), and can often be expensive, labour intensive and time-consuming due to the tides, mud, and general difficulty in accessing these remote coastal ecosystems (Lee and Lunetta, 1995).
Remote Sensing data, on the other hand, provides a fast, cost-effective, and efficient method to estimate the biological, biophysical and biochemical factors that translate into some of the services provided by mangrove ecosystems (Giri, 2016; Pham et al., 2019). The type of remote sensing platform (ground-based, airborne or satellite) and sensor (photographic, LiDAR or radar) used depend on the scale and the goal of the research (Wang et al., 2019). Remotely sensed satellite archive data offer broad-scale (nation-wide) and long-term (up to 40 years) monitoring for detection of change over time (Giri, 2016; Kuenzer et al., 2011; Pham et al., 2019). However, it might not necessarily provide the resolution and precision required for local accounting (Ruwaimana et al., 2018). To increase spatial resolution while still covering large areas, airborne LiDAR systems and ground-based platforms like terrestrial laser scanners have been used in the past for retrieval of forest inventory data of coastal wetlands (Feliciano et al., 2014; Wannasiri et al., 2013). However, they are expensive, not widely available, and data management and processing is often difficult and requires specialized software (Wallace et al., 2016; Yin and Wang, 2019).
Unmanned aerial vehicles (UAVs) paired with Structure from Motion and Multi-View Stereo photogrammetric procedures (from now on UAV–SfM) have the potential to: 1) increase fieldwork efficiency by collecting broader spatial information in less time than traditional ground-based surveys (Dandois and Ellis, 2013; Messinger et al., 2016; Murfitt et al., 2017), 2) increase spatial resolution obtained from satellite data while still covering large areas (Ruwaimana et al., 2018), and 3) provide a more cost-effective approach than other airborne systems like LiDAR or airborne photogrammetry while maintaining accuracy and resolution (Dustin, 2015; Sankey et al., 2017).
In terrestrial forests, the implementation of high-resolution imagery collected from low-cost UAVs is becoming an increasingly valuable tool for mapping above-ground carbon stock (Dandois and Ellis, 2010; Samiappan et al., 2016; Zahawi et al., 2015). Common forest inventory data derived from UAV-SfM includes tree species, height, canopy diameter and above-ground biomass (AGB), which can complement, and eventually replace, traditional forest inventory techniques (Dandois and Ellis, 2013; Messinger et al., 2016; Panagiotidis et al., 2017; Zarco-Tejada et al., 2014). Overall, there is a general consensus that UAVs are the most cost-effective solution for sites with an extent between 10 and 20 ha when compared to aircraft and satellite data (Dustin, 2015; Manfreda et al., 2018; Matese et al., 2015). However, no cost-benefit analysis has compared the benefits of using UAVs against on-ground measurements in mangrove ecosystems to date.
Despite the importance of coastal wetlands for carbon accumulation and other ecosystem services, there are few studies evaluating the use of UAV-SfM approaches for assessing the biophysical and biochemical properties of mangrove forests (Li et al., 2016; Navarro et al., 2019; Otero et al., 2018; Tian et al., 2017; Yaney-Keller et al., 2019). Out of these, only two have focused on retrieving mangrove forest inventory data for estimating mangrove AGB (Navarro et al., 2019; Otero et al., 2018). These studies demonstrated that UAV-SfM derived data has the potential for estimating AGB from mangrove plantations, but were unable to predict AGB from natural forests with densely packed mangrove trees. Otero et al. (2018) retrieved information on height from UAV-SfM data at a plantation and natural site with mixed results (only the height medians from the plantation site were significantly similar). Additionally, validation was achieved through visual interpretation of the orthomosaic, with no direct plot by plot comparison between UAV-SfM and field data. On the other hand, Navarro et al. (2019) managed to perform a tree by tree comparison of UAV-SfM data vs field data for two measurements: height and canopy diameter. However, this study is based on trees from a mangrove plantation project <8 years old with clearly defined boundaries between trees. Furthermore, only canopy diameter was not significantly different from on-ground measurements. Consequently, the challenge to effectively estimate mangrove AGB from UAV-SfM derived forest inventory data still remains.
In this study, we propose an approach that combines tree detection and canopy segmentation algorithms applied to UAV-SfM data for quantifying AGB of mangrove forests within natural and rehabilitated (25+ years) areas of the southeastern coast of Australia, and compare it to on-ground measurements at the plot level. Moreover, we perform the first cost-benefit analysis to date of the two different methods (field based measurements vs UAV-SfM data) in mangrove ecosystems to estimate above-ground biomass. The methods described in this study provide local managers with a cost-effective approach for regular monitoring of mangrove forests over larger areas than traditional on-ground surveys while maintaining forest inventory data accuracy.
Section snippets
Study area
We focused our research on two areas of the southeastern coast of Australia: Western Port (WP) in southern Victoria (Fig. 1b) and Richmond River Estuary (RRE) in northern New South Wales (Fig. 1c). WP is a large tidal bay that covers around 680 km2, of which 270 km2 are exposed mudflats at low tide. Mangrove ecosystems in WP are located near the southernmost distribution of mangrove ecosystems in the world and cover an area of around 1800 ha (Boon et al., 2011. Fig. 1d). Air temperatures in WP
UAV-SfM derived point clouds and Canopy Height Models
An average of 250 images per hectare were used to generate the dense point clouds using SfM-MVS procedures. An orthomosaic image, DSM, DTM and CHM were generated for every survey site with an average pixel resolution of 3.2 cm (Fig. 4). The geometric accuracies of the scene reconstructions averaged an RMS horizontal error (x, y) of 1.0 cm and RMS vertical error (z) of 1.2 cm. DTM generation averaged a RMSE vertical error of 6.3 cm for WP, 9.0 cm for RRE (rehabilitated areas) and 44.7 cm for RRE
Discussion
This study has demonstrated the ability of UAV-SfM data to estimate AGB of mangrove ecosystems. Low-cost UAV imagery was used for the creation of orthomosaic images, DSM, DTM and CHMs for nine survey sites in the southeastern coast of Australia using Structure from Motion procedures. There was a close correspondence between UAV-SfM derived tree heights, average canopy diameters, tree density and ultimately AGB estimates with those measured in the field. Other studies have found that UAV-SfM
Conclusions
Through this study, we showed that low-cost UAV-SfM provide an accurate and efficient method for assessing the above-ground biomass of mangrove forests in temperate and sub-tropical regions. We were able to obtain accurate tree height and canopy diameter metrics from UAV-SfM data that had a high correspondence with on-ground based data. The methods described in this study opens the possibility for easily repeatable, low-cost UAV surveys for local managers by providing a faster, more
CRediT authorship contribution statement
Alejandro Navarro:Conceptualization, Methodology, Software, Formal analysis, Investigation, Writing - original draft.Mary Young:Conceptualization, Writing - review & editing, Supervision.Blake Allan:Methodology, Writing - review & editing.Paul Carnell:Investigation, Data curation, Writing - review & editing.Peter Macreadie:Writing - review & editing, Supervision, Funding acquisition.Daniel Ierodiaconou:Conceptualization, Writing - review & editing, Supervision, Project administration, Funding
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This study is part of the Mapping Ocean Wealth project from The Nature Conservancy's Great Southern Seascapes program and supported by The Thomas Foundation, HSBC Australia, the Ian Potter Foundation, and Victorian and New South Wales governments including Parks Victoria, Department of Environment Land Water and Planning, Victorian Fisheries Authority, New South Wales Office of Environment and Heritage, and New South Wales Department of Primary Industries. Funding was also provided by an
References (68)
Mangrove forests: resilience, protection from tsunamis, and responses to global climate change
Estuarine Coastal and Shelf Science
(2008)- et al.
Economic value of wetlands-based recreation
Ecol. Econ.
(1990) - et al.
High spatial resolution three-dimensional mapping of vegetation spectral dynamics using computer vision
Remote Sens. Environ.
(2013) - et al.
Rehabilitating mangrove ecosystem services: a case study on the relative benefits of abandoned pond reversion from Panay Island, Philippines
Mar. Pollut. Bull.
(2016) - et al.
Estimation of aboveground biomass of different mangrove trees based on canopy diameter and tree height
Procedia Environ. Sci.
(2011) - et al.
A low-cost multi-sensoral mobile mapping system and its feasibility for tree measurements
ISPRS J. Photogramm. Remote Sens.
(2010) - et al.
Remote estimation of canopy height and aboveground biomass of maize using high-resolution stereo images from a low-cost unmanned aerial vehicle system
Ecol. Indic.
(2016) - et al.
Managing mangrove forests from the sky: forest inventory using field data and Unmanned Aerial Vehicle (UAV) imagery in the Matang Mangrove Forest Reserve, peninsular Malaysia
For. Ecol. Manag.
(2018) - et al.
Spatial variation of above-ground carbon storage in temperate coastal wetlands
Estuar. Coast. Shelf Sci.
(2018) - et al.
UAV lidar and hyperspectral fusion for forest monitoring in the southwestern USA
Remote Sens. Environ.
(2017)
Comparison of UAV and WorldView-2 imagery for mapping leaf area index of mangrove forest
Int. J. Appl. Earth Obs. Geoinf.
Decomposition of leaf and root litter of Avicennia marina at Westernport Bay, Victoria, Australia
Aquat. Bot.
Multi-temporal imaging using an unmanned aerial vehicle for monitoring a sunflower crop
Biosyst. Eng.
‘Structure-from-motion’ photogrammetry: a low-cost, effective tool for geoscience applications
Geomorphology
Individual mangrove tree measurement using UAV-based LiDAR data: possibilities and challenges
Remote Sens. Environ.
Using lightweight unmanned aerial vehicles to monitor tropical forest recovery
Biol. Conserv.
Tree height quantification using very high resolution imagery acquired from an unmanned aerial vehicle (UAV) and automatic 3D photo-reconstruction methods
Eur. J. Agron.
Free as a drone: ecologists can add UAVs to their toolbox
Front. Ecol. Environ.
Global patterns in mangrove soil carbon stocks and losses
Nat. Clim. Chang.
Valuing ecosystem functions: an empirical study on the storm protection function of Bhitarkanika mangrove ecosystem, India
Environ. Conserv.
Imager: Image Processing Library Based on ‘CImg’. R Package Version 0.41.1, 2
Estimating biomass of barley using crop surface models (CSMs) derived from UAV-based RGB imaging
Remote Sens.
Victorian Saltmarsh Study: Mangroves and Coastal Saltmarsh of Victoria: Distribution, Condition, Threats and Management
Remote sensing of vegetation structure using computer vision
Remote Sens.
Mangroves protected villages and reduced death toll during Indian super cyclone
Proc. Natl. Acad. Sci. U. S. A.
Mangroves among the most carbon-rich forests in the tropics
Nat. Geosci.
Monitoring Parks with Inexpensive UAVs: Cost Benefits Analysis for Monitoring and Maintaining Parks Facilities
Height and biomass of mangroves in Africa from ICESat/GLAS and SRTM
Int. J. Remote Sens.
Assessing mangrove above-ground biomass and structure using terrestrial laser scanning: a case study in the Everglades National Park
Wetlands
Observation and monitoring of mangrove forests using remote sensing: opportunities and challenges
Remote Sens.
Creation of a high spatio-temporal resolution global database of continuous mangrove forest cover for the 21st century (CGMFC-21)
Glob. Ecol. Biogeogr.
Seagrass Ecology
Lakemorpho: Lake Morphometry Metrics
Cited by (106)
UAV-RGB-image-based aboveground biomass equation for planted forest in semi-arid Inner Mongolia, China
2024, Ecological InformaticsPrediction of cotton FPAR and construction of defoliation spraying prescription map based on multi-source UAV images
2024, Computers and Electronics in AgricultureAutomatic detection tree crown and height using Mask R-CNN based on unmanned aerial vehicles images for biomass mapping
2024, Forest Ecology and ManagementMangrove individual tree detection based on the uncrewed aerial vehicle multispectral imagery
2024, Remote Sensing Applications: Society and Environment