Remote sensing-based mangrove blue carbon assessment in the Asia-Pacific: A systematic review

• Published in: The Science of the total environment Vol. 938; p. 173270
• Main Authors: Dutta Roy, Abhilash, Pitumpe Arachchige, Pavithra S., Watt, Michael S., Kale, Apoorwa, Davies, Mollie, Heng, Joe Eu, Daneil, Redeat, Galgamuwa, G.A. Pabodha, Moussa, Lara G., Timsina, Kausila, Ewane, Ewane Basil, Rogers, Kerrylee, Hendy, Ian, Edwards-Jones, Andrew, de-Miguel, Sergio, Burt, John A., Ali, Tarig, Sidik, Frida, Abdullah, Meshal, Pandi Selvam, P., Jaafar, Wan Shafrina Wan Mohd, Alawatte, Isuru, Doaemo, Willie, Cardil, Adrián, Mohan, Midhun
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 15.08.2024
• Subjects: Above-ground mangrove biomass; Coastal ecosystems; Indo-Pacific; Machine learning; Mangrove carbon; Multispectral sensors; Above-ground mangrove biomass; Multispectral sensors; Mangrove carbon; Indo-Pacific; Machine learning; Coastal ecosystems
• ISSN: 0048-9697; 1879-1026; 1879-1026
• DOI: 10.1016/j.scitotenv.2024.173270

Accurate measuring, mapping, and monitoring of mangrove forests support the sustainable management of mangrove blue carbon in the Asia-Pacific. Remote sensing coupled with modeling can efficiently and accurately estimate mangrove blue carbon stocks at larger spatiotemporal extents. This study aimed to identify trends in remote sensing/modeling employed in estimating mangrove blue carbon, attributes/variations in mangrove carbon sequestration estimated using remote sensing, and to compile research gaps and opportunities, followed by providing recommendations for future research. Using a systematic literature review approach, we reviewed 105 remote sensing-based peer-reviewed articles (1990 - June 2023). Despite their high mangrove extent, there was a paucity of studies from Myanmar, Bangladesh, and Papua New Guinea. The most frequently used sensor was Sentinel-2 MSI, accounting for 14.5 % of overall usage, followed by Landsat 8 OLI (11.5 %), ALOS-2 PALSAR-2 (7.3 %), ALOS PALSAR (7.2 %), Landsat 7 ETM+ (6.1 %), Sentinel-1 (6.7 %), Landsat 5 TM (5.5 %), SRTM DEM (5.5 %), and UAV-LiDAR (4.8 %). Although parametric methods like linear regression remain the most widely used, machine learning regression models such as Random Forest (RF) and eXtreme Gradient Boost (XGB) have become popular in recent years and have shown good accuracy. Among a variety of attributes estimated, below-ground mangrove blue carbon and the valuation of carbon stock were less studied. The variation in carbon sequestration potential as a result of location, species, and forest type was widely studied. To improve the accuracy of blue carbon measurements, standardized/coordinated and innovative methodologies accompanied by credible information and actionable data should be carried out. Technical monitoring (every 2–5 years) enhanced by remote sensing can provide accurate and precise data for sustainable mangrove management while opening ventures for voluntary carbon markets to benefit the environment and local livelihood in developing countries in the Asia-Pacific region. [Display omitted] •The potential of remote sensing in assessing blue carbon needs further exploration.•Fewer studies from mangrove-rich countries (Myanmar, Bangladesh, and Papua New Guinea)•Sentinel-2 MSI (14.5 % of overall usage) was the most used sensor.•Research on below-ground carbon and valuation of carbon stock is limited.•Improvement in accuracy and precision based on innovative methodologies is needed.