Predictive Model of Mangroves Carbon Stocks in Kedah, Malaysia using Remote Sensing

• Published in: IOP Conference Series: Earth and Environmental Science
• Main Author: Hashim T.M.Z.T.; Suratman M.N.; Singh H.R.; Jaafar J.; Bakar A.N.
• Format: Conference paper
• Language: English
• Published: Institute of Physics Publishing 2020
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85090141683&doi=10.1088%2f1755-1315%2f540%2f1%2f012033&partnerID=40&md5=00785edb19c3d289db610ca87c9253eb

Mangroves are recognized as an ecosystem that grow and dominate the coastal areas of tropical and sub-tropical regions across the world. They not only provide ecological and socio-economic support, but also play a pivotal role in offsetting an excess of carbon from the atmosphere. Despite the crucial roles provided by mangroves, the ecosystem has degraded at an alarming rate mainly due to anthropogenic activities. Remote sensing technology provides a new dimensional perspective in monitoring and estimating tree biomass and carbon stocks. Therefore, this study aimed at (1) estimating the carbon stocks of mangroves in Kedah, Malaysia, (2) investigating the relationships between mangrove stand parameters with spectral reflectance recorded from Landsat 8 Operational Land Imager (OLI) data, and (3) developing predictive models for estimating the carbon stocks of mangroves by combining the ground and Landsat 8 (OLI) data. For the purpose of this study, a total of 81 mangrove stand data sets measuring at 100 m × 100 m were collected throughout Kedah, Malaysia. Within the stand, seven randomly selected plots were established and all individual mangroves parameter (diameter at breast height (DBH) and height) were measured. The 81 stands were split into two independent data sets for developing and validating the models (56 and 25 stands, respectively). Multiple regression technique with least square approach was used in the model development process. From several good candidate models, a model consists of four predictive variables (bands 3 and 6, Normalized Difference Vegetation Index (NDVI) and simple ratio seems to be predicting reasonably well based on its simplicity and practicality (p≤0.001, R2 = 0.56). Validation of the model has resulted in Mallow's prediction criterion (Cp) value of 4.28 and Root Mean Squared Error (RMSE) of 4.11 Mg/ha. The information from this study may provide useful input for future research and can be crucial tools for the government and stakeholders in future decision making for the sustainability of mangrove resources. © Published under licence by IOP Publishing Ltd.