Mangroves are crucial in providing ecological and provisioning services, such as sequestering atmospheric CO2 (Kumari & Rathore, 2021) and providing habitat for shorebirds and fish (Buelow & Sheaves, 2015). However, they have been lost worldwide due to human disturbance, particularly in Southeast Asia (Fauzi et al., 2019). Sustainable mangrove forest replantation and management are essential. Mangroves have biophysical parameters related to ecosystem health and dynamics (Sani et al., 2018). Accurate measurements of these parameters can be achieved using field investigations, but this method is laborious and time-consuming due to the complexity of the mangrove environment and dangerous fauna (Saintilan et al., 2022).

Remote sensing offers a complementary tool for mangrove carbon measurements, offering a synoptical overview, spectral and spatial resolution, and ease of data capture (Tran et al., 2022). Using remotely sensed optical and Synthetic Aperture Radar (SAR) images has been successfully applied to develop mangrove carbon models. However, there is a need for improved accuracy in mangrove carbon models (Wang et al., 2019).

This study aims to test a novel Machine Learning (ML) method proposed by (Pham et al., 2020) to map and quantify mangrove above-ground carbon (AGC) in Indonesian mangroves using multisource free-of-charge remotely sensed datasets. The model integrates extreme gradient boosting regression (XGB) and genetical gorithm (GA) to map AGB mangroves in Northern Vietnam using optical and SAR data combined with field sampling.

This poster was presented during the 10th APN Early Career Professional Poster and Networking Session, held alongside the APN 26th Joint Intergovernmental and Scientific Planning Group Meetings on 13 June 2024 at the National Research and Innovation Agency (BRIN) in Jakarta, Indonesia.