Integrating Remote Sensing, Field-Measured Tree Heights, and Machine Learning to Enhance Mangrove Above-Ground Carbon Estimation in Baluran National Park, Indonesia

Mangroves make an important contribution to coastal ecosystem services, including the absorption of large amounts of atmospheric carbon, thereby contributing to climate change mitigation. Developing a mangrove carbon model to better understand and monitor mangrove condition and carbon stores at relevant scales is crucial. The study purpose is to estimate mangrove Above-Ground Carbon (AGC) at Baluran National Park using variables from field measurements and remote sensing combined with a Machine Learning (ML) approach. The study utilised an extensive field data collection programme of 60 sampling plots of girth at breast height, canopy cover, tree height, and tree density. Mangrove AGC was estimated using allometric equations. AGC was also modelled by processing satellite images, conducting statistical analyses, developing models with ML algorithms (Random Forest (RF), Support Vector Machine (SVM), Decision Tree (DT), k-Nearest Neighbour (k-NN), and Gradient Boost (GB)), and checking accuracy using 5-fold cross-validation (CV) of Root Mean Square Error (RMSE). The RF model using Red Edge 3 band, Green Normalised Difference Vegetation Index (GNDVI) from Sentinel-2A (S2A), and field-measured tree height achieved the best estimation of sampled AGC (R² training = 0.93, R² testing = 0.86, 5-fold CV RMSE = 8.27 Mg C ha⁻¹). We estimate mangrove AGC values ranging from 8.89 to 46.20 Mg C ha⁻¹ (average value: 27.42 ± 10.47 Mg C ha⁻¹), which is more accurate than global datasets representing the same locations. This study demonstrates that combining field-measured tree height with high-resolution satellite imagery and appropriate ML algorithms significantly improves mangrove AGC estimations.