ESTIMATION OF MANGRROVE CARBON STOCKS USING SENTINEL 2A AND FIELD- BASED DATA IN TIEN LANG DISTRICT, HAI PHONG CITY

Abstract

Coastal mangroves along the coastal zone of Tien Lang were selected to estimate above-ground, underground carbon stocks and soil organic carbon. Field data was conducted to measure mangrove structures in 15 plots with a dimension of 900 m², each plot has 3 subplots with a dimension of 100 m² at three communes, Dong Hung, Tien Hung, and Vinh Quang. To select the most suitable classification method for Tien Lang district, this study used some classification methods, namely NDVI, SAVI, TVI indices and unsupervised classification approach. The accuracies of NDVI, SAVI, TVI and Unsupervised classification were assessed at 85.6%; 67.5%; 58.3%; and 81.8%, respectively. The NDVI index confirmed is the most suitable for mangrove mapping (85.6%), followed by Unsupervised classification (81.8%). A total of carbon stocks was calculated from 42.78 ton ha^-1 to 117.95 ton ha^-1 for mixed mangrove forests, while monoculture Sonneratia caseolaris ranged from 62.38 ton ha^-1 to 389.9 ton ha^-1. Soil organic carbon stocks in 100 cm depth was estimated ranging from 100.95 ton ha^-1 to 227.65 ton ha^-1. IDW interpolation of above-ground biomass had accuracy at 61.1%, while soil organic carbon was at 89.3%. Overall, mangrove forests in Tien Lang offer to potential amount of carbon stocks both above-ground carbon stocks and soil organic carbon for PFES.