Mangrove above-ground carbon estimation from Sentinel-1A (SAR) and field-based data in Tien Yen district, Quang Ninh province

• DOI: 10.55250/jo.vnuf.9.1.2024.073-085
• Publication Date: 2024-05-15T09:25:37Z
• Authors: Nguyen Hai Hoa; Vu Van Truong; Nguyen Thi Thu Hien; Ha Tri Son; Nguyen Van Thi; Nguyen Trong Cuong; Nguyen Thi Bich Hao; Tran Thi Huong; Phan Duc Le; Phan Van Dung; Thai Thi Thuy An; Le Phu Tuan
• Keywords: Deforestation; Land Cover

Mangrove forests have been globally recognized as they play a vital role in preventing coastal erosion, mitigating the effects of wave actions, and protecting habitats adjacent shoreline land-uses from extreme events. Sentinel-1 (SAR) offers new opportunity for mangrove cover mapping biomass estimation, especially tropics where deforestation degradation are highest cloud is persistent. This study used Sentinel-1A-derived VV/VH polarizations with thresholds -23.5<VH<-9.05 -17.5<VV<-3.8. Upon using VV VH compared to PlanetScope data, it has confirmed that these suitable monitoring along coast Tien Yen an overall accuracy over 90.5% Kappa coefficient greater than 0.78 2022. also developed AGB models based on field survey data SAR estimating Yen. In fact, we evaluated capability Sentinel-1A retrieval predictive through conventional linear regression models. The findings show polarization values can be estimation. Overall, selected Model 1 R2=0.445 (p-value<0.001) provided option carbon To more accurate this suggests research should carried out advanced machine learning Sentinel-1B estimation