Scholarlink

<p>Climate change presents significant challenges to human societies, with mangrove ecosystems playing a vital role in mitigating these impacts, particularly through carbon sequestration and coastal protection. Mangroves store carbon at rates three to five times higher than most terrestrial vegetation, positioning them as key contributors to climate change mitigation. The primary objective of this study is to develop an accurate model for estimating Above-Ground Carbon (AGC) in mangrove forests, with a specific focus on Balikpapan Bay, Indonesia, by integrating multi-sensor remote sensing data to address existing knowledge gaps in carbon stock quantification, CO₂ sequestration, and spatial mapping. The model was developed and validated using data from 20 field sample plots, which were divided equally between a calibration and an independent validation set. Ground-based measurements of mangrove structure and AGC were correlated with satellite-derived metrics from Sentinel-2 and ALOS-2 PALSAR-2. Mangrove classification was performed using a random forest algorithm, while AGC and CO₂ sequestration estimates were derived through a semi‑empirical approach. This combined methodology, integrating field observations, remote sensing, and machine learning, ensures robust model calibration and reliable assessment of mangrove carbon storage and classification accuracy. Mangrove cover was estimated at 16,523.35 hectares with high classification accuracy (Overall Accuracy: 95.4%; Kappa coefficient: 0.947). This study presents a novel integration of IRECI and DVI vegetation indices with HV polarization backscatter in a multi-sensor framework, an approach rarely applied in Indonesia or Southeast Asia. The model achieved high estimation accuracy (R² = 0.94; RMSE = 33.03 Mg/ha), with strong cross-validation confirming its robustness. As one of the most accurate and innovative methods for mangrove AGC estimation in the region, it offers significant potential for advancing blue carbon mapping and informing conservation and climate strategies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A modelling approach for Mangrove carbon dynamics using integrated optical and SAR data in Balikpapan Bay, Indonesia: algorithm development and accuracy enhancement

  • 期刊论文

  • 发表日期：

    2025年11月14日

查看全文
  • A. A. Md. Ananda Putra Suardana,
  • Nanin Anggraini,
  • Muhammad Rizki Nandika,
  • Devica Natalia Br Ginting,
  • Nugraheni Setyaningrum,
  • M. Amar Sajali,
  • Azura Ulfa,
  • Kholifatul Aziz,
  • Kuncoro Teguh Setiawan,
  • Ratih Dewanti Dimyati

摘要

Climate change presents significant challenges to human societies, with mangrove ecosystems playing a vital role in mitigating these impacts, particularly through carbon sequestration and coastal protection. Mangroves store carbon at rates three to five times higher than most terrestrial vegetation, positioning them as key contributors to climate change mitigation. The primary objective of this study is to develop an accurate model for estimating Above-Ground Carbon (AGC) in mangrove forests, with a specific focus on Balikpapan Bay, Indonesia, by integrating multi-sensor remote sensing data to address existing knowledge gaps in carbon stock quantification, CO₂ sequestration, and spatial mapping. The model was developed and validated using data from 20 field sample plots, which were divided equally between a calibration and an independent validation set. Ground-based measurements of mangrove structure and AGC were correlated with satellite-derived metrics from Sentinel-2 and ALOS-2 PALSAR-2. Mangrove classification was performed using a random forest algorithm, while AGC and CO₂ sequestration estimates were derived through a semi‑empirical approach. This combined methodology, integrating field observations, remote sensing, and machine learning, ensures robust model calibration and reliable assessment of mangrove carbon storage and classification accuracy. Mangrove cover was estimated at 16,523.35 hectares with high classification accuracy (Overall Accuracy: 95.4%; Kappa coefficient: 0.947). This study presents a novel integration of IRECI and DVI vegetation indices with HV polarization backscatter in a multi-sensor framework, an approach rarely applied in Indonesia or Southeast Asia. The model achieved high estimation accuracy (R² = 0.94; RMSE = 33.03 Mg/ha), with strong cross-validation confirming its robustness. As one of the most accurate and innovative methods for mangrove AGC estimation in the region, it offers significant potential for advancing blue carbon mapping and informing conservation and climate strategies.