Scholarlink

Abstract <p>Mangrove ecosystems are vital for coastal protection and carbon storage, yet extensive degradation highlights the urgency of targeted rehabilitation. This study developed a spatially explicit habitat suitability model using freely available geospatial datasets to identify potential mangrove rehabilitation areas in the Mahakam Delta, East Kalimantan. The model integrated elevation, inundation frequency, and shoreline proximity within a random forest framework, achieving strong predictive performance (AUC = 0.898). Uncertainty analysis revealed higher variability in the lower delta regions, highlighting the need for additional field validation, while spatial autocorrelation analysis confirmed significant clustering (Moran’s I = 0.7604, p &lt; 0.001), demonstrating spatial consistency in predictions. High-probability zones accounted for 14.58% of the total predicted rehabilitation area but overlapped with 79.02% of historical mangrove loss between 1995 and 2024, indicating substantial restoration potential. The resulting predictive map provides an evidence-based tool for prioritizing rehabilitation planning and fostering collaboration among researchers, policymakers, and stakeholders. Future improvements should incorporate real-time environmental data and participatory web-based platforms to enhance accessibility, adaptability, and scalability in advancing data-driven mangrove rehabilitation and sustainable coastal ecosystem restoration.</p> Highlights <p><UnorderedList Mark="Bullet"> <ItemContent> <p>A spatial mangrove habitat model using open-access geospatial data identifies potential rehabilitation areas.</p> </ItemContent> <ItemContent> <p>Random Forest algorithm effectively mapped suitable rehabilitation sites with high predictive accuracy (AUC = 0.898).</p> </ItemContent> <ItemContent> <p>High-probability areas can restore 79.02% of mangrove loss between 1995 and 2024 in the study area of Mahakam Delta, Indonesia, showing strong rehabilitation potential.</p> </ItemContent> <ItemContent> <p>Model validation and uncertainty analysis emphasize the importance of field assessment in specific regions.</p> </ItemContent> <ItemContent> <p>The predictive map offers a robust guide for collaborative mangrove rehabilitation planning and adaptive management.</p> </ItemContent> </UnorderedList></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 Practical Framework for Mangrove Rehabilitation Sites Selection Using Scalable Biophysical Indicators

  • 期刊论文

  • 发表日期：

    2026年2月9日

查看全文
  • Bayu Prayudha,
  • Yaya I. Ulumuddin,
  • Suyadi Suyadi,
  • Suyarso Suyarso,
  • Arif Seno Adjie,
  • Udhi E. Hernawan,
  • Muhammad R. Ramadhani

摘要

Abstract

Mangrove ecosystems are vital for coastal protection and carbon storage, yet extensive degradation highlights the urgency of targeted rehabilitation. This study developed a spatially explicit habitat suitability model using freely available geospatial datasets to identify potential mangrove rehabilitation areas in the Mahakam Delta, East Kalimantan. The model integrated elevation, inundation frequency, and shoreline proximity within a random forest framework, achieving strong predictive performance (AUC = 0.898). Uncertainty analysis revealed higher variability in the lower delta regions, highlighting the need for additional field validation, while spatial autocorrelation analysis confirmed significant clustering (Moran’s I = 0.7604, p < 0.001), demonstrating spatial consistency in predictions. High-probability zones accounted for 14.58% of the total predicted rehabilitation area but overlapped with 79.02% of historical mangrove loss between 1995 and 2024, indicating substantial restoration potential. The resulting predictive map provides an evidence-based tool for prioritizing rehabilitation planning and fostering collaboration among researchers, policymakers, and stakeholders. Future improvements should incorporate real-time environmental data and participatory web-based platforms to enhance accessibility, adaptability, and scalability in advancing data-driven mangrove rehabilitation and sustainable coastal ecosystem restoration.

Highlights

A spatial mangrove habitat model using open-access geospatial data identifies potential rehabilitation areas.

Random Forest algorithm effectively mapped suitable rehabilitation sites with high predictive accuracy (AUC = 0.898).

High-probability areas can restore 79.02% of mangrove loss between 1995 and 2024 in the study area of Mahakam Delta, Indonesia, showing strong rehabilitation potential.

Model validation and uncertainty analysis emphasize the importance of field assessment in specific regions.

The predictive map offers a robust guide for collaborative mangrove rehabilitation planning and adaptive management.