Machine learning-based monitoring of mangrove ecosystem dynamics in the Indus Delta

Ying ZHOU, Zhijun DAI, Xixing LIANG, Jinping CHENG

Research output: Contribution to journalArticlespeer-review

Abstract

Mangrove forests play a vital role in carbon sequestration, typhoon-induced wave attenuation, and the provision of ecological services. However, mangrove ecosystems have experienced large-scale loss globally due to rising sea levels and anthropogenic activities. This study investigates the dynamic changes in mangrove cover within the mega-Indus delta, the largest delta in Pakistan and Southern Asia, using multi-temporal remote sensing data and machine learning techniques from 1988 to 2023. The results indicate an increasing trend in mangrove areas in the Indus Delta, with an average annual growth rate of 18.72 %. The spatial distribution of mangrove forests tends to concentrate towards the landward areas, extending along tidal channels, while losses primarily occur in the seaward regions. Rising sea levels pose a potential threat to the survival of these mangroves. The strong southwest monsoon-driven waves are the leading cause of shoreline erosion of the Indus Delta mangroves. Meanwhile, the reduction in riverine sediment discharge is not associated with the increase in mangrove area. Instead, the tidal currents influenced by the southwest monsoon carry sediments into the delta's tidal channels, causing them to fill and create suitable habitats for mangroves, which are the primary drivers of the observed mangrove expansion in the Indus Delta. Additionally, afforestation activities observed in the northwest and southwest parts of the study area have contributed to the restoration of mangroves. The loss of mangroves in the northernmost part of the northwest region was attributed to an oil spill incident. This study highlights the dynamic nature of mangrove ecosystems in the Indus Delta, characterized by an arid climate and low population density. The findings provide valuable insights into the factors influencing mangrove gain and loss and can inform management strategies for global mangrove restoration efforts. Copyright © 2024 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.

Original languageEnglish
Article number122231
JournalForest Ecology and Management
Volume571
Early online dateAug 2024
DOIs
Publication statusE-pub ahead of print - Aug 2024

Citation

Zhou, Y., Dai, Z., Liang, X., & Cheng, J. (2024). Machine learning-based monitoring of mangrove ecosystem dynamics in the Indus Delta. Forest Ecology and Management, 571, Article 122231. https://doi.org/10.1016/j.foreco.2024.122231

Keywords

  • Mangrove expansion
  • Hydro-sediment dynamic
  • Machine learning
  • Random forest
  • Tidal channel
  • Shoreline erosion

Fingerprint

Dive into the research topics of 'Machine learning-based monitoring of mangrove ecosystem dynamics in the Indus Delta'. Together they form a unique fingerprint.