Scholarlink

<p>The significance of mangroves and the challenges to their sustainability have long been understood, promoting local, governmental, and international efforts to conserve and restore them. Despite these actions, our understanding of the post-conservation and restoration dynamics within mangroves remains limited. This study develops a LSTM model to predict the mangrove change pattern and forecast the mangrove area for the period 2023–2027 in the Bay of Assassins in southwest Madagascar and Abu Dhabi, United Arab Emirates. To develop the model we employed annual Landsat satellite images to assess and monitor long-term changes in mangrove area for blue forests projects during the period 2000–2022 in these areas. The results reveal a fluctuating trend in the mangrove area of both the Bay of Assassins and Abu Dhabi, marked by a significant decline before conservation initiatives and a gradual improvement in the later years of the study period. The forecasting results indicate that, through sustainable management and conservation programs, the mangrove area in Bay of Assassins and Abu Dhabi could potentially increase by 1.4% and 5.9%, respectively, between 2023 and 2027. Our study is among the first to apply an LSTM model for forecasting post-conservation mangrove dynamics using Landsat time series data. By focusing on two contrasting regions- the Bay of Assassins in southeast Madagascar and Abu Dhabi- it provides new insights into the effectiveness of restoration efforts and their projected outcomes through 2027. </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

Using a long short-term memory neural network model to forecast mangrove change in two blue forests conservation projects

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月10日

查看全文
  • Raheleh Farzanmanesh,
  • Kourosh Khoshelham,
  • Liubov Volkova,
  • Sebastian Thomas,
  • Mohsen Bakhtiari,
  • Jaona Ravelonjatovo,
  • Christopher J. Weston

摘要

The significance of mangroves and the challenges to their sustainability have long been understood, promoting local, governmental, and international efforts to conserve and restore them. Despite these actions, our understanding of the post-conservation and restoration dynamics within mangroves remains limited. This study develops a LSTM model to predict the mangrove change pattern and forecast the mangrove area for the period 2023–2027 in the Bay of Assassins in southwest Madagascar and Abu Dhabi, United Arab Emirates. To develop the model we employed annual Landsat satellite images to assess and monitor long-term changes in mangrove area for blue forests projects during the period 2000–2022 in these areas. The results reveal a fluctuating trend in the mangrove area of both the Bay of Assassins and Abu Dhabi, marked by a significant decline before conservation initiatives and a gradual improvement in the later years of the study period. The forecasting results indicate that, through sustainable management and conservation programs, the mangrove area in Bay of Assassins and Abu Dhabi could potentially increase by 1.4% and 5.9%, respectively, between 2023 and 2027. Our study is among the first to apply an LSTM model for forecasting post-conservation mangrove dynamics using Landsat time series data. By focusing on two contrasting regions- the Bay of Assassins in southeast Madagascar and Abu Dhabi- it provides new insights into the effectiveness of restoration efforts and their projected outcomes through 2027.