Scholarlink

<p>Surface water is a critical component of the global ecological environment and plays a pivotal role in societal development. Accurate detection and mapping of these water bodies are essential for resource management, flood monitoring, and disaster response. This study aims to investigate the impact of hydro-geomorphological factors on water body extraction using Google Earth Engine (GEE) and Landsat-8 imagery. We employed the OTSU threshold algorithm and the two most widely used spectral water indices, the Normalized Difference Water Index (NDWI) and the Modified Normalized Difference Water Index (MNDWI). Our objective is to identify the most suitable water index for 12 typical test sites, including lakes/reservoirs, mountainous regions, and urban areas across various climatic zones from humid to arid. Results indicate that both spectral water indices performed well at lakes/reservoirs test sites under varying humid and arid conditions, demonstrating approximately 97% accuracy. NDWI outperformed MNDWI in mountainous and arid/semi-arid conditions, with overall accuracy in mountainous areas exceeding 85%, which is more than 12% higher than MNDWI. NDWI also achieved an average accuracy of over 86% in arid sites. Conversely, MNDWI proved more effective in urban and humid/semi-humid settings, maintaining overall accuracy above 95% in both areas.</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

Effect of Hydro-geomorphological Environments on Surface Water Areas Extraction

  • 期刊论文

  • 发表日期：

    2025年4月15日

查看全文
  • Yuan-Yuan Jia,
  • Huan-Feng Duan,
  • Xu-Feng Yan

摘要

Surface water is a critical component of the global ecological environment and plays a pivotal role in societal development. Accurate detection and mapping of these water bodies are essential for resource management, flood monitoring, and disaster response. This study aims to investigate the impact of hydro-geomorphological factors on water body extraction using Google Earth Engine (GEE) and Landsat-8 imagery. We employed the OTSU threshold algorithm and the two most widely used spectral water indices, the Normalized Difference Water Index (NDWI) and the Modified Normalized Difference Water Index (MNDWI). Our objective is to identify the most suitable water index for 12 typical test sites, including lakes/reservoirs, mountainous regions, and urban areas across various climatic zones from humid to arid. Results indicate that both spectral water indices performed well at lakes/reservoirs test sites under varying humid and arid conditions, demonstrating approximately 97% accuracy. NDWI outperformed MNDWI in mountainous and arid/semi-arid conditions, with overall accuracy in mountainous areas exceeding 85%, which is more than 12% higher than MNDWI. NDWI also achieved an average accuracy of over 86% in arid sites. Conversely, MNDWI proved more effective in urban and humid/semi-humid settings, maintaining overall accuracy above 95% in both areas.