Scholarlink

<p>Extreme hydrometeorological events are increasingly threatening high-mountain environments such as the Himalayas. Sediment-rich debris floods in high-mountain environment pose challenges for satellite-based flood detection, as high turbidity and sediment mixing obscure conventional water spectral signature. This study evaluates the performance of multiple Sentinel-2 spectral indices (normalized difference vegetation index [NDVI], normalized differences water index [NDWI], modified NDWI [MNDWI], and water ratio index [WRI]) for mapping the June 2021 Melamchi River debris flood in central Nepal and assesses post-event vegetation dynamics using NDVI time-series analysis&#xa0;(2019–2024) within Google Earth Engine&#xa0;(GEE). Index-based threshold calibration using stratified random sampling captured open-water areas but tended to overestimate inundation under turbid conditions. In contrast, NDVI decline delineated zones of vegetation removal and sediment burial associated with high-energy debris flows, achieving an overall mapping accuracy of 83% with lower commission error relative to water indices. NDVI is therefore interpreted not as direct water detectors. but as a disturbance proxy in sediment-dominated flood environment. Seasonally consistent NDVI composites were used to assess multi-year vegetation trends. Results indicated that 73.85% of the mapped flood corridors exhibited negative NDVI slope between 2019 and 2024 with degradation spatially concentrated within flood-affected zones. While regional climate variability may also influence vegetation dynamics, the spatial coincidence of degradation with mapped sediment-dominated flood corridors suggests strong geomorphic disturbance effects. The study demonstrates a reproducible, cloud-based workflow for integrating flood extent mapping with multi-year vegetation monitoring in data-scarce mountain regions. These findings highlight the importance of combining disturbance sensitive vegetation indices with conventional water indices when assessing sediment-dominated flood events.</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

Cloud-based ecological monitoring of post-flood vegetation dynamics in the Himalayan Melamchi River with Sentinel-2 and Google Earth Engine

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月23日

查看全文
  • Narayan Thapa,
  • Pawan Thapa,
  • Bipin Dulal,
  • Saurav Raj Khanal,
  • Sabin Bhattarai

摘要

Extreme hydrometeorological events are increasingly threatening high-mountain environments such as the Himalayas. Sediment-rich debris floods in high-mountain environment pose challenges for satellite-based flood detection, as high turbidity and sediment mixing obscure conventional water spectral signature. This study evaluates the performance of multiple Sentinel-2 spectral indices (normalized difference vegetation index [NDVI], normalized differences water index [NDWI], modified NDWI [MNDWI], and water ratio index [WRI]) for mapping the June 2021 Melamchi River debris flood in central Nepal and assesses post-event vegetation dynamics using NDVI time-series analysis (2019–2024) within Google Earth Engine (GEE). Index-based threshold calibration using stratified random sampling captured open-water areas but tended to overestimate inundation under turbid conditions. In contrast, NDVI decline delineated zones of vegetation removal and sediment burial associated with high-energy debris flows, achieving an overall mapping accuracy of 83% with lower commission error relative to water indices. NDVI is therefore interpreted not as direct water detectors. but as a disturbance proxy in sediment-dominated flood environment. Seasonally consistent NDVI composites were used to assess multi-year vegetation trends. Results indicated that 73.85% of the mapped flood corridors exhibited negative NDVI slope between 2019 and 2024 with degradation spatially concentrated within flood-affected zones. While regional climate variability may also influence vegetation dynamics, the spatial coincidence of degradation with mapped sediment-dominated flood corridors suggests strong geomorphic disturbance effects. The study demonstrates a reproducible, cloud-based workflow for integrating flood extent mapping with multi-year vegetation monitoring in data-scarce mountain regions. These findings highlight the importance of combining disturbance sensitive vegetation indices with conventional water indices when assessing sediment-dominated flood events.