Scholarlink

<p>Dark particulate matter, biomass-burning residues, and emissions from vehicular and industrial sources, such as ash, get deposited on Himalayan snow, reducing its albedo and accelerating melt, which disrupts seasonal water availability for irrigation, hydropower production, and drinking water for downstream communities. This study introduces an automated hyperspectral workflow for detecting and quantifying ash contamination on snow surfaces. Field measurements of pure and ash-contaminated snow (2.5–10&#xa0;g) weights were collected using an SVC 1024i spectroradiometer across 350–2500&#xa0;nm. A seven-step absorption-feature-detection algorithm identified the NIR (1034&#xa0;nm), SWIR1 (1505&#xa0;nm), and SWIR2 (1982&#xa0;nm) water-ice absorption centres, achieving 100% validation for pure snow and mean spectral shifts of − 9.0 ± 3.1&#xa0;nm (NIR, <i>p</i> &lt; 0.01) and − 13.5 ± 6.2&#xa0;nm (SWIR1, <i>p</i> &lt; 0.05) with increasing contamination. SWIR2 responses were inconsistent (mean + 6.2 ± 25.4&#xa0;nm, <i>p</i> &gt; 0.05), indicating complex nonlinear effects. A data-driven band-selection routine computed wavelength-wise standard deviations, applied a 90th percentile threshold, and grouped contiguous sensitive wavelengths into three operational bands (blue-green: 350–522&#xa0;nm; red-1: 657–659&#xa0;nm; red-2: 664–693&#xa0;nm), demonstrating superior discrimination (separability scores ≥ 3.8). The NIR band proved to be most reliable for early contamination detection. We outline future work to refine SWIR2 analysis, incorporate machine learning for adaptive banding, and apply the algorithm to PRISMA and EMIT satellite imagery over critical Western Himalayan sites (Sissu, Manali, and Koti passes). This automated, reproducible workflow is compatible with current hyperspectral missions. It provides a foundation for operational mapping of ash-induced albedo reduction, informing water resource management and hazard mitigation under climate change.</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

Automated Spectral Feature Detection for Mapping Ash Contamination on Western Himalayan Snow

  • 期刊论文

  • 发表日期：

    2026年4月7日

查看全文
  • Balla Vivek,
  • Pradeep Kumar Garg

摘要

Dark particulate matter, biomass-burning residues, and emissions from vehicular and industrial sources, such as ash, get deposited on Himalayan snow, reducing its albedo and accelerating melt, which disrupts seasonal water availability for irrigation, hydropower production, and drinking water for downstream communities. This study introduces an automated hyperspectral workflow for detecting and quantifying ash contamination on snow surfaces. Field measurements of pure and ash-contaminated snow (2.5–10 g) weights were collected using an SVC 1024i spectroradiometer across 350–2500 nm. A seven-step absorption-feature-detection algorithm identified the NIR (1034 nm), SWIR1 (1505 nm), and SWIR2 (1982 nm) water-ice absorption centres, achieving 100% validation for pure snow and mean spectral shifts of − 9.0 ± 3.1 nm (NIR, p < 0.01) and − 13.5 ± 6.2 nm (SWIR1, p < 0.05) with increasing contamination. SWIR2 responses were inconsistent (mean + 6.2 ± 25.4 nm, p > 0.05), indicating complex nonlinear effects. A data-driven band-selection routine computed wavelength-wise standard deviations, applied a 90th percentile threshold, and grouped contiguous sensitive wavelengths into three operational bands (blue-green: 350–522 nm; red-1: 657–659 nm; red-2: 664–693 nm), demonstrating superior discrimination (separability scores ≥ 3.8). The NIR band proved to be most reliable for early contamination detection. We outline future work to refine SWIR2 analysis, incorporate machine learning for adaptive banding, and apply the algorithm to PRISMA and EMIT satellite imagery over critical Western Himalayan sites (Sissu, Manali, and Koti passes). This automated, reproducible workflow is compatible with current hyperspectral missions. It provides a foundation for operational mapping of ash-induced albedo reduction, informing water resource management and hazard mitigation under climate change.