Scholarlink

<p>Soil moisture is vital for agriculture, and drought indices play a crucial role in water management and loss mitigation. Most studies use 10–40&#xa0;km resolution remote sensing data and the SMCI index, but these lack precision locally. This study develops a new Local-Scale Soil Moisture Condition Index (LS-SMCI) at 10-m resolution, outperforming coarser global products like FLDAS, GLDAS, and SMAP. Khuzestan, Iran, was chosen for developing the index due to its vital agricultural role and ongoing water scarcity. An artificial neural network (ANN) predicted soil moisture at 10-m resolution using Sentinel-1 VV and VH bands and their ratios as inputs, with soil moisture as the output. Trained on data from Khuzestan sugarcane fields, the model generated monthly predictions for the scaled LS-SMCI index. The index was validated over 2017–2023 by comparing it with the SMCI and SPI (3-, 6-, 9-month) using meteorological data, with RMSE and correlation as metrics. The LS-SMCI index has a stronger correlation with the SPI index than SMCI indices from global products, with an RMSE below 0.1, showing high accuracy. At 10-m resolution, it enables precise local drought monitoring, helping farmers and managers identify affected areas and apply timely strategies to boost resilience.</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

Developing a new high-resolution soil moisture index for local agricultural drought monitoring using Sentinel-1 data and an artificial neural network

  • 期刊论文

  • 发表日期：

    2025年12月22日

查看全文
  • Fateme Balouei,
  • Mostafa Kabolizadeh,
  • Hamidreza Rabiei-Dastjerdi

摘要

Soil moisture is vital for agriculture, and drought indices play a crucial role in water management and loss mitigation. Most studies use 10–40 km resolution remote sensing data and the SMCI index, but these lack precision locally. This study develops a new Local-Scale Soil Moisture Condition Index (LS-SMCI) at 10-m resolution, outperforming coarser global products like FLDAS, GLDAS, and SMAP. Khuzestan, Iran, was chosen for developing the index due to its vital agricultural role and ongoing water scarcity. An artificial neural network (ANN) predicted soil moisture at 10-m resolution using Sentinel-1 VV and VH bands and their ratios as inputs, with soil moisture as the output. Trained on data from Khuzestan sugarcane fields, the model generated monthly predictions for the scaled LS-SMCI index. The index was validated over 2017–2023 by comparing it with the SMCI and SPI (3-, 6-, 9-month) using meteorological data, with RMSE and correlation as metrics. The LS-SMCI index has a stronger correlation with the SPI index than SMCI indices from global products, with an RMSE below 0.1, showing high accuracy. At 10-m resolution, it enables precise local drought monitoring, helping farmers and managers identify affected areas and apply timely strategies to boost resilience.