Scholarlink

<p>Understanding spatial variations in crop yield is vital for assessing yield gaps and improving resource management. Crop models such as AquaCrop simulate biophysical crop parameters like biomass and canopy cover, but as they are inherently point‑based, their direct application is limited to specific field sites. Therefore, spatializing these point‑based models is necessary to represent yield variations across larger regions. This study aimed to spatialize the AquaCrop model for soybean yield mapping in Nanded district, Maharashtra, using Sentinel‑2 imagery and ground observations. A soybean crop mask was generated through the Spectral Matching Technique, achieving an overall classification accuracy of 89.1%. The AquaCrop model was re‑parameterized using field‑level crop cutting experiment data to minimize deviations between simulated and observed yields. The re‑parameterized model showed strong agreement with ground observations (R² &gt; 0.9 across most taluks). Spatial yield maps derived by linking model parameters with remote sensing indicators overestimated yield by 15.35%. The results demonstrate that combining Sentinel‑2‑based spectral information with AquaCrop modelling enables robust, district‑scale yield estimation for soybean, while also recognizing the necessity for adequate ground data and considering local variations in farming practices.</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

Spatialization of Aquacrop Model Using Remote Sensing for Soyabean Yield Estimation for Nanded District of Maharashtra

  • 期刊论文

  • 发表日期：

    2026年5月27日

查看全文
  • A. Rajani,
  • H. R. Shwetha,
  • Murali Krishna Gumma,
  • Pranay Panjala,
  • Roja Mandapati

摘要

Understanding spatial variations in crop yield is vital for assessing yield gaps and improving resource management. Crop models such as AquaCrop simulate biophysical crop parameters like biomass and canopy cover, but as they are inherently point‑based, their direct application is limited to specific field sites. Therefore, spatializing these point‑based models is necessary to represent yield variations across larger regions. This study aimed to spatialize the AquaCrop model for soybean yield mapping in Nanded district, Maharashtra, using Sentinel‑2 imagery and ground observations. A soybean crop mask was generated through the Spectral Matching Technique, achieving an overall classification accuracy of 89.1%. The AquaCrop model was re‑parameterized using field‑level crop cutting experiment data to minimize deviations between simulated and observed yields. The re‑parameterized model showed strong agreement with ground observations (R² > 0.9 across most taluks). Spatial yield maps derived by linking model parameters with remote sensing indicators overestimated yield by 15.35%. The results demonstrate that combining Sentinel‑2‑based spectral information with AquaCrop modelling enables robust, district‑scale yield estimation for soybean, while also recognizing the necessity for adequate ground data and considering local variations in farming practices.