Scholarlink

<p>The total chlorophyll content is a critical indicator of plant health and photosynthetic capacity in agricultural monitoring systems. This study developed and validated empirical models for estimating total chlorophyll content in sugarcane plants using hyperspectral reflectance data harmonized with Sentinel-2 satellite imagery. Field measurements were conducted in West Java, Indonesia, where 19 sugarcane leaf samples were collected and analyzed for chlorophyll content using spectrophotometric methods. Corresponding hyperspectral reflectance data (350–2500&#xa0;nm) were acquired using an ASD FieldSpec 4 Hi-Res spectroradiometer and harmonized to Sentinel-2 band configurations. Multiple linear regression models were developed using different combinations of Sentinel-2 spectral bands (B1–B12) to predict total chlorophyll content. Five empirical models were evaluated based on the coefficient of determination (R²), adjusted R² (R²adj), and RMSE when applied to Sentinel-2 imagery. The results showed that the red-edge bands (B6 and B7) provided the most reliable chlorophyll estimates with 61.2% R² and 0.79&#xa0;mg/g RMSE. While models incorporating additional bands achieved higher R² values during development (up to 73.8%), they exhibited poor performance when implemented on satellite imagery due to atmospheric effects and spatial resolution limitations (RMSE up to 5.89&#xa0;mg/g). Time-series analysis from 2020 to 2024 revealed that the two-band model maintained consistent performance and successfully captured sugarcane plantation seasonal growth patterns. These findings provide a practical and computationally simple approach for large-scale chlorophyll monitoring in commercial sugarcane cultivation systems.</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

Empirical models for estimating total chlorophyll content in sugarcane using hyperspectral reflectance and Sentinel-2 imagery

  • 期刊论文

  • 发表日期：

    2025年10月3日

查看全文
  • Gatot Nugroho,
  • Tri Muji Susantoro,
  • M. Rokhis Khomarudin,
  • Danang Surya Candra,
  • Ahmad Maryanto,
  • Suliantara Suliantara,
  • Herru Lastiadi Setiawan,
  • Sayidah Sulma,
  • Rahmat Arief,
  • Agung Budi Harto,
  • Kurdianto Kurdianto,
  • Yohanes Fridolin Hestrio,
  • Adis Jayati

摘要

The total chlorophyll content is a critical indicator of plant health and photosynthetic capacity in agricultural monitoring systems. This study developed and validated empirical models for estimating total chlorophyll content in sugarcane plants using hyperspectral reflectance data harmonized with Sentinel-2 satellite imagery. Field measurements were conducted in West Java, Indonesia, where 19 sugarcane leaf samples were collected and analyzed for chlorophyll content using spectrophotometric methods. Corresponding hyperspectral reflectance data (350–2500 nm) were acquired using an ASD FieldSpec 4 Hi-Res spectroradiometer and harmonized to Sentinel-2 band configurations. Multiple linear regression models were developed using different combinations of Sentinel-2 spectral bands (B1–B12) to predict total chlorophyll content. Five empirical models were evaluated based on the coefficient of determination (R²), adjusted R² (R²adj), and RMSE when applied to Sentinel-2 imagery. The results showed that the red-edge bands (B6 and B7) provided the most reliable chlorophyll estimates with 61.2% R² and 0.79 mg/g RMSE. While models incorporating additional bands achieved higher R² values during development (up to 73.8%), they exhibited poor performance when implemented on satellite imagery due to atmospheric effects and spatial resolution limitations (RMSE up to 5.89 mg/g). Time-series analysis from 2020 to 2024 revealed that the two-band model maintained consistent performance and successfully captured sugarcane plantation seasonal growth patterns. These findings provide a practical and computationally simple approach for large-scale chlorophyll monitoring in commercial sugarcane cultivation systems.