This study uses Sentinel-2 satellite imagery to monitor vineyard health in Peñafiel, Valladolid, Spain, from 2020 to 2023. Vegetation indices such as NDVI, GNDVI, and NDWI were used to assess key parameters like photosynthetic activity, water stress, and overall vineyard health. The results show a decline in vegetation indices, particularly in 2022, indicating possible water stress and a decrease in vineyard health. Strong positive correlations were found between July precipitation and vegetation indices, especially NDVI (0.99), and a negative correlation with August temperature (− 0.94), suggesting that higher temperatures reduce vegetation health. Among predictive models, Random Forest performed best with an MSE of 0.000675 and R2 of 0.574283, while MLP performed poorly with an MSE of 22.001097 and a negative R2. The next step will be to validate these models with additional data. The study highlights the need for improved irrigation practices and suggests incorporating additional variables such as soil conditions to enhance model predictions and help adapt to climate change, ensuring the long-term productivity of vineyards.

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

Crop Health Monitoring Using Vegetation Indices Derived from Satellite Remote Sensing in Wine Yard DO Ribera Del Duero

  • Claudia Helena Ramirez-Soler,
  • Susana Del Pozo Aguilera,
  • Juan M. Nuñez Velasco,
  • Fernando De la Prieta

摘要

This study uses Sentinel-2 satellite imagery to monitor vineyard health in Peñafiel, Valladolid, Spain, from 2020 to 2023. Vegetation indices such as NDVI, GNDVI, and NDWI were used to assess key parameters like photosynthetic activity, water stress, and overall vineyard health. The results show a decline in vegetation indices, particularly in 2022, indicating possible water stress and a decrease in vineyard health. Strong positive correlations were found between July precipitation and vegetation indices, especially NDVI (0.99), and a negative correlation with August temperature (− 0.94), suggesting that higher temperatures reduce vegetation health. Among predictive models, Random Forest performed best with an MSE of 0.000675 and R2 of 0.574283, while MLP performed poorly with an MSE of 22.001097 and a negative R2. The next step will be to validate these models with additional data. The study highlights the need for improved irrigation practices and suggests incorporating additional variables such as soil conditions to enhance model predictions and help adapt to climate change, ensuring the long-term productivity of vineyards.