Scholarlink

<p>In the Mediterranean region, agriculture accounts for 50 to 70% of total water usage, with perennial crops being among the most water-intensive. Irrigation is particularly crucial for these perennial crops, and this dependence has been exacerbated by climate change and the resulting water stress. This study aims to classify farms at the watershed scale, based on their irrigation water consumption. The main assumption is that farms with similar structural characteristics tend to adopt similar irrigation practices for the same crops. Different classification methods, including machine learning and regression techniques, were applied to various spatial databases to ensure reliable results. The study was conducted in the Ouvèze-Ventoux watershed in south-eastern France, using both farm’s surveys and watershed-scale data. Irrigation practices were detailed through farm surveys, and annual water consumption data from water managers at watershed scale were used to validate the classifications for 13 farms. The results showed that farms with high irrigation rates were identified with 90% accuracy, and specific heavily irrigated orchard plots were identified with 68% precision. Water consumption maps were created at both the watershed and municipal scales, with estimates deviating only 14% from actual usage. This approach offers a way to identify water consumption patterns at the watershed scale without relying on complex crop models, SVAT models, or remote sensing. It provides a practical, scalable tool also for policy-makers to identify farms with high irrigation demand, especially in orchard’s 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

Characterising the Diversity of Farm-Level Irrigation Practices for Water Management in North-Mediterranean Perennial Cropping Systems

  • 期刊论文

  • 发表日期：

    2026年1月24日

查看全文
  • Pierre Rouault,
  • Dominique Courault,
  • Fabrice Flamain,
  • Marta Debolini

摘要

In the Mediterranean region, agriculture accounts for 50 to 70% of total water usage, with perennial crops being among the most water-intensive. Irrigation is particularly crucial for these perennial crops, and this dependence has been exacerbated by climate change and the resulting water stress. This study aims to classify farms at the watershed scale, based on their irrigation water consumption. The main assumption is that farms with similar structural characteristics tend to adopt similar irrigation practices for the same crops. Different classification methods, including machine learning and regression techniques, were applied to various spatial databases to ensure reliable results. The study was conducted in the Ouvèze-Ventoux watershed in south-eastern France, using both farm’s surveys and watershed-scale data. Irrigation practices were detailed through farm surveys, and annual water consumption data from water managers at watershed scale were used to validate the classifications for 13 farms. The results showed that farms with high irrigation rates were identified with 90% accuracy, and specific heavily irrigated orchard plots were identified with 68% precision. Water consumption maps were created at both the watershed and municipal scales, with estimates deviating only 14% from actual usage. This approach offers a way to identify water consumption patterns at the watershed scale without relying on complex crop models, SVAT models, or remote sensing. It provides a practical, scalable tool also for policy-makers to identify farms with high irrigation demand, especially in orchard’s systems.