Abstract The performance of irrigation schemes around the world has been below expectations. The assessment of irrigation performance is an essential step towards improving agricultural water use. One of the primary performance indicators is the relative irrigation supply (RIS, the ratio between the amount of water delivered and the crop net irrigation requirements). This study presents a worldwide analysis of irrigation scheme performance by evaluating key attributes that influence the RIS. The analysis was based on a review of reports and scientific papers that yielded 264 cases belonging to 25 countries in six world regions. The database was subjected to two types of statistical analysis: k-means clustering and analysis of covariance (ANCOVA). The cluster grouped irrigation schemes which were characterized by low RIS and advanced irrigation technology. The ANCOVA showed that the RIS co-varied significantly with the variation in precipitation, delivery schedule, on-farm irrigation systems, distribution network, and region, but not with the crop. The ANCOVA also showed that modern pressurized on-farm irrigation systems and on-demand distribution systems significantly improve RIS. The ANCOVA general linear model had a good capacity to predict RIS with a coefficient of determination of R2 = 0.83.

中文翻译：

---

与需求相关的灌溉计划供水的全球分析

摘要 世界各地灌溉计划的表现一直低于预期。灌溉绩效评估是改善农业用水的重要步骤。主要绩效指标之一是相对灌溉供应（RIS，输送的水量与作物净灌溉需求之间的比率）。本研究通过评估影响 RIS 的关键属性，对灌溉计划的绩效进行了全球分析。该分析基于对报告和科学论文的审查，这些报告和科学论文产生了属于世界六个区域的 25 个国家的 264 个案例。该数据库进行了两种类型的统计分析：k 均值聚类和协方差分析 (ANCOVA)。该集群将灌溉方案分组为具有低RIS和先进灌溉技术的特点。ANCOVA 表明，RIS 与降水、交付时间表、农场灌溉系统、分配网络和地区的变化显着共同变化，但与作物无关。ANCOVA 还表明，现代加压农场灌溉系统和按需分配系统显着提高了 RIS。ANCOVA 一般线性模型具有很好的预测 RIS 的能力，其决定系数为 R2 = 0.83。ANCOVA 还表明，现代加压农场灌溉系统和按需分配系统显着提高了 RIS。ANCOVA 一般线性模型具有很好的预测 RIS 的能力，其决定系数为 R2 = 0.83。ANCOVA 还表明，现代加压农场灌溉系统和按需分配系统显着提高了 RIS。ANCOVA 一般线性模型具有很好的预测 RIS 的能力，其决定系数为 R2 = 0.83。