Grapevine yield is defined as the quantity of harvest, expressed as either grape mass or wine volumeunits, which has been collected per surface unit are and per crop cycle. The information about current and future yield, termed a yield assessment in this paper, is an essential decision-making element for the grape and wine industry. Crop management, wine-making, commercial, accounting and strategic operations are all adapted to and all impact on the expected yield of the current vintage. Numerous yield assessment approaches have been proposed in the scientific literature. However, only a few of them have considered their adaptation to the operational context under the constraints, needs and strategies of commercial vineyards and wineries. The few studies that have worked on the operational implementation of yield assessment methods have only partially addressed this issue, concentrating their improvement efforts on a single step in the yield assessment process. This paper first proposes to review the characteristics of yield development in an operational context that must be taken into account by yield assessment methodsThese characteristics are consolidated into three main challenges for yield assessment methods: (i) addressing the complex temporality of yield development, (ii) ensuring a local monitoring of yield development and (iii) fitting to the operational needs and constraints to allow for relevant decision support systems in the field. The approaches of yield estimation, prediction and forecast are discussed in the context of these challenges. In a second step, the paper proposes a generic framework for the yield assessment process, including a review of the variables that are used to explain grapevine yield. Issues and proposals from the literature associated respectively with measurement, sampling and yield modelling are reviewed and the need for improved modelling of relationships between explanatory variables and the desired, reported yield variable is discussed. The yield assessment methods found in the literature are categorised and compared according to measurement, estimation and modelling approaches, and then according to the three challenges identified for yield assessment in operational conditions, such that the yield assessment method is adapted to commercial needs and not to research objectives. In conclusion, concrete proposals for new grape yield assessment methods are discussed in order to investigate the as yet unexplored opportunities for the improvement in yield assessment in operational contexts that have been identified in the paper. These considerations could easily be transposed to other perennial crops.

葡萄产量被定义为收获的数量，表示为葡萄质量或葡萄酒体积单位，已收集到每个表面单位和每个作物周期。有关当前和未来产量的信息，在本文中称为产量评估，是葡萄和葡萄酒行业的重要决策要素。作物管理、酿酒、商业、会计和战略运营都适应并影响当前年份的预期产量。科学文献中提出了许多产量评估方法。然而，在商业葡萄园和酿酒厂的限制、需求和策略下，只有少数人考虑了他们对运营环境的适应。为数不多的关于产量评估方法的操作实施的研究仅部分解决了这个问题，将他们的改进工作集中在产量评估过程中的一个步骤上。本文首先提出在产量评估方法必须考虑的操作背景下审查产量发展的特征这些特征被合并为产量评估方法的三个主要挑战：（i）解决产量发展的复杂时间性，（ii） ) 确保对产量发展进行本地监测，以及 (iii) 适应运营需求和限制，以便在现场使用相关的决策支持系统。在这些挑战的背景下讨论了产量估计、预测和预测的方法。第二步，该文件为产量评估过程提出了一个通用框架，包括对用于解释葡萄产量的变量的审查。分别与测量、抽样和产量建模相关的文献中的问题和建议进行了审查，并讨论了改进解释变量与所需报告的产量变量之间关系建模的必要性。根据测量、估计和建模方法对文献中发现的良率评估方法进行分类和比较，然后根据在操作条件下进行良率评估确定的三个挑战，使良率评估方法适应商业需求而不是研究目标。综上所述，讨论了新的葡萄产量评估方法的具体建议，以调查在本文中确定的操作环境中改进产量评估的尚未探索的机会。这些考虑很容易转移到其他多年生作物上。