中文摘要：

从统计模型与作物机理模型的区别与联系出发，介绍了识别气候变化对农业产量贡献的3种主要统计模型，即时间序列模型、截而模型和面板模型；综述了前人在站点和区域（全球、国家、省级、地区、县级）尺度对这一问题的研究进展；总结了应用统计模型识别农业产量对气候变化响应敏感性的4个主要问题，包括时空尺度问题、产量的非气候趋势去除问题、气候要素间的自相关问题和忽略适应措施的问题；最后针对以上主要问题，提出了改进建议及今后研究的发展趋势。

英文摘要：

Statistical models and crop models are two major tools for research of contributions of climate change to agricultural production. The researchers have paid much attention to the studies of effects of climate change on crop growth based on crop models. However, the topic based on statistical models has not been fully realized. This paper starts with the difference and relation between statistical models and crop models, and introduces three main statistical methods for identifying contributions of climate change to crop yields including time-series model, cross-section model and panel model. It reviews the topic on different scales, e.g. global scale, national scale, provincial scale, regional scale, county scale and site scale. There are four problems in identifying response sensitivity of crop yields to climate change, namely spatial and temporal scale, non-climatic trend removal, co-linearity existing in climate variables and the non-consideration of adaptation.