美国作为全球最大的玉米、大豆生产国，在全球粮食供应中发挥着重要的作用。极端高温是威胁美国玉米和大豆单产的主要环境胁迫因素。虽然很多研究探讨了极端高温对上述作物单产的影响，但是对于衡量极端高温的一种重要农学指标，极端高温度日(KDD)，仍然缺乏充分认识。这是由于以往的研究通常根据固定高温阈值计算极端高温度日而忽略了该指标的空间异质性。本研究通过线性回归模型拟合了1981-2020年间极端高温度日与玉米、大豆单位产量之间的关系,确定了美国县级玉米和大豆极端高温度日的最佳温度下限。发现两种作物的格局均表现出稳健的集聚分布，玉米对极端高温的耐受能力略高于大豆。在一定温度范围内，随县域灌溉比例的提高，作物逐渐上升。灌溉能够减少高温下作物的水分胁迫、通过提高玉米和大豆，缓解极端高温对玉米和大豆的负面影响，使得产量波动更加稳定。随逐渐升高，单位KDD造成的产量损失非线性增长。不同背景环境下的均值差异不明显，这可能是受农民适应性措施的影响。本研究对极端高温度日的研究有助于进一步理解极端高温对美国玉米和大豆的影响机制，为更加准确地评估和预测极端高温提供理论基础。

As the world's largest producer of corn and soybean, the United States plays a significant role in global food supply. Extreme heat is a main environmental stress factor threatening the yield of maize and soybeans in USA. Although many studies have explored the effects of extreme heat on the yield of these crops, there is still a lack of thorough understanding of an important agricultural index which measuring extreme heat, Killing Degree Days(KDD). This is because previous studies usually calculated KDD based on fixed high temperature threshold and ignored the spatial heterogeneity of this index. In this study, a linear regression model was used to fit the relationship between Killing Degree Days(KDD) and yield of maize and soybeans during 1981-2020 in the United States and the optimum base temperature of KDD was determined. We found that of the two crops showed a robust clustering distribution, while maize had a slightly higher tolerance to extreme high temperature than soybeans. increases gradually with the increase of irrigation fraction in a certain temperature range. Irrigation is a critical way to reduce the water stress induced by heat, alleviate the negative effects of extreme high temperature of corn and soybean by raising and make the yield fluctuation more stable. As increases, the yield loss per KDD increases nonlinearly. The difference of mean in different background environments is not obvious, which may be affected by the adaptation measures of farmers. This study of KDD helps to further understand mechanism of extreme heat on maize and soybeans in the United States, and provides a theoretical basis for more accurate evaluation and prediction of extreme heat.