Abstract Distributed crop models were considered a useful tool for allocating rational water resources. However, how to upscale the crop model into the regional scale is a key issue because the associated crop parameters tend to have spatial heterogeneity according to their feedback to the environment. The remote sensing may provide a possible method for generating the spatial distribution of crop parameters, thereby improving the upscaling approach. On this basis, this study aims to achieve the spatial distribution of crop parameters and establish the distributed AquaCrop-RS model by means of remote sensing. The results showed that the distribution of two crop parameters, namely, the maximum canopy coverage CCx* and the relative biomass Brel* was significantly heterogeneous with the spatial variability coefficient of 8% and 8.5% for CCx* in 2012 and 2013, respectively, and that of 6.3% and 4.6% for Brel*. Compared with the AquaCrop-GIS model without introducing the spatial parameters, the simulation accuracy of AquaCrop-RS model was improved by 20% in regional crop water consumption and by 26–36% in regional yield. The new upscaling approach could be an effective and promising method of crop model upscaling.

中文翻译：

---

通过使用从遥感数据得出的分布式作物参数将粮农组织 AquaCrop 模型升级到区域尺度的新方法

摘要 分布式作物模型被认为是合理分配水资源的有用工具。然而，如何将作物模型升级到区域尺度是一个关键问题，因为相关的作物参数根据它们对环境的反馈往往具有空间异质性。遥感可以为生成作物参数的空间分布提供一种可能的方法，从而改进升级方法。在此基础上，本研究旨在通过遥感手段实现作物参数的空间分布，建立分布式AquaCrop-RS模型。结果表明，最大冠层覆盖度CCx*和相对生物量Brel*两个作物参数分布具有显着异质性，空间变异系数分别为8%和8%。2012 年和 2013 年 CCx* 分别为 5%，Brel* 为 6.3% 和 4.6%。与不引入空间参数的AquaCrop-GIS模型相比，AquaCrop-RS模型的模拟精度在区域作物耗水量上提高了20%，在区域产量上提高了26-36%。新的升级方法可能是一种有效且有前途的作物模型升级方法。