欢迎访问《中国农学通报》,

中国农学通报 ›› 2011, Vol. 27 ›› Issue (27): 161-167.

• 农学 农业基础科学 • 上一篇    下一篇

作物生长模型同化SAR数据模拟作物生物量时域变化特征

谭正 刘湘南 张晓倩 吴伶   

  • 收稿日期:2011-06-02 修回日期:2011-07-28 出版日期:2011-10-25 发布日期:2011-10-25

Simulation of Dynamics of Crop Biomass by Assimulation SAR Data into Crop Growth Model

  • Received:2011-06-02 Revised:2011-07-28 Online:2011-10-25 Published:2011-10-25

摘要:

生物量是监测作物长势的一个重要指标,可以反映作物的生长状况,和作物产量有密切关系。遥感获取生物量的方法之一是通过基于矢量辐射传输方程的微波冠层散射模型反演,但多数模型反演方法并未考虑作物生物量在不同阶段的变化特征。运用数据同化方法,将SAR数据提取的生物量信息和作物生长模型结合,描述作物生物量与时间变化的关系,提高生物量估测精度。通过分析生物量和SAR数据提取的后向散射系数的时域变化关系建立反演模型估算生物量。在构建代价函数的基础上,采用共轭梯度法对生长模型参数进行优化,使模型估算的生物量和SAR数据反演的生物量差值最小。结果表明,引入SAR数据修正后的作物生长模型模拟生物量和实测值的时间分布基本吻合,且比未引入SAR数据的结果精度有明显提高。因此采用SAR数据提取的作物实时生长信息可以修正作物生长模型关键参数以提高模拟生物量的精度。

关键词: 城市园林, 城市园林, 现代艺术, 园林设计

Abstract:

Biomass is one of the key indicators in crop condition monitoring, which can reflect the growth status of crop and can be closely related to crop yield. One of the methods for retrieving biomass from remotely sensed observations is through inversion of microwave canopy reflectance models based on vector radiative transfer equation. Many model inversion methods fail to account for variable biomass values at different crop growth stages. The objective of this preliminary study is to assimilate the biomass information retrieved from SAR data into the crop growth model to describe the biomass changes with crop growth and improve the accuracy of biomass estimation. Firstly the relationship of temporal biomass with backscatter coefficients retrieved from SAR data is analyzed, and then an empirical model is developed to estimate biomass. Then the Fletcher-Reeves method is developed to modify the parameters of the model by minimizing the difference between biomass simulated by the crop growth model and biomass inverted by SAR data. The results showed that simulated biomass values by the revised model introduced SAR data and the measured values were consistent, and the estimated biomass profile had been significantly improved. And the assimilation of real time and dynamic SAR data into crop growth model can calibrate crop model to actual crop growth conditions to improve accuracy of simulated biomass.