郭斐, 陈惟杰, 朱逸凡, 张小红. 一种融合相位、振幅与频率的GNSS-IR土壤湿度反演方法[J]. 武汉大学学报 ( 信息科学版). DOI: 10.13203/j.whugis20210644
引用本文: 郭斐, 陈惟杰, 朱逸凡, 张小红. 一种融合相位、振幅与频率的GNSS-IR土壤湿度反演方法[J]. 武汉大学学报 ( 信息科学版). DOI: 10.13203/j.whugis20210644
Guo Fei, Chen Weijie, Zhu Yifan, Zhang Xiaohong. A GNSS-IR soil moisture inversion method integrating phase, amplitude and frequency[J]. Geomatics and Information Science of Wuhan University. DOI: 10.13203/j.whugis20210644
Citation: Guo Fei, Chen Weijie, Zhu Yifan, Zhang Xiaohong. A GNSS-IR soil moisture inversion method integrating phase, amplitude and frequency[J]. Geomatics and Information Science of Wuhan University. DOI: 10.13203/j.whugis20210644

一种融合相位、振幅与频率的GNSS-IR土壤湿度反演方法

A GNSS-IR soil moisture inversion method integrating phase, amplitude and frequency

  • 摘要: GNSS干涉测量(GNSS Interferometric Reflectometry,GNSS-IR)技术已经成为探测地表环境特性的一种新兴被动遥感技术,本文综合利用从土壤反射的GNSS信号中提取的相位、振幅、频率特征,提出了一种多类型特征数据融合的GNSS-IR土壤湿度反演方法,采用最小二乘支持向量机(LSSVM,Least Square Support Vector Machine)、随机森林(RF,Random Forest)、BP神经网络(BPNN,Back Propagation Neural Network)三种机器学习模型,对比和验证了所提出方法的可行性与效果。结果表明多特征融合的LSSVM、RF和BPNN模型反演得到的土壤湿度与参考值的相关性系数分别为0.823、0.944和0.955,对应的均方根误差(RMSE)分别为0.045、0.035和0.032 cm3 · cm-3。相比于单一特征反演法,土壤湿度反演精度提高了6-14%,相关系数提高了2-7%。

     

    Abstract: Objectives:GNSS-IR (Global Interferometric Reflectometry) is a new passive remote sensing technique for determining surface environment parameters, which places an important part in the inversion of earth's surface properties, such as soil moisture monitoring, snow parameter retrieval, and vegetation remote sensing, etc. The GNSS-IR offers several benefits over the traditional soil moisture inversion approach, including all-weather capability, high temporal precision, and cheap cost. Method:Considering the fact that the existing soil moisture inversion algorithms only utilize one single feature of GNSS reflected signal and from the perspective of increasing data availability, this paper proposes a GNSS-IR soil moisture inversion approach that integrates multi-type feature data by utilizing phase, amplitude, and frequency extracted GNSS signals reflected by soil. The main work is to effectively filter all available features extracted from the original GNSS SNR observations. The feasibility and effect of the suggested method are compared and evaluated using three machine learning models, including LSSVM (Least Squares Support Vector Machine), RF (Random Forest), and BPNN (Back Propagation Neural Network). Results:Comparing the inversion effects of the above three models, the BP neural network model has the best inversion effect, followed by the random forest (RF) model, and the least squares support vector machine (LSSVM) model is the worst. The results showed that the correlation coefficients between the reference value and soil moisture inversed by the multi-feature fusion method LSSVM, RF, And BPNN models are 0.823, 0.944, and 0.955, respectively, and the corresponding root mean square errors (RMSE) are 0.045, 0.035 and 0.032cm3·cm-3. Conclusions:Compared with the single feature inversion method, the soil moisture inversion accuracy is increased by 6-14%, and the correlation coefficient is increased by 2-7%. The results demonstrated that the proposed method has higher inversion accuracy and reliability than the single feature inversion method.

     

/

返回文章
返回