刘立龙, 陈雨田, 黎峻宇, 田祥雨, 贺朝双. 活跃期区域电离层总电子短期预报及适用性分析[J]. 武汉大学学报 ( 信息科学版), 2019, 44(12): 1757-1764. DOI: 10.13203/j.whugis20180145
引用本文: 刘立龙, 陈雨田, 黎峻宇, 田祥雨, 贺朝双. 活跃期区域电离层总电子短期预报及适用性分析[J]. 武汉大学学报 ( 信息科学版), 2019, 44(12): 1757-1764. DOI: 10.13203/j.whugis20180145
LIU Lilong, CHEN Yutian, LI Junyu, TIAN Xiangyu, HE Chaoshuang. Short-term Prediction and Applicability Analysis of Regional Ionospheric Total Electron Content in Active Period[J]. Geomatics and Information Science of Wuhan University, 2019, 44(12): 1757-1764. DOI: 10.13203/j.whugis20180145
Citation: LIU Lilong, CHEN Yutian, LI Junyu, TIAN Xiangyu, HE Chaoshuang. Short-term Prediction and Applicability Analysis of Regional Ionospheric Total Electron Content in Active Period[J]. Geomatics and Information Science of Wuhan University, 2019, 44(12): 1757-1764. DOI: 10.13203/j.whugis20180145

活跃期区域电离层总电子短期预报及适用性分析

Short-term Prediction and Applicability Analysis of Regional Ionospheric Total Electron Content in Active Period

  • 摘要: 太阳活跃期受太阳风高能粒子影响易发生磁暴,使得电离层总电子含量异常扰动,其非平稳性与非线性特征较平静期明显增强。分别利用2011年区域内多个测站的实测数据与IGS(International GNSS Service)发布的全球电离层模型(global ionosphere model,GIM)进行逐点建模,选取db4小波基对样本序列进行分解后,采用时间序列模型对各分量进行预报并重构,实现对ARIMA(auto regressive integrated moving average)模型的改进。通过分析ARIMA模型与改进模型预报值的残差比例和实验区域内均方根误差的分布情况,来评定改进模型的预报精度与适用性。结果表明,改进模型的残差与实验区域内的均方根误差较ARIMA模型总体减小,且该模型对区域内均方根误差峰值能起到较大的削弱作用。

     

    Abstract: In the solar active period, the earth's magnetic field is easily affected by the high energy particles of the solar wind, which makes the total electron content of the ionosphere abnormally disturbed, and its non-stationary and nonlinear characteristics are obviously enhanced compared to the calm period. Using the measured data from multiple stations in the 2011 region and the GIM (global ionosphere model) published by the IGS (International GNSS Service) to perform point-by-point modeling, the db4 wavelet basis is used to decompose the sample sequence, and the time series model is used to forecast each component and forecast. Each component is reconstructed so that the ARIMA (auto regressive integrated moving average) model can be improved. The prediction accuracy and applicability of the improved model are evaluated by analyzing the residual ratio of the ARIMA model and the improved model and the distribution of the root mean square error in the experimental region. The results show that the residual error of the improved model and the root mean square error in the experimental area are reduced compared with the ARIMA model, and the improved model can greatly weaken the peak value of the root mean square error in the area.

     

/

返回文章
返回