基于电磁探测数据的时序分析模型研究

针对电磁探测数据交叉检验时对不同卫星探测数据的时间匹配需求，本文基于DEMETER卫星时序探测数据，分析了国际参考电离层(IRI)模型模拟电子浓度(N_e)数据在不同纬度区域的误差分布特征； 同时，基于自回归移动平均(ARIMA)模型构建了N_e数据时序预测模型. 在此基础上，分析比较IRI模型与ARIMA模型在N_e数据时序预测中的优缺点，结果表明: ARIMA模型模拟预测N_e数据时间序列的相对误差在短期内较低(小于10%)，且随着预测时间的增长而增大； 而IRI模型模拟预测N_e数据时间序列的相对误差不会随着预测时间的增长而增大，且在高纬度地区的预测相对误差比在中低纬度地区低. 

Time series analysis model based on electromagnetic detection data

Due to the demand for the time matching of different satellites data during cross-calibration of electromagnetic detection data, this paper analyzes the error distribution features of electron density (N_e) simulated based on the international reference ionosphere (IRI) model at different latitudes by using the time series data of DEMETER satellite. At the same time, based on the autoregre-ssive integrated moving average (ARIMA) model, the N_e data time series forecasting model is constructed. By comparing with the IRI model, the advantages and disadvantages of ARIMA model are analyzed in simulated prediction of the N_e time series data. The results show that the relative error of ARIMA forecasting model is small in the short-time (the relative error less than 10%), however, becomes greater in the long time. Whereas, the relative error of IRI model in simulated prediction of N_e time series data does not become larger in the long time, and the relative error at high latitudes is lower than that at low and moderate latitudes.