基于电磁探测数据的时序分析模型研究  

Time series analysis model based on electromagnetic detection data

作　　者：张志玮[1,2,3] 李子扬[1,2] 朱小华[1,2] 李传荣[1,2] 

机构地区：[1]中国科学院定量遥感信息技术重点实验室,中国北京100094 [2]中国科学院光电研究院,中国北京100094 [3]中国科学院大学,中国北京100049

出　　处：《地震学报》2016年第3期408-416,508,共9页Acta Seismologica Sinica

基　　金：国防科工局民用航天科研工程项目(ZH-1-DMYZ-02-04)资助

摘　　要：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(Ne)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Ne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Ne数据时序预测中的优缺点,结果表明:ARIMA模型模拟预测Ne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Ne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.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 （Ne） 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 Ne data time series forecasting model is constructed. By comparing with the IRI model, the advantages and disadvantages of ARIMA model are analyzed in simulatedprediction of the Ne 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 Ne 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.

关 键 词：DEMETER卫星 电子浓度 国际参考电离层(IRI)模型 自回归移动平均(ARIMA)模型 时间序列分析 

分 类 号：P352.7[天文地球—空间物理学]