基于DEMETER卫星观测数据的电离层离子温度时间序列预测模型
|
摘要
提出了一种利用DEMETER卫星观测量构建电离层离子温度背景场及预测模型的时间序列分析方法。首先,通过分析卫星轨道重访周期与建模格网大小的关系,确定了时间序列采样轨道数据的空间区域大小为1.75°×1.75°、时间间隔约为14天。然后,基于电离层离子温度明显的季节变化特征,采用自回归移动平均模型(ARIMA)构建时间序列预测模型来描述电离层离子温度及其周期性变化。最后,采用2006年～2009年的DE-METER卫星观测数据验证了该时间序列预测模型,结果表明ARIMA季节模型能较好地模拟电离层离子温度在时间上的变化趋势,建立较为可靠的电离层背景场。
The recent researches revealed there is some relationship of anomalous signals observed by DEMETER satellite sensors with seismic abnormalities.Since ionosphere is affected heavily by solar radiation and earth magnetic field,only after removing those background signals the true abnormalities related to seismic activities can be obtained.So,it is essential to establish a background field reflecting the changes of various features in ionosphere.This paper presents an advanced time-series analysis approach to modeling the background filed of ionosphere ion temperature(Ti) using DEMETER earth observation dataset.First,we analyzed the relations between the satellite repeat cycle and modeling grid size,and determined to prepare time-series Ti data with an 14-days interval and falling within a box of 1.75°× 1.75°centered at a particular geographic location.Then,the ARIMA model(Seasonal Autoregressive Integrated Moving Average) was adopted to model the background field of ionosphere Ti by concerning on the periodic changes of Ti values in time-series dataset.Finally,the observation data of 2006-2009 was used to calibrate and validate the ARIMA model,and the results show that the proposed time-series model could well predict the changes of Ti in future and give a high-quality background filed of ionosphere Ti.
The recent researches revealed there is some relationship of anomalous signals observed by DEMETER satellite sensors with seismic abnormalities.Since ionosphere is affected heavily by solar radiation and earth magnetic field,only after removing those background signals the true abnormalities related to seismic activities can be obtained.So,it is essential to establish a background field reflecting the changes of various features in ionosphere.This paper presents an advanced time-series analysis approach to modeling the background filed of ionosphere ion temperature(Ti) using DEMETER earth observation dataset.First,we analyzed the relations between the satellite repeat cycle and modeling grid size,and determined to prepare time-series Ti data with an 14-days interval and falling within a box of 1.75°× 1.75°centered at a particular geographic location.Then,the ARIMA model(Seasonal Autoregressive Integrated Moving Average) was adopted to model the background field of ionosphere Ti by concerning on the periodic changes of Ti values in time-series dataset.Finally,the observation data of 2006-2009 was used to calibrate and validate the ARIMA model,and the results show that the proposed time-series model could well predict the changes of Ti in future and give a high-quality background filed of ionosphere Ti.
引文
[1]Hayakawa M.Atmospheric and ionospheric electromagnetic phenomena associated with earthquakes[C]//Tokyo:TER-RAPUB,1999:997.
[2]Hayakawa M,Molchanov O A.Seismo-Electromagnetics:Lithosphere-atmosphere-ionospherer coupling[C]//Tokyo:TERRAPUB,2002:477.
[3]何宇飞.地震电磁前兆观测[J].国际地震动态,2009,32(6):28-36.
[4]朱荣,杨冬梅,荆风.DEMETER卫星观测到的云南普洱地震前的电离层扰动[J].地震学报,2008,30(1):76-81.
[5]安张辉,范莹莹,刘君.汶川M_S8.0地震前DEMETER卫星探测的离子温度变化分析[J].地震学报,2011(6):755-759.
[6]李志刚,程宗颐,冯初刚,等.电离层预报模型研究[J].地球物理学报,2007,50(2):327-337.
[7]Brockwell,P.J.,Davis,R.A..Introduction to Time Series and Forecasting[M].Springer,New York.,Sections 3.3and 8.3.
[8]Harvey,A.C.Time Series Models[M].2nd Edition,Harvester Wheatsheaf.
[9]Durbin,J.,Koopman,S.J.Time series analysis by state space methods[D].Oxford University Press.
[10]Jones,R.H.Maximum likelihood fitting of ARMA models to time series with missing observations[J].Technometrics1980(20):389-395.
[11]张树京,齐立新.时间序列分析简明教程[M].清华大学出版社,2003:137-138.
[12]潘国荣.基于时间序列分析的动态变形预测模型研究[J].武汉大学学报:信息科学版,2005,30(6):484-485.
[13]董承玮,宋现锋,芮小平,徐方舟,汪超亮.DEMETER卫星数据格式解析与数据入库效率研究[J].地理与地理信息科学,2010,26(4):42-46.
[2]Hayakawa M,Molchanov O A.Seismo-Electromagnetics:Lithosphere-atmosphere-ionospherer coupling[C]//Tokyo:TERRAPUB,2002:477.
[3]何宇飞.地震电磁前兆观测[J].国际地震动态,2009,32(6):28-36.
[4]朱荣,杨冬梅,荆风.DEMETER卫星观测到的云南普洱地震前的电离层扰动[J].地震学报,2008,30(1):76-81.
[5]安张辉,范莹莹,刘君.汶川M_S8.0地震前DEMETER卫星探测的离子温度变化分析[J].地震学报,2011(6):755-759.
[6]李志刚,程宗颐,冯初刚,等.电离层预报模型研究[J].地球物理学报,2007,50(2):327-337.
[7]Brockwell,P.J.,Davis,R.A..Introduction to Time Series and Forecasting[M].Springer,New York.,Sections 3.3and 8.3.
[8]Harvey,A.C.Time Series Models[M].2nd Edition,Harvester Wheatsheaf.
[9]Durbin,J.,Koopman,S.J.Time series analysis by state space methods[D].Oxford University Press.
[10]Jones,R.H.Maximum likelihood fitting of ARMA models to time series with missing observations[J].Technometrics1980(20):389-395.
[11]张树京,齐立新.时间序列分析简明教程[M].清华大学出版社,2003:137-138.
[12]潘国荣.基于时间序列分析的动态变形预测模型研究[J].武汉大学学报:信息科学版,2005,30(6):484-485.
[13]董承玮,宋现锋,芮小平,徐方舟,汪超亮.DEMETER卫星数据格式解析与数据入库效率研究[J].地理与地理信息科学,2010,26(4):42-46.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心