陆态网络CORS站观测时间序列分析方法研究

Research on the Methods of CMONOC CORS Station Time Series Analysis

以陆态网络CORS站中的武汉站为切入点,分别采用ARIMA模型和傅里叶拟合模型对武汉站10 a左右的观测时间序列进行分析拟合。实验表明,ARIMA模型和傅里叶拟合模型都能较好地拟合武汉站测站时间序列,但从细节来讲,ARIMA模型理论严密,拟合效果稳定;而傅里叶拟合模型更有利于直观分析时间序列中的各种周期频谱特性。

Taking the Wuhan CORS station for object, we used the ARIMA model and FFT model to fit the Wuhan station's about ten years' time series. According to the results, we can see that both ARIMA model and FFT model are good at fitting the Wuhan station time series. However, ARIMA model is very strict in theory and fits stably. FFT model is better in studying on the spectrum periods of the time series.