基础时间序列长度对ERP短期预报精度影响的分析

Analysis of Influence of Basic Time Series Size on ERP Short-Term Prediction Accuracy

高精度预报地球自转参数(Earth Rotation Parameters,ERP)对卫星导航等领域具有重要意义。不同时间长度的ERP序列作为基础序列对预报结果影响不同。分别对2～20年长度的以ERP序列作为基础时间序列的预报精度进行统计分析发现,当ERP基础时间序列为10年时,对3个参数的预报效果较优,但依据预报跨度的不同,也可适当调整基础时间序列长度。

High accuracy prediction of earth rotation parameters has great significance on satellite navigation and other fields. Different length of ERP has different effects on prediction results as base sequences. Statistical analysis of prediction accuracy of ERP sequence of 2~20 years length as base time series shows that the prediction effect of the three parameters is better when the ERP base time series length is 10. And the length of basic time series can also be adjusted appropriately according to the difference of forecast span.