Medium-long term forecasting method for earth rotation parameters considering effective angular momentum information

Objectives: The prediction accuracy of Earth rotation parameters (ERP), as one of the hot spots in geodetic research, has been attracting attention from users such as autonomous satellite navigation and deep space exploration. In order to further satisfy the users’ requirements, this paper combines the new effective angular momentum (EAM) data for the ERP forecast study. Methods: The EAM fitted residual series in the excitation domain is first transferred to the EOP domain, and then an empirical adjustment factor is added to construct a new EAM fitted residual series. Then the fitted residual series of ERP and the fitted residual series of the new EAM are subtracted to obtain the difference series. Using least-squares extrapolation and autoregressive (LS+AR) models driven by final product data and rapid data, forecast experiments are performed for the difference series and the new EAM residual series, respectively, and finally forecasts for the target ERP series are jointly introduced. Results: Based on hindcast experiments from the beginning of 2012 to the beginning of 2021, it is shown that an overall better result can be obtained for the adjustment factor of 0.7 in both the x-direction and y-direction at a forecast horizons of 1-365 days, with an average improvement of 31.86% and 21.00% over the traditional method, respectively; the average improvement over the traditional method was 15.01% when the adjustment factor of LOD was 1. Conclusions: This verifies the existing conclusion that about 70% of the high-frequency changes in the short time scale of polar motion are stimulated by EAM by means of numerical prediction, and provides a reference for ERP prediction research. In addition, considering time-varying periodic terms or replacing better models will also help to improve the prediction accuracy of the new method.