Accuracy Evaluation of ERP Prediction Based on Satellite Orbit Determination

ERP (Earth Rotation Parameters) are essential parameters for the transformation between the terrestrial reference frame and the celestial reference frame, playing a crucial role in fields such as timekeeping, satellite orbit determination, and deep space exploration. Numerous scholars have focused on achieving high-precision ERP prediction. Currently, the primary method for evaluating ERP prediction accuracy is to use the C04 series published by the IERS as a reference and measure performance by comparing Mean Absolute Error (MAE). This approach is overly simplistic and heavily dependent on the C04 series. In this study, ERP prediction data from Bulletin A, finals.daily, and three participants from the 2nd EOP PCC (Earth Orientation Parameters Prediction Comparison Campaign) are evaluated based on satellite orbit determination accuracy. The evaluation results indicate that: 1. The PMX (Polar Motion X) term in Bulletin A files exhibits significant deviations over time, which is the primary factor affecting orbital accuracy. 2. The predicted values in finals.daily files meet the requirements for satellite orbit determination. 3. For the first-day prediction accuracy, the ranking is ID 136 > ID 101 > ID 117, while for the seventh-day prediction accuracy, the ranking shifts to ID 101 > ID 136 > ID 117.