Impact of the reference series choice in analysis of the Second Earth Orientation Parameters Prediction Comparison Campaign (2nd EOP PCC) results

Earth Orientation Parameters (EOP) represent the rotational part of the transformation between the current releases of the International Celestial Reference Frame (ICRF) and the International Terrestrial Reference Frame (ITRF). The determination of EOPs requires post-processing of observational data collected from various space geodetic techniques, which causes some delays in the provision of EOP solutions.

In the light of the developments in the field of advanced geodetic data processing, modelling effective angular momentum functions and developing new prediction methods, a re-assessment of the various EOP predictions is currently pursued in the frame of the Second Earth Orientation Parameters Prediction Comparison Campaign (2nd EOP PCC). The campaign started in September 2021 and its main part finished on 28 December 2022. It was run by the Centrum Badań Kosmicznych Polskiej Akademii Nauk in cooperation with GeoForschungsZentrum and under the auspices of the International Earth Rotation and Reference Systems Service (IERS).

In this presentation, we provide initial analysis of the impact of the choice of reference series used on the assessment of predictions submitted to the 2nd EOP PCC. We will compare the single-technique EOP series provided by International Global Navigation Satellite Systems Service (IGS), International Laser Ranging Service (ILRS), International Very Long Baseline Interferometry Service (IVS) as well as selected combined series with reference to official IERS solutions (IERS 14C04) for respective EOPs. We also look closer to the input data exploited by campaign participants to process predictions by dividing them into subclasses. We conduct selected statistical analyses, including forecast horizons, to show variabilities caused by reference series. We also analyse the number of rejected predictions with use of beta parameter considering various references. Finally, we focus on Mean Absolute Error (MAE) computed for 10 days and 30 days of predictions to discuss potential effects triggered by the input data choice. This work concludes with potential impacts of the input EOP data on the accuracy of the received EOP predictions.