A Least Squares Extrapolation Model for UT1-UTC Prediction Method with Consideration of the Edge-effect

The prediction accuracy of UT1-UTC can be easily affected by the edge distortion of least squares (LS) fitting time-series, referred to as edge-effect in the data-processing domain, when periodic oscillations and residuals are separated by LS fitting. In order to alleviate the edge-effect, the original UT1-UTC time-series is first extended on both boundaries by using a time-series analysis model in this paper. A LS extrapolation model is then set up using the extended time-series. Finally UT1-UTC predictions are obtained by employing the combination of the edge-effect correlated least squares (ECLS) model and a stochastic predication technology such as neural network (NN). The numerical experiments demonstrate that the edge-effect can be noticeably alleviated with the developed method. In addition, the accuracy of the UT1-UTC short-term predictions is comparable with that by the conventional LS extrapolation-based prediction algorithm. However, the medium- and long-term predictions are significantly more accurate than those obtained by the proposed ECLS extrapolation-based prediction solution.