PhD page of M.Sc. Maciej Michalczak

The (ultra) short term prediction of length-of-day using kriging

Authors:

• Maciej Michalczak,Department of Integrated Geodesy and Cartography, Faculty of Mining Surveying and Environmental Engineering, AGH University of Science and Technology, Krakow, Poland, e-mail: mmichalc@agh.edu.pl, ORCID: https://orcid.org/0000-0001-6859-988X
• Marcin Ligas, Department of Integrated Geodesy and Cartography, Faculty of Mining Surveying and Environmental Engineering, AGH University of Science and Technology, Krakow, Poland, e-mail: ligas@agh.edu.pl, ORCID: https://orcid.org/0000-0002-0970-128X

DOI: https://doi.org/10.1016/j.asr.2022.05.007

Abstract: The contribution presents results of ultra-short-term (15 days into the future) and short-term (30 days into the future) LOD prediction using kriging. IERS EOP 14 C04 (IAU2000A) and IERS EOP 05 C04 (IAU2000A) series are used as a reference. Two strategies of semivariogram selection for prediction purposes are employed. The first one uses a single semivariogram model in the entire prediction procedure whilst the second one selects a semivariogram model, from a group of models, with the least residual sum of squares. This work’s implementation of kriging offers mean absolute prediction errors in ultra short-term forecast for days 1–15 with respect to IERS EOP 14 C04 on the level of 0.035 – 0.149 ms whilst for IERS EOP 05 C04 the values are 0.030 – 0.139 ms. For short term prediction (days 1–30) the values are 0.038 – 0.180 ms and 0.034 – 0.175 ms; respectively. The results show that LOD kriging-based prediction may compete with existing methods.

Keywords: length-of-day, LOD, prediction, kriging, semivariogram, time series

List of files and results used during my research on the prediction of length of a day:

• Data used to processing:
  • IERS EOP 14 C04 (https://www.iers.org/IERS/EN/DataProducts/EarthOrientationData/eop.html)
  • IERS EOP 05 C04 (https://www.iers.org/IERS/EN/DataProducts/EarthOrientationData/Discontinued/eop.html)
• Step 1 - set of steering parameters:
  • ordinary and simple kriging IERS EOP 14 C04 15-days
  • ordinary kriging IERS EOP 05 C04 15 days
  • ordinary kriging IERS EOP 14 C04 30-days
  • ordinary kriging IERS EOP 05 C04 30-days
  • ordinary kriging IERS EOP 14 (long as 05) C04 30-days
• Step 2 - set of predictions:
  • ordinary and simple kriging and for adaptive and single-semivariogram approach (IERS_EOP_14_C04_15_days_25_12,2)
  • ordinary and simple kriging and for adaptive and single-semivariogram approach (IERS_EOP_14_C04_15_days_25_7)
  • ordinary kriging and for adaptive and single-semivariogram approach (IERS_EOP_05_C04_30_days_21_9,2)
  • ordinary kriging and for adaptive and single-semivariogram approach (IERS_EOP_14(long_as_05)_C04_30_days_21_9,2)
  • ordinary kriging and for adaptive and single-semivariogram approach (IERS_EOP_05_C04_15_days_25_9,2)
  • ordinary kriging and for adaptive and single-semivariogram approach (IERS_EOP_14(long_as_05)_C04_15_days_25_9,2)
  • ordinary kriging and for adaptive and single-semivariogram approach (IERS_EOP_05_C04_15_days_19_9,3)
  • ordinary kriging and for adaptive and single-semivariogram approach (IERS_EOP_14(long_as_05)_C04_15_days_19_9,3)



• Operational data comparison:
  • Real operational time comparison 10- and 30-days prediction (IERS 08 C04)