Earth rotation parameter estimation from LLR

Abstract

Lunar Laser Ranging (LLR) measures the distance between observatories on Earth and retro-reflectors on Moon since 1969. In this paper, we estimate the Earth Rotation Parameters (ERP; terrestrial pole offsets, x_P and y_P , and Earth rotation phase, DeltaUT1) using LLR data. We estimate the values of DeltaUT1, and the pole offsets separately for nights selected from subsets of the LLR time series which have a minimum of 5, 10, and 15 normal points obtained per night. For the pole offsets, we estimate the values of x_P and y_P simultaneously as well as separately. Overall, the uncertainties of ERP from the new LLR data (after 2000.0) have significantly improved compared to ERP from LLR data reported previously by Biskupek, 2015, Hofmann et al., 2018. The Weighted Root Mean Square (WRMS) value (after 2000.0) of the uncertainty for estimation of DeltaUT1 lies between 17.03 µs and 24.49 µs for different subsets of nights. The WRMS values (after 2000.0) of the uncertainty for estimation of the terrestrial pole coordinates (estimated separately) for different subsets of nights lie between 1.30 mas and 3.46 mas for x_P and between 1.63 mas and 4.21 mas for y_P. The WRMS of differences between the estimated DeltaUT1 values from LLR analysis and those from the a-priori ERP series lie between 59.38 µs and 115.35 µs for a subset of NPs from all LLR observatories post 2000.0. For x_P, the differences are between 1.18 mas and 2.18 mas, and for y_P, the differences are between 1.01 mas and 1.76 mas ( x_P and y_P estimated separately). The differences of the obtained ERP from LLR analysis to the a-priori ERP series (for all subsets) become smaller with a stricter selection criteria (i.e. more number of NPs per night). Additionally, we see that the simultaneous estimation of the terrestrial pole offsets leads to high correlations between the estimated values of x_P and y_P of the same night. Furthermore, we estimate that the addition of non-tidal loading improves the uncertainties (3sigma values) of the estimated ERPs by about 1%.