Reprocessing of KBR1B data on GRACE-FO and its advantages

The K-Band Ranging system (KBR) is the key payload to measure the inter-satellite distance variations with the accuracy of sub-micrometers in GRACE and its successor GRACE-FO missions. In addition to KBR, GRACE-FO is equipped with a laser interferometer (LRI) with higher precision at the nanometer level. The KBR observations are not only used as the main input data for the gravity field recovery but also play an irreplaceable role in re-scale the biased range measured by LRI. The KBR data processing from Level-1A to Level-1B converts the measured raw phase to the inter-satellite biased range (as well as range rate and range acceleration) mainly by correcting time-tags, fusing the phase for each frequency, correcting ionospheric effects, and low-pass filtering. Furthermore, the light time correction (LTC) and antenna offset correction (AOC) are computed. Although the KBR-LRI range residuals have been recently analyzed with the conclusion that they are mainly limited by time-tag errors at low frequencies, we still attempt to better understand the minor remaining deviations. In this study, we re-process the KBR data from Level-1A to Level-1B in alternative ways by mainly investigating:

1) to use an alternative method to fill gaps that occur sporadically in either the K or Ka channel based on transplanting the phase from the available to the missing channel and linearly interpolating the ionospheric effect. If both channels are missing, we propose to use interpolation ; 

2) to smooth the overlap of time-tags at day boundaries, which can reduce the jumps in bias range at day transitions;

3) to take into account time-varying carrier frequency with intraday variability when converting the phase to range; 

4) to improve the accuracy of LTC in low-frequency areas by using the range rate from KBR measured instead of the range rate from POD and, at high frequencies, reduce the numerical limitations by using analytical equations instead of an iterative equation;

5) to consider the movement of the satellite’s center of mass when computing the AOC.

Although these adaptations to KBR1B data are minor and seem to have no obvious differences in terms of gravity field maps at the current precision level, this study is useful to better understand the relationship between KBR and LRI and for the instrument characterizations in the future more precision missions.