Research On The Raw Data Processing Technology Of The Inter-satellite Ranging System Of GRACE/GRACE-FO Gravimetric Missions

The gravity field of Earth is an important physical quantity that characterizes the distribution and variation of mass in the interior and surface of the earth,and has an important impact on national defense security,global climate change and human life.The GRACE/GRACE-FO gravimetric mission is a revolutionary advancement in gravity field measurement.It provides for the first time high-precision,all-weather,global mid-to-long space-scale monthly resolution time-varying gravity field information.In addition to benefiting from satellite scientific formation design,high-precision payloads,and highstability satellite platform,it also relies on decades of accumulation and development of its data processing technology in the Science Data System on the ground.The data products of GRACE-like missions are classified according to the processing flow: Level-0,Level-1A,Level-1B,Level-2,Level-3.Among them,Level-0 to Level-1B data processing is called raw data processing,and it is the key step to all subsequent data product processing.Since JPL(Jet Propulsion Laboratory)and GFZ(Geo Forschungs Zentrum)only published Level-1B products for GRACE missions and Level-1A,Level-1B products for GRACE-FO missions,and the key technical details of the raw data processing have been kept secret.Therefore,the published papers of the raw data processing technology are much few.Gravity satellite mission in China has been designed and the hardwares have been developed in the past decades,but it urgently needs the development of raw data processing technology.The inter-satellite ranging system is the key payload of GRACE-like missions,and its raw data processing technology directly affects the results of gravity field recovery.This article focuses on the raw data processing of the inter-satellite ranging system in GRACE-like missions.Firstly,we describe the concept of gravity field of Earth and its measurement strategy in space(eg.Gravity satellite),and then analyzes the measurement principle and different types noise of the inter-satellite ranging system on GRACE/GRACEFO.On this basis,the raw data processing technology of the inter-satellite ranging system has been researched and verified based on Level-1A and Level-1B data products of GRACE-FO.Finally,the delay problem of inter-satellite ranging signal propagation is analyzed in depth.The thesis mainly includes the research progress in the following aspects:(1)Principle and noise of inter-satellite ranging systemGRACE is equipped with a K-band ranging system(KBR),and GRACE-FO is equipped with both a K-band ranging system(KBR)and a laser ranging interferometer(LRI).First of all,this article derives the changes of the physical quantities in the ranging system from the frequency of the ranging signal to the inter-satellite distance and the phase of the ranging signal to the inter-satellite distance,so that the ranging theory of inter-satellite ranging has a wider application range.Then the noise distribution of the two ranging systems and the influence on the ranging signal are analyzed in detail.The low-frequency noise of KBR mainly comes from the noise of the Ultra-Stable Oscillator(USO),and the high-frequency noise mainly comes from the thermal noise of the system.The lowfrequency noise of LRI mainly comes from the Tilt-to-Length(TTL)noise which caused by the changes of satellite attitude coupled into the range measurements,and the highfrequency noise mainly comes from the laser frequency noise.(2)The flow and verification of the raw data processing technology of KBRThe ranging accuracy of the KBR is 1?m,but the original phase observations contain a lot of noise as well as the time-tags between two satellites are not synchronized.Firstly,we independently propose the raw data processing technology(from Level-1A to Level-1B)of KBR.The main components are the bias range which is directly observed by the ranging system,the light time correction(LTC)caused by the limited speed of light and the delay on propagation path,and the antenna offset correction caused by the misalignment of the measurement reference point of KBR with the center of mass of the satellite.The processing of the bias range mainly includes key steps such as time-tags correction(double-satellite time synchronization),phase reanalysis,elimination of ionospheric delay,low-pass filtering and down-sampling.The difficulty is that the changes of phase depend on the time system and the frequency of the ranging signal.Therefore,it is necessary to accurately calculate the phase observation change caused by the change of time and frequency and the coupling effect of the two when to obtain high-quality and low-noise phase observation products.The processing of LTC is mainly to compute the light travel time between two satellites.The processing of the antenna offset correction is mainly to convert the inter-satellite distance from the KBR reference frame(KF)to the line-of-sight frame(LOSF).Then compare official product JPL?KBR1B which is published by JPL and our product HUST?KBR1B which is based on our method.The standard deviation(STD)of the difference of bias range between two products is about rms=1nm,and the STD of the difference of the antenna offset correction range is about rms=5nm,they are far lower than the ranging noise level of KBR.Finally,the difference in gravity field inversion between using HUST?KBR1B and JPL?KBR1B products is compared.The post-fit residuals of bias range rate obtained by using the two KBR1 B products are close(the differences are about tens of nanometers),the maps of spatial distribution of the gravity field look same,and the geoid degree variances are close(the differences are below 1mm).(3)The flow and verification of the raw data processing technology of LRIThe ranging accuracy of the laser interforometer is about 2-5nm,but the original phase observation contains a lot of phase jumps in addition to a large amount of noise and asynchronization of the binary time scales.First of all,we independently propose the raw data processing technology of LRI(from Level-1A to Level-1B),and the main components are the bias range which is directly observed by the ranging system and the LTC.The processing of bias range mainly includes key steps such as time-tags correction(doublesatellite time synchronization),phase reanalysis,phase jump removement,low-pass filtering and down-sampling.The difficulty is that the changes of phase depend on the time system and the frequency of the ranging signal,and there are a lot of phase jumps in the phase observations.Therefore,in order to obtain high-quality and low-noise phase observation data products,it is necessary to accurately calculate the phase variations cause by time and frequency changes,build optimal phase jump model to remove the phase jumps in both of two satellites.The processing of LTC is mainly to compute the light travel time of laser between two satellites.Then we compare our product HUST?LRI1B with the official product JPL?LRI1B which published by JPL and the product AEI?LRI1B which published by AEI(Albert-Einstein-Institut).The results show that HUST?LRI1B is close to AEI?LRI1B in the high-frequency areas and is better than JPL?LRI1B.In the lowfrequency areas,there are big differences between the three type products at 1CPR and2 CPR due to the different estimation methods of the time scale and the scale factor.In the comparison of the gravity field inversion results,the post-fit residuals of HUST?LRI1B and AEI?LRI1B are close,and the high-frequency error is much smaller than JPL?LRI1B,but the three type products are close in spatial distribution and geoid degree variance(the differences are below 1mm).(4)The high-precision calculation of light time correctionThe propagation time of microwaves or lasers between satellites is delayed by the highspeed movement of satellites,residual atmosphere in the space,and path curvature caused by the effect of general relativity.We analyze the effects of the above three factors on the propagation time in detail.The delay of the high-speed movement of satellites has an impact on the KBR/LRI is about 4.8m.The delay of inter-satellite residual gas(mainly component is the ionosphere)has an impact on the single-way ranging of KBR by 13 mm,but after dualfrequency combined ranging is used,the impact will be measured in pm.Its influence is on the order of nm for LRI,since the frequency of laser is much higher than that of microwave.We revisit the calculation of the LTC from the photon motion equation considering general relativistic effects and state-of-the-art models of Earth's potential field.The novel analytical expressions for the LTC of KBR and LRI can circumvent numerical limitations of the classical approach.The dependency of the LTC on geopotential models and on the parameterization is studied,and afterwards the results are compared against the LTC provided in the official datasets of GRACE and GRACE Follow-On.It is shown that the new approach has a significantly lower noise,well below the instrument noise of current instruments,especially relevant for the LRI,and even if used with kinematic orbit products.This allows calculating the LTC accurate enough even for the next generation of gravimetric missions and even other space missions.