Future Mission Designs (Future Gravity Missions)

Transcription

1 Future Mission Designs (Future Gravity Missions) AEI Hannover 1

2 Content The Gravity Field How to measure it? Why is GRACE important? What comes next GRACE Follow-On NGGM Mission defining parameters Accelerometry Interferometry 2

3 Earth's Gravity Field The gravity as the acceleration towards ground is ~9.780 m/s² at the equator and ~9.832 m/s² at the poles. Approximately 0.5% of weigh change due to Earth's oblateness and the centrifugal acceleration. Earth's gravity field is formed by the total mass distribution present due to Earth's core, crust, oceans, ice, water, atmosphere,... We use the geoid, a particular equipotential surface of Earth's gravity potential, to visualize the gravity field. After removal of the best-fit ellipsoid, we obtain the geoid undulation or height Important for (water) levelling, ~ mean sea level Conversion of the spatial gravity signal into the frequency domain via spherical harmonics (SH) yields the spectral decomposition SH degree plot Potsdam Potato

4 Time Variability of Earth's Gravity Field The Earth reacts to the gravitational pull of the Moon, Sun and other planets. These forces cause tides of the Solid Earth Oceans Atmosphere If the static part and the tides are removed from the gravity field, a variability of the order of 1 mm/s² becomes visible smaller than static gravity field driven by mass redistribution in the system Earth Atmosphere Oceans Hydrology Ice Solid Earth AOHIS gravity field changes provide fruitful science later slide Gruber, doi: /essd How to measure the gravity field? Terrestrial measurements with gravimeters are possible in well-developed countries [FG5 gravimeter by LaCoste] For global coverage need to go to space...

5 Spaceborne Gravity Measurements The satellite's orbital trajectory is influenced by the gravity field Measure orbit trajectory gravity field Global coverage with polar orbits Laser Geodynamics Satellite I und II LAGEOS Launch 1976 & 1992 Satellite Laser Ranging (SLR) for orbit determination SLR stations & view required Navstar GPS became fully operational in the 1990s CHAMP: CHAllenging Minisatellite Payload Operational Global gravity field from orbit perturbations Global magnetic field recovery (solid Earth & solar-terrestrial physics) Atmosphere/Ionosphere sounding with GPS occultation Accelerometer measures non-gravitational accelerations Wikipedia: Lageos GPS/GNSS based absolute 3-d positioning over hundreds and thousands of km has a limited accuracy limiting the gravity field recovery Relative measurements between close-by objects

6 Spaceborne Gravity Measurements II Gravity Recovery and Climate Experiment (GRACE) High-Low & Low-Low Satellite-Satellite Tracking Operational in 2002 Nov 2017 (TBC) Orbit height: 500 km 380 km ( ) ~cm accuracy for orbit positioning 1 µm = m precision for inter-satellite distance of ~200 km Accelerometer measure non-gravitational accelerations Measure force required to keep the proof-mass centered GOCE: Gravity field and steady-state ocean circulation explorer The Space Ferrari Operational Orbit height: 268 km Drag-compensation in along-track direction Used 6 accelerometer with 6 proof masses Science objective was the static gravity field with very high resolution.

7 GRACE principle In GRACE-like missions, most of the gravity field signal is in the the projected differential gravitational acceleration (PDGA) Some gravity field signal is in the GNSS information. Stabilizes gravity field solutions at long wavelengths Example of the range change over Himalaya 60 µm ~ thickness of a DIN A4 sheet of paper Image credit: AIUB, Ulrich Meyer et al..

8 Current State of Gravity Fields - Spectrum 200 km resolution 100 km resolution 200 km resolution.

9 GRACE Science Results [ Velicogna, I. (2009), Increasing rates of ice mass loss from the Greenland and Antarctic ice sheets revealed by GRACE, Geophys. Res. Lett., [O. Baur (2013): Continental mass change from GRACE over and its impact on sea level ] Increase ~3.3 mm/yr Figure 20: The GRACE data reveal the severity of California's drought on water resources across the state. This map shows the trend in water storage between September 2011 and September 2014 (image credit: NASA/JPL) Wang et all (2012), Coseismic and postseismic deformation of the 2011 Tohoku-Oki earthquake constrained by GRACE gravimetry, Geophys. Res. Lett. More than 1300 GRACE related publications

10 GRACE Science Results II Monthly solutions.

11 GRACE Science Results III daily GRACE gravity field solutions GRACE shutdown can be expected in 2017

12 GRACE Follow-On Quick successor mission for GRACE initiated by NASA/JPL Contributions from GFZ and Germany Launch slot currently for March/April 2018 SpaceX Falcon 9 shared with Iridium Polar orbit with 220 km S/C separation at ~490 km height (as in GRACE) Initially indented as copy of GRACE satellites with some improvements and evolved instruments An additional star camera head No multi-layer insulation foil at bottom side Slightly improved accelerometer noise More GNSS channels / New GNSS receiver New laser ranging instrument as technical demonstrator next to the conventional microwave ranging instrument AEI involved since 2010 Important for LISA: First inter-satellite laser ranging instrument Mission Objectives: 1st Continue series of high-resolution global gravity field maps 2nd Proof feasibility and show performance of a Laser Ranging Instrument (image credit: Airbus DS GmbH, Mathias Pikelj) What is the expected gain in gravity field accuracy?

13 GRACE FO Gravity Field Performance What is the beyond GRACE Follow-On?

14 Next Generation Geodesy Missions (NGGM) Regularly, studies on future gravimetric missions are conducted to be prepared for mission calls Usually address Science Requirements Satellite Design & Instruments Satellite Constellation NASA and Germany (GFZ) performed studies on GRACE Follow-On missions in the years Quick successor - now GRACE Follow-On GRACE-2 as more advanced mission GFZ studies called GRAF & 3M4C ( ) e.motion proposal submitted for ESA's Earth Explorer 8 call in 2010 Earth System Mass Transport Mission Next Generation Geodesy Missions studied by Thales-Alenia Space (2005) Airbus/Astrium (2010) NG2 Thales-Alenia Space (2010) NGGM-D: e.motion² study ( ) NGGM study by ESA Consolidation of the System Concep for the Next Generation Gravity Mission Airbus incl. AEI (not granted in 2015/2016) Thales-Alenia (Nov 2017) China interested as well in gravity field missions...

15 General Requirements for NGGM Future missions need to tackle various aspects to reduce the errors in the gravity field measurements A laser ranging system alone produces only a modest improvement in the gravity fields The primary goal is a higher resolution of gravity field maps Spatial resolution Temporal resolution Difficult due to uncertainty principle, e.g. for a single satellite pair, the spatial resolution S and temporal resolution T is bound [Pour et al., 2013] Need more satellites...

16 Formations & Constellations Simulated Satellite Formation Flights for Detecting the Temporal Variations of the Earth s Gravity Field (PhD thesis), Basem Elsaka, Bonn, 2010 GRACE, GRACE FO Pendulum: Polesvia RAAN mismatch Equator via inclination mismatch e.g. 200 km along-track, 50 km cross-track Bender Formation: 2 pairs, one inclined Cartwheel (LISA-like) Many more... Raw GRACE gravity field maps show significant North-South striping Alternative formations add East-West component to measurement Reduces striping More isotropic measurement Basically, all other constellations/formations show better performance 16

17 E.motion² Approach Bender configuration was selected over Pendulum Was not studied in detail before Less technical demanding, since low Doppler rates Slightly better gravity field solutions More demanding in terms of costs (4 S/C) Inclination of second pair (70 ) was a trade-off to return some stand-alone benefit Information on icy regions More appealing for funding by a second space agency Particular altitude bands exist as a function of the inclination, which have resonances at high orders (and therefore degrees) Idea: Shift these resonances to d/o, where no time variable gravity field signal is expected Lower susceptibility to temporal aliasing For details see Murböck et al GRACE has periods with poor ground track repeat cycles NGGM: use fixed repeat orbits such as α/β=478/31 Requires active drag compensation. 17

18 Altitude and S/C distance First order effects of the S/C distance L Gravity signal in the range (or PDGA) is proportional to L Parts of ranging noise (laser frequency noise) ~ L Signal-To-Noise ratio indep. of L at low frequency Spatial gravitational features shorter than L (high frequencies) are slightly suppressed L = 200 km SH degrees >100 are suppressed L = 50 km SH degrees >400 are suppressed Some studies assume constant ranging noise might lead to misleading results e.motion² selected L=100 km Frequency [Hz] [constant ranging noise!] Lower altitude allows to retrieve higher SH degrees, more of the gravity fine-structure Higher roll-off frequency Increases atmospheric air drag, higher noise Trade-off for science return vs. mission lifetime, instrument noise, propellant mass constraints,

19 Results of e.motion² (Bender) Gravity fields significantly improve with two satellite pairs not simply sqrt(2), but factor 10 for time-variable signal! better global coverage with inclined pair 2 S/C Gravity background models (undersampling) limits the performance of (future) gravity missions Steady improvements of models More satellite pairs are beneficial Design instruments to be not limiting Can not be improved in orbit. 19

20 Top Level Requirement The acceleration approach offers a linear (not linearized) relation between measurements and gravity field SH coefficients. Simple propagation of instrument errors into final gravity field Centrifugal term makes approach difficult to use for real data processing Commonly used to derive the top level requirement for GRACE-like missions Gravity Field SH coeff. Ranging (blue) Accelerometry (green) Centrifugal Acceleration (not measured directly) Need also to fix the frequency band 20

21 Measurement Band Science Measurement Frequency Band typically between 0.18 mhz. 0.1 Hz Upper limit: 0.1 Hz = fs/2 5s sampling rate Data rates for down-link limited Gravity signal rolls of at high frequency, makes no sense to down-link noise (SNR < 1) Lower limit: 0.18 mhz, e.g. once per orbit Low frequencies drive costs and complexity of instruments & testing In principle, gravity signal also at lower frequencies (repeat orbit frequency) Common mistake: To measure monthly gravity field signals, one needs instrument sensitivity at 1 µhz Time-varying SH coefficients produce sidebands at ±1 µhz w.r.t. to tones Need sufficient resolution in the frequency domain determined by observation length In e.motion² a calibration bandwidth was proposed between Hz Assessment of instrument health & use for in-orbit calibrations Orbital velocity ~7600 m/s determines the frequency 21

22 Accelerometers CHAMP, GRACE, GRACE-FO & GOCE use(d) servo-accelerometer by Onera Capacitive Sensing (CS) Proof-Mass recentered with high loop gain Linear and rotational accelerations are measured To measure purely non-gravitational accelerations ACC PM CoG needs to be co-located with S/C CoG Gravity gradient G may produce bias CoM/CoG trim unit on S/C General model for linear acceleration measurement:.

23 GRACE (FO) - Accelerometer GRACE SuperSTAR evolved from STAR (CHAMP) ~10x improved resolution Gap size 175 µm (CHAMP 75 µm) Reduced dynamic range higher resolution GRACE Follow-On Modest improvement in noise expected Separated Front End Electronic from SU change from 7 µm diam gold wire to 6.5 µm platinum wire after test failure Christophe, Onera, Development status of GRACE FollowOn accelerometers, GSTM 2015 SuperSTAR accelerometer with the sensor unit (right) and the ICU (left), image credit: ONERA; Frommknecht,Schlicht; The GRACE Gravity Sensor System Najda Peterseim, TWANGS High-Frequency Disturbing Signals in 10 Hz Accelerometer Data of the GRACE Satellites. 24

24 Accelerometry in Future Missions High-frequency twang artifacts hopefully not present in GRACE Follow-On Dynamic range (max acceleration) might be insufficient e.motion² finding: saturation above 400 km unlikely Future instruments might reach level of 1e-11 m/s² Hz Need a DC-scale factor stability of 0.1% 0.01% GRACE-FO scale factor: 2% Scale factor accuracy and variability might be problematic in future missions Driving is Voltage to Force & Acceleration conversion, Loop gain stability,.. e.motion² proposal for accuracy (DC) Drag compensation requirement within MBW Calibration scheme to get 0.2% Scale factor stability (AC): reduce the DC-load of accelerometer with drag compensation system. 25

25 Calibration Scheme ACC scale factor Assumption: Drag-compensation is available. Average air drag of 50 µn. 50 µn drag compensation with thrusters required to keep orbit height and ground-track repeat pattern Calibration scheme idea: Do not use constant 50 µn thrust, but modulate it sinusoidally between 5µN 95 µn with modulation frequency of 0.1 Hz (S/C1) and 0.16 Hz (S/C2) Each accelerometer will measure the sinusoidal (calibration tone) amplitude The calibration tones will appear also in the LRI distance measurement LRI-based ranging is sufficiently precise to determine ACC DC scale factor to 0.2% Calibration does not increase propellant consumption. 26

26 Drag-Free vs. Drag-Compensation Distinction in literature not established Proposed by myself Gradual difference between drag-compensation and drag-free possible Two nested control loops Bandwidth of the AOCS loop main loop Ideally both concepts are equivalent (under optimal conditions) GOCE used drag-compensation LISA pathfinder is prototype of drag-free Performance not applicable to LEO orbit: temperature stability, non-gravitational disturbances Advanced techniques: gravitational balancing, thermal modeling Future missions might change philosophy of the reference point CoG/CoM of S/C not well defined, subject to thermo-elastic deformations, propellant consumption, CoG/CoM of PM well defined Reduce load on the accelerometer by drag reduction is a good idea 27

27 Interferometry - Basics GNSS/GPS (phase) Ranging, Microwave Ranging, Laser Ranging, Doppler Tracking - but not SLR rely on the same principle Readout of the phase of an electro-magnetic wave (SLR: measures time difference) The phase of EM waves changes linearly with Time: propagation distance (to first order): Phase Readout Mix your received EM wave (RX) with a Local Oscillator (LO) of the same frequency Low-Pass Filter Result: Voltage/Signal depends on the phase difference! Interferometry measures phase differences What type of EM waves do we use? One characteristic is the wavelength (frequency) GPS: 19cm/24cm ( / MHz) K/Ka band Microwave Ranging: 0.9/1.2cm (24/32 GHz) Laser Ranging Interferometry: 1064 nm = µm (282 THz) Coarse vs Fine Ruler! RX LO GPS (20cm) K/Ka Band (1cm) 28 Laser (0.001mm = 1µm) Interferometry (phase) always a biased range integer ambiguity 28

28 GRACE (FO) - Interferometry How/Where is the EM radiation produced? GPS Electronics: ~ 50 W on TX S/C, LO: electronics Frequency Stability: Atomic Clock for RX, USO for LO Microwave Ranging (GRACE like) Electronics (K/Ka Transmitter), 23 dbm = 200mW Frequency Stability: USO For TX & LO Laser Ranging Interferometer (GFO) Laser, ~25 mw optically Frequency Stability: ULE Cavity For TX & LO TESAT GRACE Follow-On Laser USO OXCO by Oszilloquartz 29 Laser frequency stabilization cavity and optical bench mounted on vacuum flange (image credit: NASA/JPL) Frequency or Phase fluctuations = varying conversion/scale factor 29

29 Frequency Standards Stability of frequency standards given in Allan deviation Spectral density Optical cavity higher stability than USO Can't we use an optical cavity as electrical frequency standard? Conversion from optical frequencies to electrical not easily possible The frequency noise for all traces is referred to 1064 nm wavelength. Laser frequency standards Optical cavities: provide stability but many resonant frequencies Atomic or Molecular standards: absolute frequency Should be considered in future missions: accuracy better and for laser acquisition NGGM: Large improvement in stability for spacequalified reference unlikely Frequency converts radian to physical distance (meter) 30

30 GRACE (FO) Interferometry Where does the mixing occur? GPS/GNSS Directly to DC Phase-Readout in the GPS Receiver K/Ka Ranging 1st: Mixing to ~1MHz (electrical) 2nd: Phase-Readout in the GPS Receiver Laser Ranging: 1st: Optical Mixing to ~4..20 MHz (electrical) 2nd: Phase-Readout in the Laser Ranging Processor Dual Band Ranging (K + Ka, L1 + L2 GPS) EM waves at GHz affected by Ionosphere Phase advance Effect at optical frequencies (282 THz) negligible (1/f² dependency) 31 Phase tracking internally with high sampling rate (~MHz) 31

31 Phase Measurement Noise Sources S/C Pointing Related Noise LRI: S/C attitude jitter noise, a dominant noise coupling ~ 100 µm/rad KBR: Antenna Offset Correction, coupling 1.5m/rad GNSS: Antenna Offset Correction, coupling: 0.4m/rad Phase-Center Variation Map (PCV), coupling ~mm/rad Readout-Noise (low received power) LRI: Negligible KBR: called system-noise, 1 μm/rthz, dominant at high frequencies GNSS: important / dominating at high frequencies Frequency Noise (wavelength fluctuations) LRI: a dominant noise source KBR: small/negligible GNSS: not dominant (?), atomic clocks, well measured with stations Timing Jitter Noise via USO Frequency Noise i.e. 1 m/s with 1ns/rtHz 1 nm/rthz noise LRI: not limiting, maybe visible at 0.1 mhz KBR: typically called oscillator noise GNSS: not important, clock offset estimation S/C pointing jitter noise can be reduced in post-processing Estimation of precision difficult as signal >> error; correction/fit disturbed by signal LRI mitigated this contribution by design, but remaining part still considered dominant Pointing noise not shown

32 GFO Noise Estimates Pointing jitter difficult to estimate due to proprietary AOCS 33

33 More Noise & Error Estimates Time Of Flight Corrections Special Relativity Ionospheric Effect in LRI General Relativistic Effects Neutral Atmosphere Effect in LRI. 34

34 Effects to be considered. 35

35 GRACE Follow-On Initially accommodation issue for LRI: Line-of-sight not available 36

36 GRACE FO LRI Optical Layout Racetrack configuration with corner-cube retro-reflector (Triple Mirror Assembly, TMA) 60 cm lateral separation, space constraints due to KBR & cold gas tanks TMA vertex ( phase center ) separated from physical structure co-located with CoM/ACC, [wikipedia/corner-cube] lever arm for rotation coupling only ~100 µm Corner-Cube Optical Bench Quadrant Photodiodes measure interfered light 2-axes steering mirror (SM) automatic beam alignment acquisition 2-lens beam compressor image aperture & SM onto PD, resize beam suppress beam walk & diffraction rings D. Schütze/AEI Distance measured between TMA vertices (= S/C CoM) 37

37 OBA & Triple Mirror Assembly. 38

38 LRI Automatic Beam Alignment and Pointing Quadrant photodiodes and multi-channel phasemeter enable to form Average phase of segments ~ ranging Differential phase of segments (DWS) ~ relative beam tilt & tip between RX & LO LRI utilizes a control loop to zero DWS by means of a steering mirror with high gain and bandwidth Optimal wavefront overlap between LO & RX Maximizes SNR for phase readout Higher common-mode error rejection 39

39 LRI Automatic Beam Alignment and Pointing Quadrant photodiodes and multi-channel phasemeter enable to form Average phase of segments ~ ranging Differential phase of segments (DWS) ~ relative beam tilt & tip between RX & LO LRI utilizes a control loop to zero DWS by means of a steering mirror with high gain and bandwidth Optimal wavefront overlap between LO & RX Maximizes SNR for phase readout Higher common-mode error rejection Beam tilt & tip caused by local S/C yaw & pitch misalignment DWS zeroed Dedicated steering mirror angle sensing Closed Loop Operation OB in&out waves parallel OB enhances light power, no light deflection TMA retro-reflection sends beam to distant S/C Optimal TX beam pointing Allows: S/C pointing error > LRI beam pointing requirement 40

40 Ideas for LRI in NGGM Smaller retro-reflector e.g. 38 cm Higher stability, better accuracy in terms of co-alignment Acquisition search of laser links in GFO LRI is a complex 5 d.o.f. search Dedicated acquisition sensors would be beneficial Absolute frequency standard on both S/C Stability comparable to cavity Frequency shifting components required Reduction of complexity in acquisition phase Higher accuracy for wavelength = scale factor Dual One-Way Ranging scheme would allow 2 lower frequency noise Fallback redundancy option: Transponder Retro-reflection on-axis with alternative optical layout possible Allows in principle a telescope More complex due to more polarization optics Thilo Schuldt et al 2017 J. Phys.: Conf. Ser

41 Gravitational Physics Ranging Part Gravimetric space missions, such as GRACE Follow-On, and gravitational-wave observatories, such as LISA,... measure distance variations in terms of light's phase (photon time of flight) between probes measurement of phase, not time (such as in Lunar Ranging) measure with a low noise in the picometer (LISA) or nanometer (GRACE FO) regime are interested in signals with frequencies in the band: ~0.1 mhz 0.1 Hz Gravitational waves are ripples in space-time Probes are quasi-static but free-falling Ripples in space-time propagate with c Gravity field corresponds to space-time curvature Structure of space-time is quasi-static Probes are moving (free-falling) with orbital velocity [LIGO] Both effects cause a modulation of the light's phase, i.e. of the distance 42

42 Sensitivity of Space-Time Measurements. 43