GRACE

1
GRACE GRAVITY RECOVERY AND CLIMATE EXPERIMENT
The GRACE Mission Status and Latest Results M.
M. Watkins JPL B. D. Tapley University of Texas

Hydrology/Surface Water WG Meeting Irvine,
CA March 22, 2004
2
(No Transcript)
3
Background - Gravity Mapping
4
Gravity Mapping - Current State of the Art

• Currently, the best global gravity field models
  come from a combination of tracking data from a
  large number of spacecraft, satellite altimeter
  data, and surface gravimetry data
• Example
• JGM-3 (Joint Gravity Model 3) - produced
  collaboratively by the University of Texas and
  NASA/GSFC used
• tracking from over 30 satellites (and going back
  30 years!),
• ocean altimetry from GEOS-3, Seasat, and Geosat
• land surface gravimetry
• RMS geoid error over ocean 25 cm (cumulative
  to 70x70) due to extensive cheating by using
  altimetry data. Errors are worse over land.

5
How to Improve Things?

• Fly a low altitude spacecraft that can be tracked
  very accurately
• CHAMP, various other LEO GPS s/c, STEP (maybe)
• Fly two lower s/c whose relative motion can be
  tracked very accurately
• GRM, GravSat, GAMES, GRACE
• Fly a gradiometer
• Aristoteles, QuickSTEP, M3 STEP, SGGM, EGGS,
  GEOID, GP-B (1 axis), GOCE,...?

6
NASA/European Gravity Program

• Gravsat - 1970s
• GRM - 1980s
• SGGM - 1980s
• EGGS - 1995
• TIDES - 1991
• CIGAR - 1990s
• ARISTOTELES - 1980s - 90s
• GAMES - 1990s
• GRACE - launched 2002
• GOCE - expected launch 2006

7
GRACE Measurement Concept

• GRACE Measurement Distance Change Between KBR
  Antenna Phase Centers (P.C.)
• ??(t) ? r1(t) - r2(t) True
  C.G.-to-C.G. Range Change
• C.G. Correction C.G.-to-P.C. Vector
  Baseline
• (Calibrated relative to Star Camera)
• Dimensional Variations e.g. - Thermal
  distortion of structure
• - C.G. Variations in satellite frame
• Measurement Errors e.g. - Multipath Errors
• - Electronic Noise
• - Time-tag Errors, etc.

8
Expected Accuracy
Kaula's
Power Law
Mean Ocean
(POCM)
Geoid Height ( mm )
EGM96
GRACE
90d, 450 km
Hydrology (Ann)
Ocean (Ann)
GRACE
90d, 300 km
Half Wavelength ( km )
9
GRACE Expected Performance (2)
Imagine a disk of some radius. How thick must
that disk be so that GRACE can feel its
gravitational effect?
10
State of the Mission
11
GRACE MISSION
Science Goals High resolution, mean and time
variable gravity field for Earth System Science
applications.
Mission Systems Instruments HAIRS
(JPL/SSL/APL) SuperSTAR (ONERA)
Star Cameras (DTU) GPS Receiver
(JPL) Satellite (JPL/Astrium) Launcher
(DLR/Eurockot) Operations (DLR/GSOC) Science
(CSR/JPL/GFZ)
Orbit Launched March 17, 2002 Initial Altitude
500 km Current Altitude 476 km (30
m/d) Inclination 89 deg Eccentricity
0.001 Separation Distance 220 km Nominal
Mission 5 years Non-Repeat Ground Track, Earth
Pointed, 3-Axis Stable
12
GRACE Project Status

• Spacecraft System
• Launched 0921 UTC, March 17, 2002
• Reached target orbit
• Perigee/Apogee 495.526/501.163 km
• Eccentricity 0.0004
• Inclination 89.025
• Satellites in science collection mode
• Mission is in Validation Phase
• AOCS and Thermal Tuning
• Flight Software Improvement
• Loss of some redundancy on GRACE-1
• Lifetime Estimate is greater than eight years
• Mission Operations
• GSOC successfully operating twin satellites in a
  multi-mission environment
• Over 99 science data recovered from satellites
  (science housekeeping)
• Science Data System (JPL,CSR,GFZ)
• Second Generation Gravity Models

13
State of the Flight Segment
(Or, how can you tell if youre actually ranging
between two spacecraft 200 km apart with an
accuracy of less than the diameter of a red blood
cell?)
14
K-band Ranging performance

• Requirement
• 10-4 cycles _at_ 1-Hz, single link
• lt 1 mm/s range-rate, 4 links 0.2 Hz
• 4 links are K/Ka from each GRACE s/c

(K 0.75KA)A (K 0.75KA)B _at_ 10Hz
1-Hz RMS 0.887 microns 0.2 Hz RMS 0.396 microns!
15
Histogram KBR Range-Rate Residuals, Gravity Fit
Dual 1-Hz Star Camera
1-Hz, 0.2 Hz Star Camera
16
Relative Calibration using Dual ACC Data

• The instruments are aligned by position in
  Earth-fixed frame
• Aligned data can be used for relative bias
  scale calibration
• Residual relative to calibration is an upper
  bound on noise

17
Relative Calibration Residuals

• Relative calibration residuals depend on
• ACC measurement errors
• measurement noise
• thermal variations
• Residual variability of attitude density
• Differences from flying forwards backwards
• Upper Bound Error RMS in 1 to 35 mHz bandwidth
• X (cross-track) 0.44 nm/s2
• Y (radial) 0.50 nm/s2
• Z (along-track) 0.40 nm/s2

Figure S. Bettadpur
18
GRACE Satellite-Satellite Range Performance
Full sat-sat Range - Bias
After removing long period part
Topography Along Groundtrack
19
GRACE range rate response to gravity field
20
State of the Science
21
GRACE Gravity Solutions (GGM02)

• GGM02S
• Estimate 160x160 using only data from GRACE
• 13 months of GPS, KBR, ACC and SCA data used
  (April 02 - December 03)
• No Kaula constraint, no other satellite
  information, no surface gravity information and
  no other a prior conditioning
• GGM02C
• Combine GGM02S with surface information to
  200x200 and consistent with EGM96
• Designed to be extendable to 360x360 with EGM96
  coefficients

Geoid height ( m )
The geoid is the level (equipotential) surface
that best coincides with mean sea level The geoid
height varies by 200 m, but oceanographic
applications need this to be determined to cm
accuracy
22
Progress in Gravity Field Resolution
Decades of tracking to geodetic satellites
111 days of GRACE data
13 months of GRACE data
Detailed geophysical features are being detected
by GRACE with no surface gravity inputs and no
satellite altimetry
Latest model shows more detail because less
smoothing is required to remove artifacts
23
Progress in Gravity Field Resolution
Details of Tonga/Kermadec Region
Decades of tracking to geodetic satellites
111 days of GRACE data
13 months of GRACE data
24
Zonal Geostrophic Currents
Determined from hrel to 3000-4000m (hrel
calculated from WOA by V. Zlotnicki)
CSRMSS98 - EGM96
CSRMSS98-GGM 01
eastward
25
Improvements to Marine Geoid
GGM01C
GGM02S
Zonal Currents
Meridional Currents
Currents computed as CSR MSS - marine
geoid Residuals relative to Levitus long-term
hydrography
26
Gravity Errors Predicted by Full Covariance
Geoid errors from GRACE are much more uniform,
without land/sea discrimination
Predicted geoid height errors for EGM96
Predicted geoid height errors for GGM01S
Errors as large as 50 cm
Errors less than 2 cm
Predicted geoid height errors for GGM02S
Differences EGM96-GGM01S
Differences gt 1m max
Errors less than 9 mm
at 300 km resolution (degree/order 70)
27
Progress in GRACE Gravity Models
Previous error estimates were generally
consistent, if not pessimistic, with the
now-known errors (as determined by comparing them
to most accurate current model GGM02)
Feb 04
Sep 02
Feb 03
July 02
28
Time Varying Gravity Results
GRACE TIME VARYING SIGNALS (PRELIMINARY GRACE
RESULTS)
Preliminary results (Tapley, Watkins, and
Reigber, Science, 2004) for the observation of
the time varying field show great promise GRACE
observations (top) match the broadly expected
signals (bottom) from land hydrology and ocean
models (ECCO/JPL) constrained with data.
29
Time Variable Gravity Comparison
August 2002 to April 2003 GRACE using current
data release
Annual variability from GRACE is based on 14
months (4 in 2002, 10 in 2003) Smoothed to 600
km resolution (previous GRACE results shown to
2000 km resolution) Not everything in these maps
will be continental hydrology there will be
unmodeled variability in all the Earths systems,
such as the baroclinic ocean variability
Annual cosine component
Annual sine component
30
Time Variable Gravity Comparison
August 2002 to April 2003 GRACE using current
data release
August 2002 to April 2003 Hydrology model
Annual cosine component
Annual sine component
The major difference is due to an incorrect phase
from the NCEP model which does not model snow
cover properly.
31
Conclusions

• GRACE is on orbit and operating well
• Exciting early results with the best still to be
  achieved
• Significant improvement in mean field
• GGM02S satisfies NASA Minimum Mission
  requirement
• Very significant impact already on oceanographic
  modeling and data analysis
• Time varying gravity signal has been measured
• Good correlation with expected hydrology signal
• Errors in Hydrology Models can be observed
• Eleven monthly solutions determined preliminary
  annual signal
• Feel free to contact us
• michael.watkins_at_jpl.nasa.gov
• tapley_at_csr.utexas.edu
• srinivas_at_csr.utexas.edu