Time Variations of the European Gravity Field, 1997-2001

1
Time Variations of the European Gravity Field,
1997-2001

• David Crossley, Saint Louis University, Missouri,
  USA
• Jacques Hinderer, EOST / IPG, Strasbourg, France
• Jean-Paul Boy, NASA Goddard Space Flight Center,
  Maryland, USA

2
Global Geodynamics Project Phase 1 1997-2003
IUGG SEDI initiative Workshops Brussels 1997,
Munsbach 1999, Jena 2002
3
GGP Stations 1997 - 2003
4
GGP Satellite Project

• CHAMP and GRACE satellite calibration and
  validation
• Provides surface gravity measurements that are
  independent of satellite observations, compared
  to other methods that rely on modeling
• Goal is to find and interpret coherent seasonal
  gravity effects using European GGP ground stations

5
GGP and Satellite Missions
6
Ground and SatelliteGravity Contributions
7
Objectives

• Use Superconducting Gravimeter (SG) data from the
  European sub-network of GGP
• Compute residual gravity series for each station,
  July 1997 to December 2001
• Combine series spatially into a surface that
  approximates the length scale of GRACE data over
  Europe (200 -1000 km)
• Estimate the error in this surface and compare to
  GRACE accuracy predictions (0.4 mgal at 300 km)
• Compare with existing CHAMP satellite data
• Set up a semi-automatic procedure for producing
  European surface gravity for the duration of the
  CHAMP and GRACE missions

8
Simulation of Hydrology Recovery 5 year model
for Manaus, Amazon Basin (Wahr et al., 1998)
9
GGP Stations Europe 1997 - 2003
BE Belgium BR Brasimone PO Potsdam MB Membach MC
Medicina ME Metsahovi MO Moxa ST Strasbourg VI
Vienna WE Wettzell BH Bad Homberg WA Walferdange
10
Processing GGP Data

• Start with uncorrected ICET 1 minute files
• Fix pressure - linear interpolation for gaps
• Remove local tide (solid and oceanic components)
  and local barometric pressure
• Fix gaps and remove spikes replace with linear
  interpolation
• Decimate to 1 hour, remove IERS polar motion

11
Tides, Local Pressure and Polar Motion Removed
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
Processing Drift and Offsets

• Simultaneous Drift and Offset Estimation VERY
  IMPORTANT TO DO THIS RIGHT
• Offsets need to be checked against station log
  files
• Drift and offsets can differ even between two
  spheres of a single instrument
• Drift and offsets should be consistent between SG
  residuals and AG measurements

17
Wettzell - Initial Drift and Offsets CD029
Uncorrected GGP 1 minute data from ICET
18
Strasbourg Residuals - SG, AG and Polar Motion
19
Medicina Residuals, SG AG
Corrected for tides, air pressure, polar motion
(Romagnoli et al. 2003)
20
SG Drift Estimation (mgal / yr)
21
Station Residuals BE to MO
22
Station Residuals PO to WE2
23
Atmospheric Loading Models

• Local pressure, admittance 0.3 mgal / hPa
  includes direct attraction loading
• Global pressure using only surface data, e.g.
  ECMWF. Includes treatment of oceans as static,
  IB, or non-IB. Thin atmosphere no vertical
  structure (e.g. Boy et al., 1998)
• Global (p,T) - as above, but model r(h) as a
  function of surface (P,T) perfect gas to 20
  km (e.g. Boy et al., 2002)
• Full 3-D vertical structure from actual 3-D
  meteorological data. Computationally intensive,
  but important at seasonal periods (Boy et al.,
  2002).

24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
Atmospheric loading, station ME
28
Combined Daily Residuals
29
Minimum Curvature Surface

• For each day fit a minimum curvature surface to
  all the stations
• This surface goes through every station and
  therefore does no spatial averaging

30
Minimum Curvature Surface, All Stations
31
6-month Samples, all Stations
32
Spatial Smoothing

• Insufficient data to reconstruct a smoothed
  field from spherical harmonic analysis
• Insufficient station data to fit a surface
  directly
• Instead, fit an nth degree polynomial 2-D
  surface to the minimum curvature surface

33
Comparison of Surfaces
Minimum Curvature
3rd Degree Polynomial
34
3rd Degree Surface All Stations
35
Reconstruct Gravity at Stations

• Select the gravity values on the surface at the
  stations
• Average all stations for each day (mean surface)

36
(No Transcript)
37
(No Transcript)
38
Select Central Stations
39
New Surface Using Central Stations
40
Central Stations, 6-Month Intervals
41
(No Transcript)
42
Northward Event 99
Apr 99
Jan 99
Mar 99
Feb 99
May 99
43
Northward Event 99
Jun 99
May 99
Aug 99
Jul 99
44
So far

• GGP data shows clear annual signal for 3 years
  (97-00), but not for 2000-01
• Accuracy at 0.8 mgal, comparable to predicted
  GRACE (0.4 mgal) at 300 km.
• Atmospheric loading - difference between global
  (p,T) and local p - affects the mean gravity
  field at about the 0.5 mgal level.
• No attempt yet to estimate secular changes over
  Europe

45
Interpretation of Annual Signals

• Local
• Instrument effects, thermal anomalies,
  vegetation, groundwater, surface water, soil
  moisture
• Regional and global
• Atmospheric pressure (3-D) attraction / loading
• Ocean circulation, loading
• Hydrology
• Soil compaction
• Zerbini S., B. Richter, M. Negusini, C.
  Romagnoli, D. Simon, F. Domenichini, W. Schwahn
  (2001) EPSL 192.
• C. Romagnoli, S. Zerbini, L. Lago, B. Richter, D.
  Simon, F. Domenichini, C. Elmi, and M. Ghirotti
  (2003) in press.

46
Annual Signals at Medicina
47
GPS and Gravity at Medicina

• GPS (upper), gravity (lower)
• Expected anti-correlation
• Reasonable admittance -18 mm vs 6 mgal (peak to
  peak)

48
Continental Water Storage

• Model uses estimated precipitation, downwelling
  radiation and near surface atmospheric conditions
• Soil saturates and recharges groundwater that
  partially recharges surface water model includes
  evaporation
• Monthly water storage is convolved with a Greens
  function to estimate gravity attraction and
  elastic deformation.
• van Dam, T., Wahr, J. Milly, C., and Francis
  O., (2001) J. Geodet. Soc Japan.

49
Global loading results

• Model results from some GGP stations
• Includes attraction and loading
• Note maxima usually in winter

50
Water Storage Statistics
51
(No Transcript)
52
Comment

• Residual annual gravity in Europe has a similar
  amplitude 2.5 mgal
• and phase (probably)
• as estimated water loading
• To proceed further, we need CHAMP and GRACE
  satellite data

53
Other Possibilities (1) Japan (2) Greenland
54
Japan Ocean Hemisphere Project
National Astronomical Observatory, Misuzawa
(Esashi, Canberra and Ny-Alesund) Kyoto
University (Bandung) National Institute for Polar
Research, Tokyo (Syowa)
55
GGP Stations - W. Pacific
BA Bandung CB Canberra ES Esashi KY Kyoto MA
Matsushiro WU Wuhan
56
Greenland Uplift
Kellyville

• Post-glacial uplift from 3 different models
• ICE-3G Greenland component only
• HUY2 ignoring ice changes during last 4000 yr
• ICE-3G - variability outside Greenland
• J. Wahr, T. van Dam, K. Larson, and O. Francis,
  (2001) JGR, 106

Kulusuk
57
Greenland GPS
58
Greenland AG
59
Greenland Interpretation

• GPS uplift rate as a function of increasing data
  length
• Kellyville multi-day averages
• Kellyville daily averages
• , (d) same for Kulusuk

60
New Greenland Proposal

• Network of 5 new SGs around coast
• Should confirm gravity variations to lt 1 mgal
• Tied to AG measurements
• Monitored by GRACE

61
New GWR field instrument
(requires no He refills)
62
Larger European GGP Array

• New sites would be useful in Europe to fill gap
  between ME and existing stations
• e.g. Central France, Denmark, Southern Sweden,
  Poland

63
Conclusions

• GGP database will monitor gravity variations for
  satellite missions
• Both SGs and AGs are required to confirm drift
  and offsets at the 1 mgal level
• GPS measurements required to correct for ground
  vertical deflection requires gt 4 years to define
  secular trends
• Atmospheric loading should be done with full 3-D
  modeling, as for GRACE
• More hydrological studies, including soil
  moisture and continental water loading, are
  required.

64
Thanks to all members of the GGP Support
Groups who have dedicated their efforts to
maintaining the gravity stations