Comparing GEOID99 to the GSD95 Canadian geoid model

1
Comparing GEOID99 to the GSD95 Canadian geoid
model

• Dru A. Smith
• Daniel R. Roman
• National Geodetic Survey, U.S.A.
• Presented at the 2000 Meeting of the CGU
• Banff, Alberta, Canada

2
Topics of Discussion

• Models NGS (USA) and GSD (Canada)
• Comparison and discussion
• Future plans

3
G99SSS NGS gravimetric model

• 2.6 Million gravity points
• 30 arcsecond DEM
• 1 arcsecond DEM for PNW USA
• EGM96 underlying structure
• Faye anomalies (not Helmert)
• 1-D FFT
• Geocentric (ITRF96/GRS-80)

4
New for G99SSS (vs. G96SSS)

• KMS98 altimetry
• Ellipsoidal solution to the Stokes Integral (Fei
  and Sideris)
• Multi-band 2-D FFT TCs
• 3 arcsecond TCs for PNW USA
• 1 arcminute resolution
• North to 58 degrees

5
New for G99SSS Ellipsoidal solution to the
Stokes Integral

• Fei, Z.L. and M.G. Sideris, 2000 A new method
  for computing the ellipsoidal correction for
  Stokess formula, Journal of Geodesy, in review.
• Significant reduction in N/S tilts

6
Ellipsoidal Corrections used in G99SSS (cm)
7
New for G99SSS Multi-band 2-D FFT TCs

• Lowers N/S errors in TCs Not as good as 1-D FFT
• In PNW (3 arcsecond)
• 10 areas (5 x 2)
• 18 x 18 degrees (keep 6 x 6)
• Overlap 2 degrees E/W and N/S
• In CONUS (30 arcsecond)
• 3 areas (1 x 3)
• 164 x 46 deg (keep 76 x 17)
• Overlap 8.5 degrees N/S

8
Basic Statistics of TC Differences using
overlapping 2-D FFT regions
9
GEOID99 NGS hybrid model

• Begin with G99SSS
• 6169 (NAD 83) GPS heights on (NAVD 88) leveled
  benchmarks inside USA
• Collocation and transformations
• Non-geocentric
• NAD 83 lt--gt NAVD 88

10
PICTURE OF 6169 GPSBMS
11
PICTURE OF GEOID99
12
G99BM NGS geocentric hybrid model

• Need a geocentric hybrid model to compare to
  GSD95 (geocentric)
• G99BM geocentric GEOID99
• Transformed NAD 83 into ITRF96

13
G99SSS/G99BM vs GSD95

• Re-grid GSD95 from 5 to 1
• G99BM valid only inside USA

14
(No Transcript)
15
G99SSS - GSD95

• Ave -63 cm /- 33 cm
• Trend 0.15 ppm (72 cm E/W)
• Detrended
• Ave 0 cm /- 19 cm
• PNW, BC, Hudson Bay remain outliers

16
G99BM - GSD95

• Valid in USA only
• Ave -78 cm /- 49 cm
• New issues
• WA/OR tilt
• New England rise

17
Conclusions

• E/W tilt remains, but reduced from 150 cm to 72
  cm
• GPS on BM do not significantly help the NGS/GSD
  disagreement
• Data differences in altimetry, and DEM yield
  systematics that cause tilts
• NGSDEM99 in PNW
• Quebec DEM

18
FUTURE PLANS

• Helmert Anomalies, not Faye
• GPS on BM through all of North America
• Re-adjustment in 2003(?)
• Density anomalies
• 1 arcsecond DEM over all of CONUS (USGS)
• Possibly NA (SRTM)