Elevation change of the southern Greenland ice sheet

Science

Volumn 279, Issue 5359, 1998, Pages 2086-2088

  Davis, Curt H ^a   Kluever, Craig A ^a   Haines, Bruce J ^b  

^a UNIVERSITY OF MISSOURI KANSAS CITY   (United States)

^b CALIFORNIA INSTITUTE OF TECHNOLOGY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

ICE;

ELEVATION; ICE SHEET;

ARTICLE; CLIMATE; GEOLOGY; GREENLAND; PRIORITY JOURNAL;

GREENLAND;

EID: 0032571564     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.279.5359.2086     Document Type: Article

References (38)

1. B. C. Douglas, Rev. Geophys. (Suppl.) 33, 1425 (1995).
2. E. Barron, USGCRP Report 95-02 (U.S. Global Change Research Program, Washington, DC, 1995).
3. H. J. Zwally, A. C. Brenner, J. A. Major, R. A. Bindschadler, J. G. Marsh, Science 246, 1587 (1989).
4. H. J. Zwally, ibid., p. 1589.
5. B. C. Douglas, R. E. Cheney, L. Miller, R. W. Agreen, ibid. 248, 288 (1990).
6. H. J. Zwally, A. C. Brenner, J. A. Major, R. A. Bindschadler, J. G. Marsh, ibid., p. 288.
7. C. H. Davis, IEEE Trans. Geosci. Remote Sens. 33, 1108 (1995).
8. C. J. Van der Veen, Clim. Change 23, 383 (1993).
9. T. V. Martin, H. J. Zwally, A. C. Brenner, R. A. Bindschadler, J. Geophys. Res. 88, 1608 (1983).
10. For example, H. J. Zwally, A. C. Brenner, J. DiMarzio, T. Seiss, Proceedings of Second ERS-1 Symposium ESA SP-367, Hamburg, Germany, 11 to 14 October 1993 (European Space Agency, 1994), p. 159; see also (22).
11. C. H. Davis, IEEE Trans. Geosci. Remote Sens. 35, 974 (1997).
12. J. G. Marsh and R. G. Williamson, J. Astronaut. Sci. 87, 3207 (1982); S. Smith III, G. West, C. Malyevac, Johns Hopkins APL Tech. Dig. 8, 198 (1987).
13. J. G. Marsh and R. G. Williamson, J. Astronaut. Sci. 87, 3207 (1982); S. Smith III, G. West, C. Malyevac, Johns Hopkins APL Tech. Dig. 8, 198 (1987).
14. B. D. Tapley et al., J. Geophys. Res. 101, 28209 (1996).
15. The Geosat JGM-3 orbit solutions show subdecimeter accuracy in colinear ocean altimeter data [F. G. Lemoine et al., Eos (Spring Suppl.) 78, 103 (1997)] and comparisons of monthly averaged tide gauge and altimeter sea level differ at the level of a few centimeters (http://ibis.grdl.noaa.goV/SAT/gdrs/ geosat_handbook/docs/chap_1.htm).
16. For example, H. J. Zwally et al., NASA Ref. Publ. 1233 (NASA, Washington, DC, 1990), vol. 1; C. S. Lingle et al., Ann. Glaciol. 20, 26 (1994).
17. For example, H. J. Zwally et al., NASA Ref. Publ. 1233 (NASA, Washington, DC, 1990), vol. 1; C. S. Lingle et al., Ann. Glaciol. 20, 26 (1994).
18. D. B. Chelton and M. G. Schlax, J. Geophys. Res. 98, 12579 (1993).
19. Ocean Altimeter Pathfinder home page, http:// neptune.gsfc.nasa.gov/ocean.html
20. Residuals were formed at the intersection (crossover points) between the satellite ground tracks and the reference network. The reference network was created by averaging the first 2 years of colinear ERM sea heights after correcting for orbit errors. Only the first 2 years of ERM data were used, because the accuracy of the orbit solutions were severely degraded by solar activity beginning in the third year.
21. The radial orbit errors, E(t), within each orbit solution (typically 6 days) were parameterized by E(t) = A(t)cosΩt + B(t)sinΩt + C, where Ω = 2π/orbital period. In the stochastic filter [S. M. Lichten, Manuscripta Geodaetica 15, 159 (1990)], the 1/rev orbit-error amplitudes, A(t) and B(t), were treated as exponentially correlated noise processes with a time constant of 6 days [C. A. Kluever, B. J. Haines, C. H. Davis, Y. T. Yoon, Eos (Spring Suppl.) 78, 103 (1997)]. The bias parameter, C, was held constant over each 6-day Seasat orbit solution and was used to account for a global-scale difference between Seasat and Geosat sea heights because of measurement-system biases.
22. The radial orbit errors, E(t), within each orbit solution (typically 6 days) were parameterized by E(t) = A(t)cosΩt + B(t)sinΩt + C, where Ω = 2π/orbital period. In the stochastic filter [S. M. Lichten, Manuscripta Geodaetica 15, 159 (1990)], the 1/rev orbit-error amplitudes, A(t) and B(t), were treated as exponentially correlated noise processes with a time constant of 6 days [C. A. Kluever, B. J. Haines, C. H. Davis, Y. T. Yoon, Eos (Spring Suppl.) 78, 103 (1997)]. The bias parameter, C, was held constant over each 6-day Seasat orbit solution and was used to account for a global-scale difference between Seasat and Geosat sea heights because of measurement-system biases.
23. There were 17 separate 6-day Seasat orbit solutions used in our study. The mean and SD of the Seasat bias coefficients were 27 ± 4 cm.
24. Other investigations have indicated that the Seasat altimeter yielded lower sea-surface heights relative to Geosat [B. Haines, thesis, University of Colorado, Boulder (1991); C. Wagner and R. Cheney, J. Geophys. Res. 97, 15607 (1992)]. A recent independent study by G. Kruizinga [thesis, University of Texas, Austin (1997)] reported a Seasat-Geosat relative bias of 22 ± 4 cm.
25. Other investigations have indicated that the Seasat altimeter yielded lower sea-surface heights relative to Geosat [B. Haines, thesis, University of Colorado, Boulder (1991); C. Wagner and R. Cheney, J. Geophys. Res. 97, 15607 (1992)]. A recent independent study by G. Kruizinga [thesis, University of Texas, Austin (1997)] reported a Seasat-Geosat relative bias of 22 ± 4 cm.
26. Other investigations have indicated that the Seasat altimeter yielded lower sea-surface heights relative to Geosat [B. Haines, thesis, University of Colorado, Boulder (1991); C. Wagner and R. Cheney, J. Geophys. Res. 97, 15607 (1992)]. A recent independent study by G. Kruizinga [thesis, University of Texas, Austin (1997)] reported a Seasat-Geosat relative bias of 22 ± 4 cm.
27. Ice-Sheet Altimeter Pathfinder home page, http:// crevasse.stx.com/ia_home/
28. 2 = 24.7 cm.
29. A)(1 - cosα), OH are the orbital heights, and α is the regional ice-sheet slope. The regional slope was determined using a 10 km by 10 km grid computed from a digital elevation model of Greenland produced by S. Ekholm at Kort-og Matrikelstyrelsen, Denmark. The mean and SD of all corrections applied to 35,600 ERM × Seasat dH values was -1.4 ± 4.6 cm.
30. The ascending-descending (A-D) orbit bias is accounted for by using the method described by Zwally et al. (3). Crossover differences greater than 3 SD of the primary Gaussian distribution were discarded to eliminate data outliers due to irregular ice-sheet topography; see note 8 in (3).
31. An average dH and dt value was computed for each cell before computing the spatial average of all cells. For a cell to be used, a minimum of 10 crossovers had to be present, with at least five from Geosat-Seasat (A-D) and five from Seasat-Geosat (A-D).
32. J. F. Bolzan, Eos (Fall Suppl.) 73, 203 (1992).
33. W. Krabill, R. Thomas, K. Jezek, K. Kuivinen, S. Manizade, Geophys. Res. Lett. 22, 2341 (1995).
34. A correction for vertical velocity of the ice base because of crustal deformation was subtracted from the dH/dt value of each 50 km by 50 km cell, and the spatial average was recalculated (see http:// cfageod4.harvard.edu/calc_def.html for details of the vertical velocity calculations).
35. The relative measurement system bias (20, 21) is estimated from the global ocean data and as such is not perfectly separable from the sea-state bias (SSB) resulting from the noncoincidence of electromagnetic and geometric sea level as well as tracker and skewness biases. Although the SSB model for the Geosat data has been rigorously estimated [P. Gaspar, F. Ogor, and M. Hamdaoui, http://neptune.gsfc.nasa.gov/∼krachlin/ opf/algorithms/ssb_geosat.html], the corresponding model for the Seasat data is still under investigation. Older models [for example, G. H. Born, M. A. Richards, G. W. Rosborough, J. Geophys. Res. 87, 3221 (1982)] may not be appropriate because the Seasat data have been retracked as part of the Pathfinder effort (17). We estimated a global SSB for Seasat as a percentage of the significant wave height (SWH). Conservatively assuming our estimate of 0.044 × SWH is accurate to within 40%, the relative measurement system bias would be accurate to ∼5 cm (global SWH averages 2.7 m). A 5-cm error in the relative bias translates to 0.5 cm/year in terms of the overall ice-sheet growth rate.
36. The relative measurement system bias (20, 21) is estimated from the global ocean data and as such is not perfectly separable from the sea-state bias (SSB) resulting from the noncoincidence of electromagnetic and geometric sea level as well as tracker and skewness biases. Although the SSB model for the Geosat data has been rigorously estimated [P. Gaspar, F. Ogor, and M. Hamdaoui, http://neptune.gsfc.nasa.gov/∼krachlin/ opf/algorithms/ssb_geosat.html], the corresponding model for the Seasat data is still under investigation. Older models [for example, G. H. Born, M. A. Richards, G. W. Rosborough, J. Geophys. Res. 87, 3221 (1982)] may not be appropriate because the Seasat data have been retracked as part of the Pathfinder effort (17). We estimated a global SSB for Seasat as a percentage of the significant wave height (SWH). Conservatively assuming our estimate of 0.044 × SWH is accurate to within 40%, the relative measurement system bias would be accurate to ∼5 cm (global SWH averages 2.7 m). A 5-cm error in the relative bias translates to 0.5 cm/year in terms of the overall ice-sheet growth rate.
37. The seasonal signal cycle is approximately sinusoidal and is observed in monthly variations in the dH values over the 2-year period. Interannual variations were computed by subdividing each year of data into four 3-month seasonal periods and performing a same-season crossover analysis (for example, Fall 1988 × Fall 1987).
38. We thank R. Thomas for suggesting the satellite height correction (24). We thank the NASA Pathfinder Program for the ocean and ice sheet altimeter datasets; in particular, we acknowledge the help of H. J. Zwally, C. Koblinsky, A. Brenner, J. DiMarzio, and B. Beckley for their assistance in obtaining these data. We thank C. Perez and Y. Yoon for processing the data for this study. Supported by NASA's Polar Program and the Office of Mission to Planet Earth under grants NAGW-5010 and NAGW-5243. A portion of this work was conducted by the Jet Propulsion Laboratory, California Institute of Technology, under contract with NASA on RTOP #622-83-1740.