datum discrepancies incorporated in GTOPO30 on the order of
10 m.
With the incorporation of improved and fully analysed
backscatter waveforms in the SLA-02 data set (Carabajal et al.,
2000), a more rigorous assessment of regional biases can be
performed. The waveforms record the within-footprint height
distribution of backscattered laser energy, characterising surface
relief caused by vegetation, buildings and ground slope and
roughness. Comparison of highest, mean and lowest detected
elevations within SLA-02 footprints to DEM's will reduce bias
effects due to first-return ranging.
The standard deviations in Table 2 and 3 are likely a
consequence of four sources of difference. One is the sampling
issue whereby the SLA point observation is not equivalent to
the representative GTOPO30 grid cell value; as local relief
increases, the sampling difference will cause larger standard
deviations. A second source is the spatially heterogeneous
nature of vegetation and urban cover causing a ‘random’ SLA
error; in some places SLA is measuring canopy or building tops
whereas in other locations bare ground is measured. A third
source is actual random error in the SLA elevation results. For
flat surfaces this error is small, as indicated by the narrow
distribution of residuals with respect to the ocean surface
(Garvin et al., 1998), but as surface slope increases random
error due to pointing uncertainty increases (Harding et al.,
1994). The final source is any random error in the GTOPO30
product.
It is not possible from this analysis to separate these four
contributions to the observed standard deviations of elevation
differences. However, the analysis does show that the raster
based source material (DTED in the regions studied) does have
less error as compared to the sources based on 1:1,000,000
scale contour maps, as expected (Table 1). Separation of the
four sources of elevation difference could be achieved by
examining SLA elevation repeatability in the vicinity of ground-
track cross-overs as a function of local relief, land cover, and
distance between laser footprints and comparing that to SLA to
GTOPO30 differences as a function of local relief and land
cover.
4. CONCLUSION
The flight of SLA has provided the first opportunity to utilise
orbital laser altimeter data in an accuracy assessment of global
DEM's of the Earth. The consistent reference frame, high
absolute accuracy, and ability to range to all types of land
surfaces, regardless of cover or relief conditions, makes orbital
laser altimeter observations well suited for characterisation of
systematic biases in global DEM's. However, sampling
differences between the laser altimeter data and 1 km gridded
DEM's lead to differences in the manner in which topography is
represented and thus contribute to the variation observed in
altimeter to DEM elevation differences.
Through this study methodologies have been developed which
will be applied using Vegetation Canopy Lidar (VCL) and Ice,
International Archives of Photogrammetry and Remote Sensing, Vol. 32, Part 3W14, La Jolla, CA, 9-11 Nov. 1999
Cloud and land Elevation Satellite (ICESat) laser altimeter
profiles to validate the accuracy of a 30 m resolution global
DEM to be produced by the Shuttle Radar Topography Mission
(SRTM). SRTM, scheduled for launch in January 2000, is a
joint project between NIMA, the National Aeronautics and
Space Administration (NASA), the California Institute of
Technology’s Jet Propulsion Laboratory and DLR. VCL, the
first in NASA’s Earth System Science Pathfinder spacecraft
series, is led by the University of Maryland and is scheduled for
launch in September 2000. ICESat, a part of NASA’s EOS
flight program, is scheduled for launch in 2001. Integration of
VCL, ICESat and SRTM topographic data will lead to a global
representation of Earth topography with unprecedented
resolution and documented accuracy that will greatly contribute
to Earth science studies.
ACKNOWLEDGEMENTS
The SLA team consisted of a large number of individuals who
made this pathfinder experiment possible, lead by Jim Garvin,
Project Scientist, and Jack Bufton, Project Engineer. Flight of
the SLA instrument was made possible by the infrastructure and
personnel of the Shuttle Small Payloads Hitchhiker Program.
Funding and hardware for SLA was provided by the NASA
Earth Science Enterprise, the Goddard Director’s Discretionary
Fund, and the ICESat and Mars Observer Laser Altimeter
projects.
The following organisations participated in the GTOPO30
project by contributing funding or source data: the National
Aeronautics and Space Administration (NASA), the United
Nations Environment Programme/Global Resource Information
Database (UNEP/GRID), the U.S. Agency for International
Development (USAID), the Instituto Nacional de Estadistica
Geografica e Informatica (INEGI) of Mexico, the Geographical
Survey Institute (GSI) of Japan, Manaaki Whenua Landcare
Research of New Zealand, and the Scientific Committee on
Antarctic Research (SCAR).
REFERENCES
Bufton, J.L., J.B. Blair, J. Cavanaugh, J.B. Garvin, D. J.
Harding, D. Hopf, K. Kirks, S. Rabine, and N. Walsh, 1995.
Shuttle Laser Altimeter (SLA): a pathfinder for space-based
laser altimetry and lidar, In: Proc. Shuttle Small Payloads
Symposium, NASA Conference Publication CR-3310, pp. 83-
91.
Bufton, J.L., D.H. Harding, and J.B. Garvin, 1999. Shuttle
Laser Altimeter: Mission Results and Pathfinder
Accomplishments, In: Proc. Shuttle Small Payloads
Symposium, NASA Conference Publication CP-1999-209476,
pp. 337-345.
Internatior
Carabajal, C.C., D..
Rowton, and J. J. F
Altimeter Range anc
In: International. A
Sensing, this volume
Garvin, J., J. Buft
Frawley, and D. Rc
Topography from S
Echo-recovery Mea
Chem. Earth, 23(9-1
Gesch, D.B., and
development of glot
Pecora Thirteen, Hi
Perspectives from $
20-22, 1996.
Harding, D.J., LL
Altimetry of Terres
Function of Surface
Trans. Geoscience aı
Hutchinson, M.F.,
elevation and strea
spurious pits. Journ:
Lemoine, F.G., S.C
Pavlis, D.S. Chin, C
Torrence, Y.M. W
Rapp, and T.R. Ol
NASA GSFC and t
(NIMA) Geopotenti