U.S. Geological Survey --EROS Data Center
Development of a 30 ARC-Second Digital Elevation Model of South America
Norman B. Bliss and Lisa M. Olsen, USGS/EROS Data Center
e-mail: bliss@dg1.cr.usgs.govWork your way through the poster using the imagemap, or use the links below.
Abstract | Introduction | Processing | Special Issues | Evaluation | Applications
A 30 arc second digital elevation model (DEM) of South America was
developed from five data sources:- DTED: Digital Terrain Elevation Data (Defense Mapping Agency)
- DCW: Digital Chart of the World (Defense Mapping Agency)
- IMW: International Map of the World
- AMS: Army Map Service
- PERU: a map of Peru (Government of Peru)
Figure 1 (17K) is an index map showing the primary data source for each area.
A few areas did not have detailed data, and were interpolated from
neighboring areas.
The result is a raster data product. A graphic representation of the
data (figure 6 (50K) is based on shading elevations by color, and blending a
shaded relief view. There are 120 of the 30-arc second data points per
degree of latitude and longitude. The cell size on the ground is
variable, but is roughly 1 kilometer on a side near the equator, and
smaller at high latitudes.
The DTED were generalized from a 3 arc second resolution to a 30 arc second
resolution by systematic sampling. These data were used directly in the
final data set. All of the other data types were processed using Australian
National University's
Digital Elevation Model (ANUDEM) software. This software iteratively
applies a spline interpolation algorithm to the data, resulting in a
gridded surface. The algorithm is able to incorporate a stream network
data set (without elevation values) and use it to maintain hydrologic
consistency in the resulting elevation grid.
Examples of the input data are shown for DTED (figure 2 [17K]), DCW contours,
point elevations, and drainage network (figure 3 [33K]), and other sources
(figure 4 [33K]).
These other sources include IMW contours and points,
AMS contours digitized by Geomatics, Inc., and a river network derived from
both DCW and IMW. A 40-percent sample of DTED was used to constrain
the interpolation of adjacent areas. Where there were conflicts between
data types, we excluded what appeared to be unreliable data (figure 5 [33K]).
The data from figures 2 through 4 were combined by ANUDEM to create a DEM
of the area (figure 6 [50K]). The results were reviewed in a series of small
areas, and then several runs of large areas were made to cover the continent.
The Amazon river system is very large and has very little elevation change
over long distances. Consequently, not much elevation data was available
to control the interpolation process. A technique was developed to
use a very few elevation points along the main stem to interpolate the
river heights. The interpolated river elevations were then used to
constrain the interpolation of the surrounding topography using ANUDEM,
resulting in a more realistic landscape pattern than when the gridding
was done without this constraint.
The gridding algorithm in ANUDEM can create spurious hills or valleys,
especially in areas of little data with very strong relief nearby. In some
cases, such as right next to the Andes mountains in Bolivia, the contour
data for the mountains were removed, ANUDEM was run, and a portion of the
output grid without the spurious hills was then used as input to another
ANUDEM run with all of the data included.
The resultant data set has 10 times more points along each line of
latitude or longitude than the best continental data set previously available
to the public (ETOPO5). This represents a 100-fold increase of resolution
on an areal basis. The accuracy of the grid is limited by the accuracy of
the source materials used to create it. The typical contour interval in
the DCW data is 1000 feet (305 meters). The stated limits of accuracy for
the DCW as defined by the DMA are 2000 meters circular error (horizontal)
and plus-or-minus 650 meters liner error (vertical) at 90-percent
confidence. The accuracy for the result has not been measured or calculated,
although it is probably better than the stated accuracy and within
about 300 meters vertical in most cases.
Ideally, the accuracy of the interpolation should approach one-half the
contour interval of the source data. However, various artifacts are
introduced by the processing. For example, if point data occur in
an area with little other control, a mound is created around the
elevation point. If these points were coded at peaks (as they often are)
then the result may be realistic. However, if the points were in a
relatively level area, then the mounds provide a misleading representation
of the topography, and the elevation of the surrounding area is underestimated.
There is a stair-step effect at contour lines in areas with gradual elevation
change. There is a trade-off between horizontal accuracy in matching the
contours and vertical realism in expected topographic profiles. The
maintenance of horizontal accuracy at the expense of vertical realism was
chosen for this project. Users may want to apply post-processing to create
more realistic profiles from these data.
The data set is expected to be useful for geometric registration of satellite
images of the earth, and for studies in the fields of geology, hydrology,
ecology, and agronomy. A shaded relief map derived from the data (figure 7 [50K])
may be useful as a backdrop for many other types of data, including
socioeconomic patterns.
