The Shuttle Radar Topography Mission - Remote Sensing Application - Completely Remote Sensing, GPS, and GPS Tutorial
The Shuttle Radar Topography Mission

A precise overview of the SRTM program is found at the JPL video site. Access through the JPL Video Site, then the pathway Format-->Video -->Search to bring up the list that includes "Mapping the World", July 11, 2002. To start it, once found, click on the blue RealVideo link.

It may come as a surprise to most who've never worked on or utilized topographic data sets or maps to learn that much of the global land surface (as well as parts of ocean floors) has not been mapped at other than coarse (large value) contour intervals. Better maps at closer intervals are very much needed in many enterprises, both civilian and military. It should predictable, then, that topographic mapping from satellites can overcome this deficiency since polar orbiters and Shuttle missions pass over much of the land masses of the World. Experience with radar mapping from the Shuttle led geographers, cartographers, and other specialists to advocate a mission that could systematically map large regional terrains in need of more detailed elevation data for the land. Such a mission, covering both well- and poorly-mapped surfaces, would provide a uniform and coherent data set that would serve as a database suited to many applications.

The response from the U.S. community, through NASA and joined by a German organization, is the Shuttle Radar Topography Mission, which was successfully launched (Mission STS-99, Shuttle Endeavor) on February 11, 2000 and operated for 10 days, gathering data that covers approximately 80% of the land surface. It will take about 2 years to fully reduce all data and provide topographic maps in various formats. C-band and X-band radars (see page 8-1 for wavelengths) operate from one pair of transmission-receiving antennas in the Shuttle payload bay and a matching second pair (receiving only) at the end of a collapsible boom of 60 m length (actual distance between inboard and outboard antennas is 83 m), so that dual signal return simulates the separation needed to provide parallax-like data. Here is a view of the boom (right) center as it was extended during the mission.

Photo taken by an astronaut on board Shuttle mission STS-99 showing the inboard end of the 60 meter boom on which radar units were mounted.

NASA JPL supplied the C-band instrument pair. The data from 159 orbits (at angles to the equator that allow monitoring the Earth between 65° N and 50° S), at a Shuttle altitude nominally at 233 km (145 miles), extend for each pass over a swath width of 225 km (141 miles), so that continuous operation for 10 days provided nearly total coverage (with some duplication) of the land traversed. Under these conditions, the radar image has a ground resolution of 30 x 30 m and a vertical (elevation) resolution of 16 m (absolute) and 10 m (relative).

Processing the data into a stereo-mode which, through the principle of interferometry, can be converted into elevation differences and then restructured as a DEM data set by incorporating bench marks (known elevation points), is summarized in this diagram (somewhat degraded by low Internet resolution):

Schematic illustrating how radar band holograms are converted to an image pair from which phase information is extracted to obtain an interferogram, from which a DEM is derived that can be displayed as a shaded relief map.

In essence, radar holograms from both antennas are generated (and can be converted to images). Then, while still in digital mode, these utilize phase information to construct interferograms (again, displayable as images). After introducing known elevation points to the data set, a digital height model (DEM type) can then be used to contruct a relief or shaded relief image. That can be combined with other data types (e.g., Landsat imagery) or converted to other map forms.

Let us look at a typical C-band image, of an area covering part of Dallas, Texas:

 A C-band image from the SRTM mission covering the Dallas, Texas area.

Next, examine this interferogram that shows the distinctive color banding such a data type presents. The two islands are West Maui and Lanai in the Hawaiian chain. Each cycle of colors (from pink to blue to pink) passes through an equal amount of elevation difference (400 m [1300]) in a manner similar to broad contour lines.

Interferogram from the SRTM data applied to two of the Hawaiian Islands.

One of the practical uses of SRTM data resides in the monitoring of changes of elevation over time. This can be important as a tool for observing the swelling of a volcano with newly emplaced lavas. This causes the volcanic edifice to undergo upward swelling of the surface. This next illustration shows changes in the surface in and around the Three Sisters, volcanic stratocones in the Oregon Cascades near Bend, OR. To construct the change plot, data from several collection times over an extended period are used.

Interferogram of the Three Sisters volcanic region in Oregon, showing changes in elevation of the surface over time.

To illustrate some of the different image types derivable from SRTM data, we will display coverage of a pass across the Kamchatka Peninsula, a volcanic mountain chain in eastern Siberia. The first view is the interferogram whose color fringes suggest relief.

SRTM interferogram of part of the Kamchatka Peninsula, Siberia.

From that, a shaded relief (computer-generated artificial light imposes a pattern of shadow and light) and a shaded relief perspective view are derived:

Shaded relief map of the above Kamchatka Peninsula.
Shaded relief perspective view (colors chosen to roughly simulate natural appearances) of the Kamchatka scene.

When SRTM data are combined with Landsat, these views of the Kamchatka mountains ensue:

The Kamchatka scene showing a small area as portrayed using both a Landsat natural color image and SRTM data to provide the perspective view.
 Same approach as just above, but showing the volcanic range in the Kamchatka Peninsula over a wider area.

Still another SRTM example from the same part of the World is Hokkaido Island in Japan, with its active volcano, Usu.

A special type of stereo image known as an analglyph is made by projecting one of the image pair through red and the other through blue filters to give the resulting superimposed image. To see relief in stereo, for this segment of the Kamchatka peninsula, you would need a pair of glasses made usually of cardboard with the right eye opening covered by red cellophane and the left with blue. A few users of this Tutorial may have one from a 3-D movie experience or can make the glasses in the obvious way.

An analglyph version of part of the Kamchatka Peninsula, made from C-band and L-band images; viewing requires you to have a pair of special (usually cardboard) glasses, with one eye opening cover with blue cellophane and the other with red.

The X-band radar on SRTM is provided and managed by DLR - the Deutches Fernerkundungs-datenzentrum (the German Aerospace Center). Its antenna setup is similar to JPL's C-band system, the difference being that its antenna in the Shuttle Bay is fixed to look straight down, rather than inclined. Thus its swath width is an unchanging 50 km. Data processing to produce elevation maps is essentially the same. Here is an X-band image of part of the Kamchatka Peninsula; the highest point in the mountains is 4755 m (15690 ft):

A shaded relief map made from X-band data obtained by the DLR radar on the SRTM configuration during the February, 2000 STS-99 mission.

For those interested in additional information and updates and more imagery, tie into the JPL and German Aerospace SRTM Web sites.

How much of an improvement in small-scale map quality in those parts of Earth where topographic maps are lacking in details found in, say, coverage of the United States? This next image is convincing: On the left is a colorized map of the Uatuma River Basin near Manaus in eastern Brazil based on the best elevation data available; on the right is a SRTM image processed to display the optimum topographic representation extractable from the radar data. The black areas in the map on the right depict the water buildup in Lake Balbina created by damming the river.

Left: Color-coded map showing elevation data in the lowlands of the Amazon River Basin, using surveys prior to SRTM; Right: The same area with elevations reconstructed from SRTM data.

Another example is this pair of maps of Coos Bay in Oregon. The left map is taken from the USGS's 10 meter series covering the U.S.; the right map shows the additional contour lines, at a 5 meter interval, that could be drawn with SRTM data. The coloring in both is an assessment of the potential for landslides:

Two maps of Coos Bay, OR.

By mid-2003, the SRTM task force had completed a topographic map of the entire land surface between 60°N and 60°S over the world. This results from the efforts involved in the SRTM30 project. Although more accurate maps exist for some parts of this surface, this global map represents the best and most uniform quality map of the continental elevations now extant.

SRTM map of land surfaces across the globe.

Dark green denotes the lowest regional elevations. Brown grading into white defines the highest elevations - in southern Asia and the West Coasts of North and South America.

To gain an insight into the improvements resulting from SRTM30, consider these two maps of the Guiana Highlands in the Guyana Shield of northern South America. The map on the left below was compiled from the best published data prior to 2000; its surface resolution is 986 meters. On the right is the SRTM30 version, at 30 m ground resolution and 10 m vertical resolution. This latter is far more accurate; note how the higher elevations shown in the left map are greatly diminished (i.e., were initially exaggerated) as the truer configuration from SRTM replaces the older version.

Left map: composite of elevations from varied data sources; Right map: more accurate version developed from SRTM data.

Follow-up programs are "in the works" to use radar interferometry and laser profiling to continually upgrade topographic/elevation mapping of the entire world.

Suffice to close this Section is this direct observation: Our world has always been a 3-D one; now a variety of space imaging systems that can produce end products in three dimensions will benefit cartographers and others who generate many kinds of maps that depict the true nature of the land surface.