surface and lead to a better understanding of geologic 
and tectonic histories of Mars. Topographic mapping 
of Mars, however, differs in many ways from the 
mapping of Earth. It involves solving unprecedented 
problems: the absence of ocean to provide a zero- 
elevation reference surface, the lack of precise ground 
control, methods of data acquisition, or characteristics 
of data, and so forth. These unconventional factors 
require new approaches and the development of new 
mapping techniques. 
As Mars has no seas hence no sea level, the most 
appropriate definition for its topographic datum, the 
zero-elevation surface, is a 6.1-millibar pressure 
surface (Wu, 1981) with a gravity field represented by 
fourth-degree and fourth-order spherical harmonics 
(Jordan and Lorell, 1975, Christensen, 1975.) 
Topographic data used for mars topographic mapping 
include imaging and non-imaging remotely sensed 
data. Imaging data are Mars images taken by the two 
narrow-angle cameras on board of the two Viking 
missions. Despite the extremely narrow field-of-view 
of the Viking Orbiter images, special photogrammetric 
techniques have been developed enabling 
stereomodels to be established for map compilation 
on AS11-AM analytical stereoplotters (Wu et al, 
1982). Using these techniques, a total of 140 contour 
maps of 1:2,000,000-scale with contour interval of 1 
km has been compiled and covers the entire Martian 
surface. 
Non-imaging topographic data include the Ultraviolet 
Spectrometer(UVS), the Infrared Radiometer(IRR), 
and the Infrared Interferometer Spectrometer(IRIS) on 
board spacecraft of Mariner 9, and the S-band radio 
occultation measurements, from both Mariner 9 and 
Viking missions. Earth-based radar observations of 
Mars have been used and also played an important 
role in Mars topographic mapping. 
From 1,157 high-altitude Viking Orbiter images 
together with occultation measurements and Earth- 
based radar data, a planetwide topographic control net 
was established and has produced 4,502 control 
points for the control of systematically mapping of the 
entire Martian surface (Wu and Schafer,1984). 
Mars small-scale topographic maps , i. e., 1:5,000,000, 
1:15,000,000, and 1:25,000,000 were compiled with 
the combination of topographic data extracted from 
the 1:2,000,000-scale contour maps, and the non- 
imaging topographic data including Earth-based radar 
observations of Mars as described above. 
By digitizing 1:2,000,000-scale contour maps, a Mars 
Terrain Model (MDTM) has been derived with a 
resolution of 1/59.226 degree, exactly 1 km per pixel at 
the equator. 
964 
Despite the coarse data from limited data sources, 
these topographic products are results of series of 
Mars projects and are the first comprehensive 
topographic maps. 
2. MARS TOPOGRAPHIC DATA SOURCES FROM 
MARINER 9 AND VIKING MISSIONS 
Topographic data used for Mars topographic mapping 
are derived from remotely sensed data of various 
devices of experiments from Mariner 9 and Viking 
missions. For example, two television cameras on 
board the Mariner 9 spacecraft photographed the 
entire Martian surface (Masursky et al., 1970). But it 
was the two Viking Orbiter missions that sent back 
almost 60,000 images of Mars to Earth, made it 
possible to systematically map Mars' topography in 
great detail. Scientific experiments, which have 
significance to the topographic data, imaging or non- 
imaging, are briefly described in the following two 
sections. 
2.1. Imaging Data 
The primary goal of Mariner 9 was to photographically 
map the planet Mars. More than 7,300 images of Mars 
were taken by the two cameras (wide- and narrow- 
angle cameras). But it is the 60,000 images of Mars, 
which were taken by the Visual Imaging subsystem 
(VIS) on the two Viking Orbiters, to have been used for 
the control network and for stereoscopically compiling 
Mars topography. The VIS on each of the two Viking 
Orbiters consists of two identical vidicon cameras with 
nominal focal lengths of 475 mm. Each image 
consists of 1,056 lines (along y-axis) with 1,182 
samples (along x-axis) in each line, and thus contains 
approximately 1,200,000 picture elements. At a 
12.494 mm x 14.021 mm format, each pixel is 11.8 
micrometers square. The field of view is 1.538 x 1.592 
degrees, or 2.286 degrees diagonally. A resau grid 
consisting of 103 resau marks is incorporated in each 
camera. The resau marks are 36 micrometers, or 3.1 
pixels, on one side. The calibrated coordinates of 
resau marks are known to better than 2 micrometers. 
Geometric distortion are less than 3 pixels at the 
center and increase outward. After the IPL 
decalibration is made to the Viking Orbiter images, the 
r.m.s. errors are reduced drastically to less than 2 um 
(Ruiz,1976). Therefore, optical distortion can be 
ignored for images processed by IPL's dicalibration 
software. At the altitude of 1,500 km, the ground 
coverage of a vertical photograph is about 39.6 x 44.4 
km. Each pixel is 37.5 m square. 
2.2 Non-imaging Data 
Non-imaging topographic data, such as spectroscopic 
sensors and Earth-based radar, can contribute 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996