The shape of Mars from re-analysis of Mariner 9 and Viking Orbiter occultations [Smith and Zuber, 1996].
The pre-MOLA knowledge of the long wavelength shape of Mars is derived from radio occultation data from the Mariner 9 and Viking Orbiter spacecraft, as well as from Earth-based radar range data. The uncertainty of the occultation measurements is approximately 500 m (1 standard deviation) due to the combined uncertainties in orbital determination (~400 m), atmospheric refraction (~300 m), and timing errors in occultation measurements and planetary ephemeris. The radar range data are also subject to errors in ephemeris, as well as solar plasma effects, for a total uncertainty of ~300 m. Comparison of radii from occultations, which tend to measure topographic highs, and radar returns, which reflect from valley floors, yields a systematic difference of 730 m.
The actual state of knowledge is limited by the sparse sampling of the planet, with only a few hundred independent measurements. The elevation of the highest peak, Olympus Mons, is uncertain by several kilometers. Our knowledge is sufficient to identify the long wavelength topography, notably the offset of the center-of-figure from the center-of-mass of 3561 m, mainly along the z-axis. The south pole is approximately 6 km further from the COM than the north pole. The implications of the shape for the nature of Mars' perplexing hemispheric dichotomy are discussed in Smith and Zuber [1996].
A digital terrain model has been published for Mars by the USGS/Flagstaff, with elevations based on occultations, UV and IR spectral measurements, stereo and radar data. The model is referenced to an areoid, derived from atmospheric pressure estimates, rather than a more accurate geoid. This model has much better spatial resolution at low latitudes than the global shape model, but it has poor long wavelength control. Comparison with MOLA data indicates that errors are on average about 1 km and are non-systematic (i.e. can't be adjusted by a simple correction factor), but in some areas the errors are as large as 5 km.
Smith, D.E., and M.T. Zuber, The shape of Mars and the topographic signature of the hemispheric dichotomy, Science, 271, 184-188, 1996.