THE SELECTION OF LANDING SITES ON MARS

The selection of the five successful Mars landing sites in the modern era reflects the state of scientific knowledge at the time and the ability to gather new information during selection and can be used as a prism to view both the growth of our scientific understanding of the red planet and the enormous impact the landed results have had on it. For the Mars Pathfinder lander, little new information existed since the Viking missions or could be acquired, but a greater appreciation of how the Viking landing surfaces related to orbital data and a clear Earth analog were helpful. Mineralogical information was scarce. The Mars Exploration Rover landing site selection benefited tremendously from Mars Global Surveyor (MGS) data and initial data from Mars Odyssey and one of the two sites that were selected (Meridiani Planum) was based on a unique mineralogical identification of hematite in thermal spectra. The selection of the Phoenix landing site in the high northern plains used thermal, gamma ray and neutron spectrometer data from MGS and Odyssey and models to successfully identify locations with thin soil overlying ground ice. High-resolution imaging on Mars Reconnaissance Orbiter (MRO) fairly late in the process enabled the identification of rocks and rock size-frequency distributions from orbit that could be related to lander rock distributions and prompted an anxious search for an acceptably low rock abundance site. Selection of the Mars Science Laboratory (MSL) landing site made extensive use of targeted MRO images and visible and near infrared spectra to identify aqueous minerals and locations that preserved the record of these early environments. Nearly complete medium- and high-resolution stereo images allowed the identification of not only landing hazards, but also plausible traverse routes for MSL into potentially difficult terrain. Even though the fidelity of estimating lander hazards from orbit has progressed tremendously with the increase in orbital imaging capabilities, available data did accurately predict landing hazards and surface characteristics found after landing at all five landing sites. The ability to predict exactly what kinds of geological materials that would be available at the surface at a landing site from remote sensing data has been less uniformly successful.