Location and sampling of aqueous and hydrothermal deposits in martian impact craters

Do large craters on Mars represent sites that contain aqueous and hydrothermal deposits that provide clues to astrobiological processes? Are these materials available for sampling in large craters? Several lines of evidence strongly support the exploration of large impact craters to study deposits important for astrobiology. The great depth of impact craters, up to several kilometers relative to the surrounding terrain, can allow the breaching of local aquifers, providing a source of water for lakes and hydrothermal systems. Craters can also be filled with water from outflow channels and valley networks to form large lakes with accompanying sedimentation. Impact melt and uplifted basement heat sources in craters > 50 km in diameter should be sufficient to drive substantial hydrothermal activity and keep crater lakes from freezing for thousands of years, even under cold climatic conditions. Fluid flow in hydrothermal systems is focused at the edges of large planar impact melt sheets, suggesting that the edge of the melt sheets will have experienced substantial hydrothermal alteration and mineral deposition. Hydrothermal deposits, fine-grained lacustrine sediments, and playa evaporite deposits may preserve evidence for biogeochemical processes that occurred in the aquifers and craters. Therefore, large craters may represent giant Petri dishes for culturing preexisting life on Mars and promoting biogeochemical processes. Landing sites must be identified in craters where access to the buried lacustrine sediments and impact melt deposits is provided by processes such as erosion from outflow channels, faulting, aeolian erosion, or excavation by later superimposed cratering events. Very recent gully formation and small impacts within craters may allow surface sampling of organic materials exposed only recently to the harsh oxidizing surface environment.     

Determination of Measurement Limits to Optimize Production Losses

The dollar consequence functions associated with consumer's and producer's risks are developed and optimized through use of computer technology. The bivariate normal distribution is used to compute the probabilities resulting from establishment of test limits inside or outside of specffication limits after the manner of Grubbs and Coon [1] and Eagle [2]. Methods of application and precautions during use are described. Practical limits on parameters are discussed, but these are shown to exert no mathematical restrictions on the solutions presented.