We have no idea yet what discoveries near-future Mars missions will make, and how new discoveries will force revisions in our picture of the planet, and how that changed picture will in turn impact upcoming in-situ precursor missions. Which in turn, will be used to pick the best locations to return samples from. The precursor missions could also help design future in-situ missions that may tell us about the geological and biological processes of Mars at a cheaper cost than sample return missions.

What does make sense is to develop a set of possible alternate mission sequences, depending on the possible discoveries that may be made by those near-future Mars missions. One of the three "scientific subgroups" of MEPAG (the "Pathways" science steering group) has done just that, developing four preliminary possible scenarios for later U.S. Mars missions depending on what the probes through 2005 reveal.

Of course, depending on the real discoveries, other such "exploration pathways" are possible. In this situation -- as General Eisenhower said about war -- any individual advance plan will turn out to be partly inapplicable, but developing a successful planning process is crucial.

All such "pathways" start with the 2009 "Smart Lander" mission, which has now turned into a central linchpin of the Mars program -- it's already been delayed two years, and this reporter wouldn't be surprised to see it delayed for two more, for it's very important to get it right.

It's crucial for two reasons. First, there's the Lander itself, which is intended to test a whole set of new technologies absolutely crucial to any adequate exploration of Mars, and which is also likely to be mass-produced to land a whole series of various big payloads on Mars in the future -- including, probably, all sample return missions for a long time.

First, it's intended to vastly increase the precision of Mars landings, which currently can be confidently targeted only within a big ellipse of about 100 by 20 km -- a fact which has very seriously limited the scientifically interesting sites open to the 2003 MER Mars rovers.

The Smart Lander, by contrast, is scheduled to come down within only 5 km of its target point, by actively changing its tilt during atmospheric entry to steer itself toward its target in accord with last-minute final instructions from Earth.

The precise navigational data needed for those instructions will come both from last-minute radio tracking, and from a new camera which will send back photos of the exact locations of Mars' moons as seen from the spacecraft during approach.

The cancelled 2001 Mars Surveyor Lander would have tested all this system except for the camera, which remains scheduled to be tested on the 2005 Mars orbiter.

Second, it will test a new landing hazard avoidance system, using both a radar system and a scanning laser rangefinder during its final descent to repeatedly build detailed 3-D maps of the terrain below it, and then steering itself as much as 100-200 meters to the side during its final descent on throttled rockets to avoid any dangerously large boulders or steep slopes.

Finally, its landing gear will be radically redesigned so that it can survive a landing on even a meter-tall boulder or a 30-degree slope without harm. This may be done by making the entire lower equipment section of the Lander, including its rocket engines, expendable and crushable, and equipping it with long outriggers to keep it from tipping over.

All these changes will make it possible for such a lander to land a big payload at a scientifically interesting spot virtually anywhere on the planet -- something that even the hard-landing airbag system used by Pathfinder and the MERs is incapable of doing.

The first payload carried by this Lander will be the 600 to 800-kg "Mars Geobiology Explorer", a rover to do the most detailed scientific survey of Mars yet.

It will land at a site indicated by earlier missions where there are accessible sedimentary rock layers, and will use its 70 to 100 kg of complex scientific instruments to obtain detailed information about the climate and water conditions that existed when the rocks were laid down, and thus about the suitability of such an environment for the development of life.

MGE will survey the landscape with an extendable mast carrying cameras, IR spectrometers, and (probably) a new instrument that can fire a small laser beam to create a spark on a rock up to 10 meters away and analyze its spectrum to precisely measure the elements in it.

It can then rove about to any interesting rock or soil target it finds, use a "long arm" carrying various element and mineral spectrometers and magnifying cameras to study it in detail, and -- if it chooses -- use a shorter "strong arm" carrying a soil scoop and a drill to extract small samples from rocks, to collect a sample and insert it into an interior package of instruments to do a still more detailed analysis. It will probably also carry a one-meter drill for subsurface soil sampling.

MGE's instruments are so complex that they will need about six years' head start to be properly developed -- so NASA is planning to issue this mission's "Announcement of Opportunity" for scientists to submit instrument proposals within six months.

It probably will not carry any instruments to try to detect actual living Martian organisms -- but its internal sensors will carry out very detailed chemical analyses of the samples to determine the precise composition of any traces of complex organic compounds found, in order to look for evidence that they are "chemical fossils" of former Martian life or of those prebiotic compounds that served as an intermediate stage in the appearance of life.

These will include a search for unusual ratios of trace isotopes in the elements within such compounds and a check for "homochirality" - the tendency of complex organic molecules that can take two mirror-image shapes to actually exist in only one of those shapes. Either of these, if found, could be a strong indicator that the organic compounds are actually the remains of once-living organisms rather than having nonliving origins.

MGE's instruments will also do very detailed studies of the non-organic minerals in the samples -- including microscopy of rock crystals and soil grains with a one-micron resolution -- to determine their geological nature and the climates and possible liquid-water environments to which they have been exposed during their history. Many of these compounds -- containing nitrogen, sulfur, phosphorus and iron -- are also directly relevant to the question of whether life could have evolved on ancient Mars.

Its microscopes may also be used to try to search directly for fossil microbes. And it will try to determine the exact nature of those powerful oxidant compounds apparently found by the Viking landers, which seem to destroy most organic compounds in Mars' upper soil layer.

Finally, the third and last MEPAG subgroup -- in its report on the best "astrobiological" strategy for the new Mars program -- has recommended that MGE could carry some relatively lightweight sensors to try to measure the extent to which it has accidentally carried Earth germs and their remains to Mars.

MGE was originally supposed to be solar-powered, which would force it to land within 30 degrees of Mars' equator to obtain enough daily sunlight to power its operations -- and which would also limit its lifetime to about 6 to 12 months before enough Mars dust accumulated on the solar panels to choke off their power production.

Like the smaller 2003 rovers, it was also supposed to travel at most 100 meters per day, with its Earth controllers using the photos it took the previous day to pick out a "waypoint" for it to drive toward, and the rover using stereo cameras and other sensors to automatically steer around obstacles to reach that point. After a rock or soil patch worthy of sampling was found, it would then take several days to jockey the rover precisely up to that target.

A study last year concluded that this rover would take about eight months to drive 6 km across the Martian surface, pausing en route at only three sites in which it would spend several weeks poking around in detail for samples. But early this year, the White House announced that it would support providing the rover with two plutonium-fueled RTGs as a nuclear power supply instead.

If Congress concludes that the safety risk from launching such RTGs is acceptable, it would revolutionize the mission, allowing it to function for three years and drive over 20 km. It would also be able to land almost anywhere on the planet -- including the "layered deposits" just beyond the edges of the polar caps, and the small gullies found by MGS which may have been formed by very recent liquid-water flows across the surface - most of which are at latitudes higher than 30 degrees.

Consideration is also being given to improving its onboard autonomous navigation sensors and computers and using the proposed 2009 Mars comsat to allow Earth stations to stay in continuous contact with it for most of each day, which could allow several different waypoints to be picked out each day enabling it to drive up to several hundred meters in that day.

These improvements could also automatically enable the rover to drive up precisely to a rock or soil patch up to 10 meters away without further Earth instructions to position it for sampling. If this technology as well as the RTGs is incorporated into the MGE, it could end up driving for fully 40-50 km across Mars and stopping at dozens of sampling sites during its lifetime. And in fact either nuclear power or autonomous navigation will by itself vastly improve the mission.