For NASA’s Mars 2020 rover, now being assembled at the agency’s Jet Propulsion Laboratory for launch in July 2020, the stakes could not be much higher. The $2.4 billion nuclear-powered rover is the most complex piece of machinery to ever make a ballistic beeline for the Red Planet. And after it lands in February 2021, its completion of one high-profile objective—collecting rock samples for eventual transport back to Earth—would ensure the rover sets the course for future Mars exploration for decades to come. But first, mission planners have to decide where on Mars this ambitious machine should actually go.

Last week, at a conference just north of Los Angeles, an international group of more than 150 scientists came together for two and a half days of deliberation on exactly that—their debates at times seeming less like a a careful scientific conversation and more a geological and astrobiological food fight. This contentious meeting is the fourth and final such workshop to be held since 2014, and considered four potentially life-friendly candidates for Mars 2020’s coveted landing site:

  • Columbia Hills/Gusev Crater, a site previously surveyed by NASA’s now-dead Spirit rover. It is flush with hot-spring deposits that, on Earth, would offer a comfy home for microbial life.
  • Jezero Crater, a site that hosts a dried-up lake and river deltas that could have sheltered—and subsequently preserved—a wealth of microbial Martians.
  • Northeast Syrtis Major, a giant shield volcano abutting a huge impact crater bearing some of the planet’s most ancient rocks, laid down when Mars was a warmer, wetter place.
  • Midway, another destination strewn with ancient rocks, appropriately named as it lies between landing sites at Jezero and at Northeast Syrtis.

“It is mind-boggling, the amount of information we have available to study another planet,” says Michael Meyer, lead scientist for NASA’s Mars Exploration Program. The workshop differentiated between potential landing sites “based on what we know, what we don’t know, and what we want to know,” he told Scientific American, “but we only get one shot.”

Matthew Golombek, a geologist who co-chairs the Mars 2020 Landing Site Steering Committee, says the four finalist landing-site candidates are now the most carefully studied locations on all of Mars. Each has been saturated with extensive remote observations by orbiters circling the planet far overhead. These have given researchers torrents of new information about geological history, rock composition, surface conditions and more. Mulling over all that data, the workshop’s participants ultimately had to vote for their preferred landing site—the place they believed offered Mars 2020 the best chances of success.

By the workshop’s conclusion, the combined tallies suggested a consensus preference for a hybrid approach—one in which the Mars 2020 rover would visit and explore the dry lakebeds and deltas of Jezero Crater as well as the ancient rocks of the Midway site, which is only some 28 kilometers away. That’s not too far, as the crow flies, but still a potentially tall order for a robotic rover trundling across uneven alien terrain. Making the trek would be a stretch goal for the rover, as the traverse could easily require time in excess of its 2.35-year primary mission. Also, even though this two-for-one approach is scientifically compelling, it is not set in stone: The decision of where exactly to send Mars 2020 rests with NASA’s top scientist, Thomas Zurbuchen, who is expected to review the workshop’s findings and announce his choice by year’s end. His review will then go up the NASA leadership chain for a final announcement.

The preference for two sites was influenced by how rich the extended mission could be, says astrobiologist and workshop participant David Des Marais of NASA’s Ames Research Center. “That really made a difference, much more so than in previous missions.” In particular, Mars 2020 will take advantage of several new technologies—such as better terrain mapping and a more precise and responsive triggering of its parachute during descent—that should allow the rover to achieve a pinpoint landing surpassing those performed by all its robotic predecessors. “Those technologies have opened the door to sites that people would have loved to have gone to earlier but couldn’t,” Des Marais says.

Still to be sorted out is when and how Mars 2020’s samples would be retrieved, as well as the order in which the rover would reconnoiter the workshop’s favored exploration sites. Like many of the workshop’s participants, Des Marais personally favors Mars 2020 landing first at Jezero, using some fraction of its 42 pencil-sized sample tubes to collect rocks there, then trekking to the Midway site for a repeat performance. “The workshop scoring reflected a pretty symmetrical opinion about where do you start and end up,” Des Marais adds, further noting that the Northeast Syrtis landing site also had a strong showing in the votes.

Jezero Crater’s lead advocate at the workshop was John Mustard, a planetary scientist at Brown University in Providence, Rhode Island. “Jezero scored highest in primary rover science, and Northeast Syrtis higher in value of the returned samples. The real winner is the extended mission from Midway to Jezero, which had broad and deep science-value support,” he says.

Regarding Jezero and Midway, it is the possibility of getting twice as much science accomplished with just one rover that is so supremely alluring, Mustard says: “It’s the best of both worlds. If we can sample both sites then we’re farther ahead and able to explore a more diverse range of Mars science than we would otherwise.”

In some important ways the Mars 2020 mission will involve a repeat of the August 2012 landing of NASA’s Curiosity rover. That means the mission’s scientists and an eager audience back on Earth must breathlessly endure a rerun of Curiosity’s so-called “seven minutes of terror” entry, descent and landing on Mars in 2012. For the finale of this nail-biting process, a robotic sky crane hovers on rockets and then lowers the rover on a tether to the surface. Curiosity’s sky crane worked flawlessly; whether the approach will do so again in 2021 remains to be seen.

Despite these operational uncertainties, the mission’s scientists are optimistic. The rover’s overall assembly and testing is progressing as planned at JPL, with no showstoppers in sight, says Matt Wallace, Mars 2020’s deputy project manager.

Even if, against all hopes, the rover’s high-tension landing on Mars goes awry, “the Mars program will go forward,” says Ray Arvidson of Washington University in St. Louis, a veteran of multiple missions to the Red Planet. “[Mars] is NASA’s long-term objective. Whether or not Mars 2020 works, there will be continued exploration.”