N ASA Administrator Mike Griffin’s unprecedented decision to launch Space Shuttle Discovery July 1 despite red flags raised by his chief engineering and safety officials is bound to be controversial in some circles. After all, astronaut safety is always a highly sensitive issue, never more so than in the wake of accidents like the February 2003 Columbia disaster.

But the publicly available information in this case suggests that Mr. Griffin is not taking any undue risks with astronaut safety and that, when viewed in the broader context of NASA’s circumstances, he made the proper call. Most of NASA’s senior space shuttle and space station officials agreed, and even the two dissenters did not object strongly enough to appeal Mr. Griffin’s decision.

At issue was whether potential shedding of insulation foam from 37 areas of Discovery’s external tank called ice-frost ramps posed an unacceptable risk to the mission. Bryan O’Connor, NASA director of safety and mission assurance, and Chris Scolese, NASA chief engineer, believed it did, and at Discovery’s June 16-17 flight readiness review they recommended against proceeding with the launch. Their so-called no-go recommendation kicked the final decision up to the administrator, who normally does not participate in flight readiness reviews.

Mr. Griffin elected to press ahead based on the assessment of everyone involved that the threat was to the re-entry phase of the mission. The worst-case scenario was similar to that of the Columbia accident, in which foam shedding during liftoff damaged the orbiter’s heat shielding, rendering it unable to withstand the extreme temperatures encountered during atmospheric re-entry.

NASA now has the tools to inspect and evaluate damage to the shuttles on orbit, and unlike the Columbia mission, Discovery’s flight is to the international space station. Mr. Griffin judged that if an on-orbit inspection revealed irreparable damage to Discovery’s hull, the crew could simply remain aboard the space station until rescued.

On Discovery’s June 17 certification of flight readiness, the document that all relevant NASA officials must sign to clear the shuttle for launch, Messrs. O’Connor and Scolese said their recommendation not to fly was based on the risk to the orbiter, not the crew. During a press conference June 21, they reiterated that point and also said that in their view the risk is at the lower end of the unacceptable spectrum.

It is for these reasons that they concurred with Mr. Griffin’s decision to go ahead with the flight. They also said during the press conference that their no-go recommendation was at least in part procedural — a way to formally convey to the NASA administrator the point that the ice-frost ramp issue must not be ignored with some 17 shuttle flights remaining.

This is not to say that Mr. Griffin’s decision was an easy one. If the worst-case scenario does come to pass, sending up a second shuttle — one with the same type of ice-frost ramps — to bring Discovery’s crew home from the space station is no trivial matter. Neither is ditching Discovery, a multibillion-dollar asset that will come down in pieces large enough to cause serious damage if not properly guided into an uninhabited area. It would be the end of the shuttle program, and in all likelihood, to Mr. Griffin’s NASA career.

But the risk that a chunk of foam falling from the ice-frost ramp will actually cause irreparable damage to the orbiter during liftoff remains relatively low. The unacceptable designation means it is deemed likely to result in the loss of an orbiter over the course of 100 missions. As Mr. Griffin noted during a press conference after Discovery’s flight readiness review, “I have a great deal of trouble believing that a statistically sound statement would be to say that this is a probable event to be seen over the next 16 flights.”

Under different circumstances, Mr. Griffin might have elected to stand down and wait until a solution that reduces the risk associated with the ice-frost ramps can be implemented, something agency officials estimate will take about six months. That should still be a top priority, even though it is clear that NASA still does not understand the external-tank foam-shedding problem well enough to come up with a foolproof solution. But if NASA is to have any hope of completing the space station and retiring the shuttle fleet by 2010, it does not have the luxury of time.

The space shuttle is a cantankerous, high-maintenance machine whose time demands can only be curbed by NASA’s willingness to accept risk. The known risks of spaceflight can never be eliminated entirely, and given its enormous complexity, the shuttle holds myriad hazards, known and unknown. It is for this reason, plus the fact that the orbiter’s abilities as an exploration vehicle are severely limited, that the fleet is being retired.

In approving the Discovery launch, Mr. Griffin has drawn a clear distinction in terms of risk tolerance between hardware and people. And he is willing to accept the possibility, however remote, that the orbiter will be lost. Under current circumstances, that is a reasonable standard if meaningful exploration and not indulgence of the space shuttle is to be the purpose of America’s civil space program.