It took a long time, probably too long to avoid a break in the 30-year record of global land-observation data, but the White House has finally gotten it right with its latest Landsat strategy.

By directing Dec. 23 that NASA and the U.S. Geological Survey build and launch a dedicated Landsat-type satellite, White House Office of Science and Technology Policy Director John Marburger has taken the necessary step to minimize the data gap that is likely to result from several years of indecision and ill-advised tinkering with this program. As the old saying goes, better late than never.

Assuming all goes well, the nation’s newest Landsat satellite should be in orbit by 2009, perhaps earlier. That will be a decade removed from the launch of Landsat 7, which was designed to operate five to seven years and has been returning degraded data since its main sensor malfunctioned in mid-2003 .

With the venerable Landsat 5 satellite on its last legs — after a remarkable 20-year run — and Landsat 7 having gyroscope issues in addition to its sensor problem, the decision to go with a free-flying follow-on satellite was long overdue.

It follows a succession of missteps in the planning for long-term Landsat data continuity, particularly the adoption of a highly risky proposal to add land-imaging sensors to the next generation of U.S. polar-orbiting weather satellites.

NASA and the U.S. Geological Survey, who respectively will procure and operate the new satellite, should prepare now to move out once the president’s 2007 budget proposal is unveiled with a competitive procurement designed to get the satellite and sensor under contract as quickly as possible.

For the sake of dispatch, it makes sense to go through NASA’s Rapid Spacecraft Development Office, which maintains a catalog of pre-approved satellite platforms, and limit the competition to three or four companies with approved designs already in the catalog. NASA also should be very wary of attempting to improve on the capabilities of Landsat 7 — prior to the sensor glitch — lest the program run into technical problems that cause delays and widen the likely data gap.

Finally, the agencies should put in a strong effort in response to the Mr. Marburger’s direction to come with a realistic long-term plan for Landsat data continuity. Options such as a government-industry partnership and international cooperation are still worth exploring, even though neither has proved feasible in the near term.

With that in mind, the agencies would be wise to include options in the next Landsat contract for additional satellites, just in case the search for an alternative solution once again hits a dead end.