Sometimes something that starts out looking like a good idea turns out to be a very bad idea. That certainly seems to be the case with the current U.S. plan to put a land imaging sensor on the first of a new generation of polar-orbiting weather satellites, instead of building another dedicated Landsat spacecraft.

Landsat satellites have been providing scientists with multispectral imagery of the Earth continuously for more than three decades. The data has proved over that time to be extremely valuable for a variety of uses including environmental monitoring, economic studies and national and global security.

Because the data is considered so useful, the Land Remote Sensing Policy Act of 1992 committed the U.S. government to making sure that the delivery of Landsat data is not interrupted.

A policy approved by the White House last year directed the agencies responsible for providing Landsat data to include a Landsat sensor on the first satellite in the planned constellation of satellites for the National Polar-orbiting Environmental Satellite System.

Unfortunately, despite that decision, the prospect of a significant gap in Landsat data is now very real.

In the view of some scientists, the gap already exists because the data gathered by the latest land remote sensing spacecraft, Landsat 7, is flawed permanently by a sensor glitch that makes the data difficult to use for some things and impossible to use for others.

Landsat 7 also is suffering problems with the gyroscopes that make it possible to point the spacecraft and keep it properly oriented in orbit. Landsat 5, which was launched in 1984, is certain to run out of fuel before the end of the decade. Now the growing cost and schedule worries surrounding the NPOESS program have only increased the anxiety of Landsat users and prompted a new look at the viability of launching a gap filler satellite.

One of the technical problems with integrating a Landsat sensor onto the NPOESS platform is concern that many of the other instruments planned for NPOESS will create too much vibration and other interference with the sensitive Landsat instrument, requiring extensive and costly engineering workarounds.

In addition, the cost estimate for completing NPOESS already has grown to $8.1 billion, an increase of $1.2 billion over the previous estimate. Government officials now say the latest cost estimate is even higher. The U.S. Government Accountability Office, at the request of the House Science Committee’s environment, technology and standards subcommittee, is initiating its third NPOESS review since 2003.

While last year’s White House policy directed NASA, NOAA and the U.S. Geological Surve y to transition Landsat measurements to NPOESS, it also directed them to continue to assess options for minimizing any gap in the delivery of Landsat data. It seems clear at this point that the best way to do that is to go with a free flyer. The current policy leaves that option open and the government should pursue it without any further delay.

With the cost of integrating the Landsat imager onto NPOESS Charlie 1 and the schedule for Charlie 1’s launch uncertain, it makes sense to end the uncertainty and authorize the construction of a small, lower cost, dedicated Landsat spacecraft. Too much time has gone by. This decision needs to be made soon.

The best way to avoid a prolonged gap in Landsat data would be to immediately scrap the idea of putting a Landsat sensor on the first NPOESS satellite and instead build a small satellite with a Landsat sensor based on a proven design. The agencies should work to keep the cost of the dedicated gap filler satellite considerably less than the original estimates, but the cost also must be weighed against the growing prospect of a long gap in Landsat data.

The Landsat and NPOESS missions both are critical to the United States and its allies. What the U.S. government must avoid at all costs is creating a mess that ends up compromising and delaying NPOESS and also creating a lengthy gap in the continuity of Landsat data.

Once a gap filler satellite is up and running, the White House Office of Science and Technology Policy needs to decide whether a series of small dedicated Landsat satellites or the use of NPOESS makes the most sense for the long term. The best solution will be one that minimizes costs without creating the possibility of a lengthy gap in Landsat data and one that also does not create unacceptable technical risk for the NPOESS satellites.