The loss of the Mars Climate Orbiter, presumed to have broken up in the planet's atmosphere on 23 September, is acutely embarrassing for NASA, coming as the agency is battling to stave off budget cuts.

Worse, some researchers suggest that the error is a result of NASA's policy of launching "faster, better, cheaper" missions.

The Mars Climate Orbiter was supposed to enter an orbit that would have brought it no closer than 155 kilometres from the surface, after a course-altering rocket burn on 15 September.

The burn went according to plan. Yet the craft descended to within 57 kilometres of the planet's surface, where it could not withstand the friction caused by the Martian atmosphere.

This is NASA's second Mars probe to be lost on arrival, and joins a list of failed missions in the 1980s and 1990s (see Table).

Having ruled out a malfunction on the probe, NASA officials say that the problem must have been with the calculation of its trajectory.

This was plotted by feeding tracking data from NASA's Deep Space Network antennas into a computer program that takes account of the changing positions of the Earth and Mars--information that was generated by another software package, called the ephemeris model.

When navigators at NASA's Jet Propulsion Laboratory (JPL) in Pasadena calculate a trajectory they also make an error estimate, which might be 10 kilometres either side of the projected path.

But a 100-kilometre error is unprecedented--and points to either a software bug, or someone entering the wrong data. "It's unbelievable that they made this mistake," says Francis Rocard, an astrophysicist with the French space agency in Paris.

Troubleshooters are scrambling to pin down the source of the error before December, when another craft must rely on the same software. "We need to understand because Mars Polar Lander is two months away from Mars," says Phil Knocke, a JPL mission controller.

NASA officials are putting a brave face on the loss. "It is not devastating to the Mars Surveyor programme as a whole," claims Carl Pilcher, science director for Solar System exploration at the agency's headquarters in Washington DC.

But in addition to studying the Martian atmosphere and climate, the missing craft was supposed to relay data from the Mars Polar Lander. This will study surface weather and probe the soil near the planet's south pole for three months.

Now the data must be sent directly to Earth--in which case half will be lost--or be relayed back via Mars Global Surveyor, forcing that craft to cut its own observations short.

NASA's previous Mars mishap, the loss of the $1 billion Mars Observer in 1993, was instrumental in causing the switch to more frequent, cheaper missions. But some scientists wonder if the pendulum has swung too far.

"We've been saying all along they were going to lose one of these things," says one. "With 'faster, better, cheaper' you work your people to death," agrees Peter Smith of the University of Arizona, principal investigator for Mars Polar Lander.

This article will appear in the October 2 issue of New Scientist New Scientist. Copyright 1999 - All rights reserved. The material on this page is provided by New Scientist and may not be published, broadcast, rewritten or redistributed without written authorization from New Scientist.