NASA officials say the best chance to hear from the napping Spirit rover again will be in September or October, but the timing of the robot's revival from winter hibernation is an engineering guessing game.

Spirit's last panoramic mosaic before entering hibernation shows a mound scientists named "Von Braun" approximately 500 feet from the rover's position. The Columbia Hills serve as a backdrop in this image. This image was provided to Spaceflight Now by Kenneth Kremer. Credit: Marco Di Lorenzo, Kenneth Kremer, NASA/JPL/Cornell See larger image here

Spirit was forced to sleep by the cold winter in the Martian southern hemisphere, where low sun angles were not sufficient to power the rover through solar panels.

The stranded rover last communicated with Earth on March 22. Spirit has been stuck in a sand pit known as Troy since April 2009, leaving the rover tilted away from the sun and limiting its ability to produce electricity.

The winter solstice at Spirit's location was May 13, and conditions should now be improving. But the rover's batteries likely won't be collecting enough sunlight to begin communicating again until September or October.

Spirit's energy production had dipped to 134 watt hours before controllers lost communications March 22.

"While we've passed winter solstice, and the sun is getting a little higher in the sky, the intensity of the sun is still very low," said Doug McCuistion, the director of NASA's Mars exploration program. "So we actually don't think we're going to have enough power to hear from it for another month-and-a-half or two months. The peak probability is going to be in late September or early October."

But that's just a best guess, according to Steve Squyres, the top scientist for the Spirit and Opportunity rovers from Cornell University.

"It depends on the power projections, which are obviously uncertain," Squyres said in an interivew last week. "It depends very much on how much dust is on the solar arrays, and we have no way of monitoring that at the current time. Our best guess is probably like October-ish, but that's got a lot of uncertainty in it."

NASA's large communications antennas are regularly listening for messages from Spirit, just in case the rover wakes up earlier than predicted.

Then there's the concern that Martian dust has accumulated on Spirit's solar arrays, the robot's lifeline to wake up from its winter slumber.

"The solar array panels were pretty dirty," McCuistion said. "If there has not been a cleaning event, and through winter typically you don't get those, the dust build-up on the arrays could be pretty significant. So we don't know when we'll be getting enough power into the arrays to actually get the batteries charged up and get the computers back online."

The dust reduces the efficiency of the craft's fixed solar arrays. Occasional gusts of wind blow dust off the solar panels, giving the rovers a jolt of electricity.

But the fortuitous wind gusts aren't common in winter, and if Spirit's solar panels have collected more dust since March, the rover could face a master clock fault.

This mosaic from the Opportunity rover shows an example of a dust cleaning event in early July, when wind cleared the craft's solar panels. This image was provided to Spaceflight Now by Kenneth Kremer. Credit: Kenneth Kremer, NASA/JPL/Cornell See larger image here

"If you stack worst case on top of worst case, there is one failure mode we could get into, in principle," Squyres said. "We think it's unlikely, but it's possible. It's called a master clock fault. If we have a master clock fault, we probably wouldn't hear from the vehicle until the next time we had one of these cleaning events -- the gusts of wind that clean the solar arrays."

Engineers believe Spirit is now in a low-power fault, in which the craft only powers its master clock to periodically check its power status until there is enough electricity to wake up and radio Earth or an orbiting satellite, according to NASA.

"There are two different levels of faults," Squyres said. "One is low-power fault mode, which we know the brand. We know that we've tripped that, and we think if that's the fault mode that we're in, we will come out of it sometime probably in October, with big error bars. If the power has dipped lower than our projections say, which is possible if there was some big dust event, the next level of fault protection is the master clock fault. If that happens, it gets much harder to predict when we might hear from it again."

McCuistion, NASA's top Mars official, said Spirit's sensitive electronics are being exposed to temperatures they have never seen before, even lower than worst-case testing conducted before the craft launched.

"It's an environment Spirit's never encountered before," McCuistion said. "Some of this is crossing your fingers and some of it is good engineering guesses, but none of it is hard science because we just haven't experienced this before."

If Spirit survives the winter, NASA is planning a series of geophysical science experiments probing the Red Planet's interior, monitoring weather and studying the composition of nearby soil.

Studying the deep interior of Mars has long been a high priority for researchers, according to McCuistion.

Spirit will be used to track tiny wobbles in the rotation of Mars, which could tell scientists whether the planet has a molten or solid core.

NASA gave up on removing Spirit from its sandy trap in January after 4.8 miles of driving across the bed of Gusev crater and into a range of highlands called the Columbia Hills.

McCuistion said the value of Spirit's new science mission, buoyed by its high ranking in researchers' decadal survey reviews, means the rover isn't in danger of being shut down anytime soon.

"The cost is negligible, frankly," McCuistion said in an interview last week. "It's not that expensive to do that kind of science."

Like most missions in NASA's portfolio, the rovers are the subject of an annual review by independent scientists to gauge their research value.

"Getting an asset on the surface of Mars is so hard, that when you have them there doing good science, it's a lot cheaper to keep them going," McCuistion said. "As long as they're doing good science, we're not going to turn them off unless they turn themselves off."