The shuttle Discovery has been delayed as it heads home to Florida's Kennedy Space Center from Edwards Air Force Base in California.
Discovery's homecoming has been delayed yet again
The orbiter is riding piggyback on a modified jumbo jet, more than a week after it landed in the Mojave desert.
The pair arrived in Louisana on Friday for an overnight stop, but bad weather has delayed departure until Sunday.
The 3,591km (2,232 mile) trip is expected to cost the US space agency a hefty $1m (£560,000).
Discovery and its seven member crew touched down on 9 August at the Edwards Air Force Base after a 14-day mission to service the International Space Station.
Nasa diverted the landing to California after poor weather prevented the shuttle from returning to Florida.
After landing, Discovery underwent maintenance and crews worked around the clock to prepare the shuttle for departure by purging it of hazardous substances and removing fuel from the on-board tanks.
Technicians attached an aluminium tail cone to the shuttle to reduce drag during the flight, and coupled Discovery to the 747 jet just hours before takeoff.
Discovery's homecoming has been tempered by uncertainties about the future of the shuttle programme.
Nasa has announced that its shuttle fleet will not fly again before March at the earliest.
DISCOVERY FLIGHT TIMELINE
26 July: Lift-off
30 July: First spacewalk to check repair techniques
1 August: Second spacewalk to fix faulty gear
3 August: Makeshift repairs completed
9 August: Touch-down
The fleet was grounded after a large piece of foam was shed from Discovery's external fuel tank during lift-off on 26 July.
A similar problem caused the shuttle Columbia to break up on re-entry to the atmosphere in 2003, killing all seven astronauts on board.
"From an overall standpoint we think really 4 March is the time frame we are looking at," said Bill Gerstenmaier, Nasa's new head of space operations and the official overseeing the foam fix.
"The teams are making very good progress. But we are still not complete by any stretch of the imagination."
Nasa chief Michael Griffin told journalists in Washington that there had been complacency in the past but that there was now a new culture at Nasa.
"For good or ill - and obviously, it was for ill, a poor choice of words on my part - we in Nasa didn't look in detail at foam shedding from the tank for 113 flights - and shame on us," Dr Griffin said.
Technicians attached an aluminium tail cone to the shuttle to reduce drag
Space shuttle Atlantis had been due to blast off in September. But Nasa engineers will now have to make modifications to the shuttle's external fuel tank, particularly to an area known as the Protuberance Air Load (Pal) ramp.
Seven members of an oversight panel said Nasa had not learned key lessons from the Columbia disaster.
Their "minority report" was contained within the final report by the 26-member Return to Flight task group appointed to evaluate how the US space agency meets the recommendations by the Columbia Accident Investigation Board (Caib).
So much emphasis was placed on trying to meet unrealistic launch dates that some safety improvements were skipped, said the seven members.
"We expected that Nasa leadership would set high standards for post-Columbia work... we were, overall, disappointed," the panellists wrote in the report.
The seven critics included a former shuttle astronaut, former undersecretary of the Navy, a former congressional budget office director, former moon rocket engineer, a retired nuclear engineer and two university professors.
Dr Griffin said he was "changing the game" on thinking regarding the shuttle's usage by Nasa ahead of its September 2010 retirement.
It was originally calculated that about 28 further shuttle flights would be needed to complete the International Space Station. That prediction was later reduced to about 15.
Now, Dr Griffin said, Nasa was "not trying to get a specific number of flights out of the shuttle system".
He added: "The United States has a commitment to its partners to complete the station. We believe that, absent of major problems, we...can essentially complete assembly of the station with the shuttle fleet in the time that we have remaining."
DAMAGE TO THE EXTERNAL TANK
The shuttle's two Protuberance Air Load (PAL) ramps act as aerodynamic covers to the various cables and air lines running up the side of the external tank
The PAL ramps are sprayed with insulating foam, like the rest of the tank, to prevent the formation of ice when it is filled with freezing liquid hydrogen fuel
During Discovery's launch a 0.5m long chunk of foam weighing about 450g broke away from one of the PAL ramps. It caused no damage - but a similar incident of foam loss is believed to have caused the Columbia accident in 2003