Nasa announces technical failure shortly after blast-off
Nasa's first mission to measure carbon dioxide (CO2) from space has failed following a rocket malfunction.
Officials said the fairing - the part of the rocket which covers the satellite on top of the launcher - had failed to separate properly.
Officials said the satellite had now crashed in Antarctica.
The Orbiting Carbon Observatory (OCO) was intended to help pinpoint the key locations on our planet's surface where the gas is being emitted and absorbed.
Nasa officials confirmed the launch had failed at a press conference held at 1300 GMT.
The $270m mission was launched on a Taurus XL - the smallest ground-launched rocket currently in use by the US space agency.
This type of rocket has flown eight times, with two failures including this launch. But this is the first time Nasa has used the Taurus XL.
Nasa will now put together an investigation board to determine the root cause of the problem.
Onlookers watched the launcher soar into the sky from the Vandenberg Air Force Base in California at 0951 GMT on Tuesday.
The first indication of a problem came in an announcement made by the Nasa launch commentator, George Diller.
"This is Taurus launch control. We have declared a launch contingency, meaning that we did not have a successful launch tonight," he said.
"The OCO spacecraft did not achieve orbit successfully in a way that we could have a mission. They're still looking at the telemetry data here very carefully. It appears that we were getting indications that the fairing was having problems separating.
"It either did not separate or did not separate in the way that it should, but at any rate we're still trying to evaluate exactly what the status of the spacecraft is at this point."
Separation of the fairing was one of the last technical hurdles faced by the satellite as it flew into orbit. Orbital said there had been no changes to the design of the fairing since previous launches.
John Brunschwyler, from Orbital Sciences Corporation, the rocket's manufacturer, cast doubt on any suggestion of a link between the failure and a power glitch which occurred to the vehicle before launch.
"That was on a separate system, so I do not believe there was any connection," Mr Brunschwyler told journalists at the Nasa press conference.
Dr Paul Palmer, a scientist from the University of Edinburgh, who was collaborating on the mission, told BBC News: "I am bitterly disappointed about the loss of OCO. My thoughts go out to the science team that have dedicated the past seven years to building and testing the instrument."
Scientists had hoped the OCO mission would improve models of the Earth's climate and help researchers determine where the greenhouse gas is coming from and how much is being absorbed by forests and oceans.
Only about 50% of carbon emitted into the atmosphere, for example from fossil fuel combustion and land use, stays there. Most of the remainder is mopped up by the forests and oceans, which act as "sinks".
However, there appears to be a large carbon sink missing.
"All eyes are now on the Japanese Gosat instrument to search for the missing carbon sink," said Dr Palmer.
Gosat was launched in January from Tanegashima in Japan. It is also designed to monitor atmospheric greenhouse gases.
Nasa's Glory satellite, which is designed to measure carbon soot and other aerosols in the Earth's atmosphere, is due to launch on a Taurus XL from California in June.
But the space agency said it would not fly Glory until the cause of OCO's failure had been investigated.
When the European Space Agency's Cryosat spacecraft was destroyed on launch in 2006, officials decided to re-build it; the launch is scheduled for later in the year. However, the future of the OCO mission remains unclear at this stage.