North Korea informed the United States of its nuclear test less than an hour before the widely condemned explosion, a US official confirmed Monday.

A senior administration official said on condition of anonymity that the North informed the US State Department of "its intention to conduct a nuclear test, without citing a specific timing."

The official did not say how Pyongyang informed Washington, but communication between the two countries -- which do not have diplomatic ties -- is generally conducted through their UN missions in New York.

North Korea's notice reached the State Department between 8:00 and 9:00 pm Sunday (0000 and 0100 GMT Monday), the official said. According to the US Geological Survey, a 4.7-magnitude tremor shook North Korea at 0054 GMT.

The official said Obama was informed of the seismic observation in North Korea at 11:15 pm (0315 GMT) and was on the telephone 45 minutes later with his national security adviser, Jim Jones, a retired general.

A South Korean official had earlier said the North had given prior notice of its nuclear test both to the United States and China, Pyongyang's main ally.

North Korea also gave prior notice last month of its defiant long-range missile launch in May.

Some analysts say the advance notice shows that forces in North Korea are determined to carry out the tests and hope to lessen the country's international opprobrium.

Full confirmation of NKorea blast will take time: scientists
Confirmation that North Korea has carried out a full-throated nuclear blast - as opposed to a fizzler or a conventional explosion designed to fool its enemies - could take a number of days, say scientists.

Verification experts have a panoply of techniques to determine when and where underground detonations take place and how big they are.

But only one - detection of nuclear particles or rare gases vented into the atmosphere from deep beneath the ground - can "unambiguously" show a blast is nuclear in origin, says Vertic, an independent non-governmental organisation in London.

Confirmation of North Korea's second claimed nuclear test will depend on the weather for blowing such evidence towards monitoring stations, the Comprehensive Nuclear Test Ban Treaty Organisation (CTBTO) said in Vienna on Monday.

"It would take at least two days for closest stations to be reached, and the farther we go, the more time it takes," said CTBTO chief Tibor Toth.

The monitors' frontline tool is the seismological sensor, of the kind that also listens out for earthquakes.

The sensitivity of these instruments, and their expanding network around the world, has vastly improved surveillance over the past two decades.

Even so, the technology has flaws and this was demonstrated by North Korea's first nuclear test on October 9, 2006.

Writing in the journal Nature Physics in 2007, Paul Richards and Won-Young Kim of the Lamont-Doherty Earth Observatory at New York's Columbia University reported how a seismic "event" was swiftly picked up by a station at Mudanjiang, northern China, 370 kilometres (230 miles) north of North Korea's test site.

Within five hours, other stations gave an excellent fix as to location and depth and estimated the "event" as a low 4.0 magnitude on the seismic scale.

But this was the easy part, said Richards and Kim: "Identifying the event as an explosion, not an earthquake - and as a nuclear explosion - was more difficult."

A signature from a typical 4.0-magnitude earthquake begins with a rumble of waves that amplifies into a thick squiggle on the seismogram, followed by another packet of waves that then tails away.

The North Korean signal, though, comprised two sharp, early bursts that swiftly faded to an angry buzz, and this showed it was man-made.

"When they are of comparable magnitude, an explosion has a more condensed signal than an earthquake," explained Bruno Seignier of France's Atomic Energy Commission (CEA).

"In an explosion, energy is released in an extremely violent fashion, signaled by a release in higher frequencies."

After determining that a bomb was indeed the cause, the scientists next had to estimate its yield and what kind of device made the blast.

On the first score, the estimates were at first wide-ranging, possibly confused by sensors' distance from the blast and the amplitude of signals carried through Earth's crust.

The 2006 blast was initially estimated by Russia at the equivalent of between 5,000 and 15,000 tonnes of TNT.

That would have potentially have put it in the same range as "Little Boy," the US atomic bomb that destroyed Hiroshima in August 1945, which was around 12.5 kilotonnes.

But other experts swiftly downgraded the test to a kilotonne or even less.

Monday's blast, measured at around 4.5 magnitude, was estimated by Russia at around 10 and 20 kilotonnes, but the CTBTO put it "in the low kilotonne range for this particular magnitude."

As for the type of bomb, seismograms may give little help at very low blast levels.

In fact, the seismic signature in 2006 "remarkably" mirrored that of a blast by 2,000 tonnes of conventional explosive, said Richards and Kim.

Confusion was such that experts speculated for several days that Pyongyang had simply detonated a huge pile of TNT to con the world into believing it had joined the nuclear club.

What weighed against this argument was that North Korea would have had to dig a big tunnel and truck in huge amounts of explosives, which would have been spotted by spy satellites.

The speculation was laid to rest several days later when radionuclide sensors proved the blast was nuclear.

But this evidence failed to still debate, continuing to this day, that the bomb may have been a dud - a big device that fizzled out after failing to achieve a sustained nuclear reaction.