Astronomers believe they have found a new class of so-called variable stars, which change in brightness.

Brighter than some stellar outbursts called novae but dimmer than the cataclysmic explosions known as supernovae, the brightening events may be caused by the merger of two stars – or by a star swallowing a giant planet.

Astronomers already know of several different phenomena that cause stars to brighten suddenly and then dim.

Novae occur when the dense corpses of stars like our Sun siphon enough hydrogen from a companion star to trigger nuclear fusion on their surfaces. And supernovae brighten dramatically when stars die in enormous explosions.

Now, researchers led by Shrinivas Kulkarni of Caltech in Pasadena, US, have found an object that briefly became about 10 times brighter than the brightest nova but stayed at least 10 times dimmer than a supernova.

Stellar pair

It is the latest in a handful of discoveries that point to a class of variable stars of intermediate brightness. While hunting for such objects, the team found an object in the galaxy M85, which lies about 60 million light years away, that was not bright enough to qualify as a supernova. The initial observations came from California’s Lick Observatory Supernova Search on 7 January 2006.

Using the Palomar Observatory in California, the Keck I telescope in Hawaii and the Spitzer Space Telescope, the team found that the object, known as M85 OT 2006-1, stayed at about the same brightness for more than two months and then gradually faded away.

Kulkarni believes the observations point to a stellar merger. Stars are often found in pairs, and based on the observations, Kulkarni estimates that the stars in this binary system were roughly as massive as the Sun.

Other outliers

They have dubbed the object a “luminous red nova” because of its brightness and ruby hue.

It adds to a growing category of temporarily bright objects in the sky that are neither novae nor supernovae (see Cannibal star: first sighting of a celestial glutton).

These stars seem to range in brightness and age. They include the stars M31 RV and V4332 Sagittarii, which also both turned red when they brightened, and one called V838 Monocerotis (V838 Mon). The light it blasted out illuminated a shell of dust around it, creating a dramatic spectacle that was captured by Hubble (see image below).

Although V838 Mon was about 100 times dimmer than the newly observed object in M85, Kulkarni thinks they may both result from stellar mergers. “I think V838 Mon is like a pale version of what we see, but there must be some underlying commonality,” Kulkarni told ��è����Ƶ.

Swallowing planets

“We thus may have a fundamentally new class of variable stars, which join the novae, supernovae, and other categories,” says Howard Bond, an astronomer at the Space Telescope Science Institute in Baltimore, Maryland, US, who has observed V838 Mon with Hubble.

He says colliding stars could explain all the objects in the new class. “In these objects, the energy being released may be gravitational, rather than thermonuclear.”

However, other astronomers say the brightening in stars such as V838 Mon could be caused by a star swallowing one or more giant planets, possibly as the star bloats up and expands into a red giant – a phase the Sun will one day go through.

But Kulkarni says the object in M85 released too much energy to be easily explained by a planet falling into its star.

Alon Retter, an astronomer formerly with the Pennsylvania State University in University Park, US, disagrees. If the planets sink deep enough into the star, they could release enough energy to explain the observed flare-ups, he says.

“I think this matter hasn’t been solved yet,” Retter told ��è����Ƶ.

The rings around Saturn may be two or three times as massive as previously thought, according to new observations by the Cassini spacecraft.

Previous estimates of the rings’ mass were based on observations by the Voyager spacecraft in 1981. Now, those estimates have been revised upwards with new Cassini measurements of the B ring, one of the planet’s two brightest rings.

They reveal that the ring is quite clumpy and filled with gaps, which Voyager was not able to see.

“We originally thought we would see a uniform cloud of particles. Instead we find that the particles are clumped together with empty spaces in between,” says Larry Esposito, principal investigator for Cassini’s Ultraviolet Imaging Spectrograph (UVIS), which made the observations.

“If you were flying under Saturn’s rings in an airplane,” he continues, “you would see these flashes of sunlight come through the gaps, followed by dark and so forth.”

To make the observations, UVIS watched as the rings appeared to cross in front of a background star. Cassini saw this happen 11 times with stars and twice with the Sun.

Stretch like taffy

This dimming of the starlight or sunlight told researchers how much material was between Cassini and the star. Clumps of ring material blocked more of the light and nearly transparent gaps in the ring blocked less.

These clumps, known as self-gravity wakes, are created by the gravitational attraction of the ring particles to one another. If Saturn was smaller or the rings were farther away from Saturn, these clumps might be able to aggregate and form a moon.

Instead, Saturn’s gravity keeps the particles from clumping together further once they reach about 30 to 50 metres in diameter. At that point, the clumps are stretched apart like taffy because different parts of the clumps are accelerated at different speeds around Saturn. Then the particles move on to form new clumps and repeat the cycle.

“At any given time, most particles are going to be in one of the clumps, but the particles keep moving from clump to clump as clumps are destroyed and new ones are formed,” says team member Joshua Colwell of the University of Central Florida in Orlando, US.

Sheet-like clumps

Similar clumps had been known to exist in Saturn’s outer bright A ring, but this is the first time the clumps were observed in the B ring. In the A ring, the spaces between the clumps are greater, making them easier to spot. In the B ring, the clumps are tightly packed and are just a few metres thick.

“It’s not shocking that these clumps exist in the rings,” Colwell told ��è����Ƶ. “What we’ve done is measured just how flat they are and how prevalent they are in the B ring.”

The newly observed clumps suggest the rings contain two or three times as much mass as previous estimates, which were based on Voyager spacecraft observations of a star passing behind the rings in 1981.

Previously, Cassini, the Hubble Space Telescope and Voyager had spotted “spokes” in the B ring (see Mystery of Saturn’s vanishing ‘spokes’ illuminated).

��è����Ƶs believe spokes are produced when micron-sized dust grains on the surface of boulders in the ring become charged and float above the ring plane. But they do not agree on how the dust particles become charged in the first place.

Journal reference: Icarus (doi:10.1016/j.icarus.2007.03.018)

Cassini: Mission to Saturn – Learn more in our continually updated .

After drifting apart due to an unknown glitch, the two US satellites that are part of the experimental Orbital Express mission rejoined as they flew nearly 500 kilometres above the Earth on Saturday.

During a test on 11 May, the mission’s two satellites, ASTRO and NextSat, undocked and separated to a distance of 10 metres as planned. Then, ASTRO experienced a serious problem with its sensor flight computer. In response, its emergency system kicked in and ASTRO flew 120 metres away to avoid an inadvertent collision with NextSat (see Paired satellites drift dangerously apart).

But ASTRO (Autonomous Space Transfer and Robotic Orbiter) then coasted up to 6 kilometres away from NextSat. Because of the failure, ASTRO was unable to keep track of NextSat’s location on its own, so ground observers had to tell the two spacecraft where they were in relation to each other.

One of the goals of this mission, led by the US Defense Advance Research Projects Agency (DARPA), which manages the mission, is to demonstrate that spacecraft can work somewhat independently. This could eventually be used for a satellite that could fix and refuel ailing satellites in Earth’s orbit.

In this case, mission managers had to come to the aid of ASTRO. Ground controllers fired ASTRO’s thrusters to bring it within 3 kilometres of NextSat. From that distance, ASTRO’s infrared camera and laser rangefinder were able to spot NextSat. On Friday, Boeing, the company that built ASTRO, sent its satellite new guidance information.

“After that, it was all ASTRO’s show,” says Lt. Col. Fred Kennedy, DARPA’s Orbital Express programme manager. “We watched in the mission control centre as ASTRO performed a series of small manoeuvres, bringing it first to within 1 kilometre of NextSat, then 320 metres, 120 metres and finally into the approach corridor.”

ASTRO rejoined NextSat, which serves as an experimental target satellite and depot station, as the pair flew over an area where it was not possible to send real-time images to the ground. So mission controllers had to rely on bits of data to assure them that the docking occurred as planned.

DARPA is analysing what caused ASTRO’s initial computer problem.

For the first time, astronomers have glimpsed one of Uranus’s 27 moons passing in front of one of its siblings – a fleeting alignment that can reveal information about the moons, such as their mass, that cannot be gleaned in any other way.

Researchers hope this will be the first in a bonanza of data returned from Uranus in the next year. That is because Uranus, which orbits the Sun every 84 years, is in an ideal geometry to view its moons. Such periods come around only once every 42 years, and the next will not occur until 2049.

“At best in a human lifetime, you can observe it a maximum of three times – the first time when you’re about two years old, then again in your mid-40s, then assuming that your eyes still hold, you can observe it in your late 80s, if you phase your life correctly,” says Apostolos Christou, an astronomer at Armagh Observatory in Ireland who organised the observations.

The events are so rare because Uranus, unlike the solar system’s other planets, is tilted almost completely on its side, with its rotational axis lying nearly in its orbital plane.

Its moons, however, orbit the icy giant above its equator. So when either of Uranus’s poles is facing the Sun during the planet’s 84-year orbit, the moons do not cross its disc at all as seen from Earth. Instead, they circle it on orbits that resemble the rings around a bull’s eye (scroll down for diagrams).

Technological advances

Now, Uranus is in its equinox, in which the Sun shines directly over the planet’s equator. This means Uranus’s moons and Earth lie in nearly the same plane, providing ideal conditions to view its moons overtaking each other.

On 4 May, Marton Hidas and Tim Brown of Las Cumbres Observatory Global Telescope in Santa Barbara, California, US, did just that.

Using the Faulkes Telescope South in Australia, they saw the 1500-kilometre-wide moon Oberon pass in front of the 1200-km-wide Umbriel. During the event, known as an occultation, the light from the moons dimmed by about 30% as Oberon blocked reflected sunlight from Umbriel.

Astronomers could not make a similar observation during Uranus’s last equinox in the late 1960s because the occultations only last 5 to 10 minutes and existing technologies could not take enough pictures in that short time to adequately measure the dimming.

Seasonal changes

Evaluating data from these occultations could lead to better maps of the major moons. Researchers hope to look at how the brightness of these satellites changes over time, perhaps with the seasons. Brightness changes could suggest that frost, made of volatile compounds, was evaporating and resettling elsewhere on the moon.

The timing of the dimming could also help astronomers better determine the orbits of the moons. By learning more about their orbits, scientists could infer the gravitational tug Uranus has on them and thus the mass of the moons.

There will be 321 total occultations and eclipses of the Uranian satellites. With many telescopes pointed at Uranus during this period, Christou estimates that about 150 of these events will be measured, greatly adding to what is known about Uranus.

Biggest moons

Uranus has 27 moons, but scientists will mainly be concerned with the largest five moons – Miranda, Ariel, Umbriel, Titania and Oberon – since they are easier to observe. Christou’s team hopes to observe another moon dimming event on 29 May. Then, Arial will cast a shadow on Miranda in an eclipse.

In 2006, the Hubble Space Telescope caught that same moon, Arial, casting a shadow on Uranus for the first time (see Hubble takes first image of solar eclipse on Uranus).

The prime observation period lasts until 2008. During this time, researchers will also be taking a close look at the Uranian ring system as the rings will appear “edge on” from Earth’s vantage point.

Two nearby, gas-rich galaxies have been caught crashing into each other in unprecedented detail, pinpointing the location of the two colossal black holes within them. The new observations could shed light on how easy it will be for the black holes themselves to merge one day.

Two disc-shaped galaxies are thought to have brushed past one another before circling around and beginning to merge about 30 million years ago. But studying the resulting galaxy, which lies 320 million light years away and is known as NGC 6240, can be difficult because it is shrouded in patchy dust that blocks light at optical wavelengths.

Now, a team using the 10-metre Keck II telescope on Mauna Kea in Hawaii, US – the largest infrared telescope in the world – has revealed the location of the two supermassive black holes within the merging galaxies to unprecedented precision by the infrared glow of the dense dust and stars around them.

The new observations also show clumps of young stars that formed as a result of the merger (seen as aqua dots in the image). When the gas clouds from each galaxy crashed into one another, shockwaves formed that caused the gas to clump together, triggering intense star formation. “They’re very close to where the black holes are,” says team leader Claire Max, an astronomer at the University of California in Santa Cruz, US.

Get stuck

All of the material surrounding the black holes may play a role in how the black holes will ultimately merge. Some theories suggest that if they cannot lose angular momentum by flinging nearby stars and gas outwards at high speed, they may get “stuck” when they are a few light years apart and just continue spinning around one another. So having a lot of material close to the black holes might enable them to merge into one colossal black hole.

When the two black holes do merge, “it’s not going to be pretty”, says Max. When the gravitational wave detector Laser Interferometer Space Antenna (LISA) launches around 2015, it may be able to measure the ripples in space-time caused by such a massive black hole collision. LISA will not be able to witness the effects of this collision, however – it is not predicted to occur for another 50 to 150 million years.

Max says NGC 6240 is a good example of how supermassive black holes and their surrounding galaxies evolve together over time. Both grow with repeated mergers.

In 2002, the Chandra X-ray Observatory peered behind the dust shrouding the galaxy and saw the two black holes (see Black holes are double trouble for galaxy). The new Keck II observations use adaptive optics, a technique that allows the telescopes to correct for turbulence in Earth’s atmosphere, making them just as precise as if the telescope were in space.

Science Express (doi:10.1126/science.1136205)

Saturn’s gravity causes ice on its moon Enceladus to grind together, generating the plumes of ice crystals and water vapour seen in recent years by the Cassini spacecraft, new calculations suggest.

The findings suggest that any liquid ocean on the moon may be buried beneath an icy shell several kilometres thick, making it difficult to ever retrieve a liquid sample that could be tested for life.

In 2005, the Cassini spacecraft first spied plumes venting from the south pole of Enceladus. They appeared to be coming from so-called “tiger stripes” – 130-kilometre-long cracks in the surface that are 10º to 17º Celsius warmer than the surrounding area.

Several explanations have been proposed for what is powering the plumes, including a heat-producing chemical reaction triggered by charged particles from space and the decay of radioactive isotopes in the moon’s core.

Some researchers suggested that reservoirs of liquid water must lie close to the moon’s surface to supply the geysers with a water source (see Saturn’s watery moon could harbour life). But that conflicted with estimates that the ice would be several kilometres thick, given temperatures on the moon of -193° Celsius.

Now, a group led by Francis Nimmo at the University of California in Santa Cruz, US, calculates that a back-and-forth shearing motion along the tiger stripes – similar to the heat you can create by rubbing your hands back and forth – could generate enough heat to turn ice directly into water vapour.

Good news/bad news

This would eliminate the need for a water ocean close to the moon’s surface, Nimmo says. But he says liquid water probably does lie at depths greater than a few kilometres, since the thick ice shell still has to have a body of water to float on to create these faults. The depth of the ice means it is unlikely, though not impossible, for life to survive in the oceans of Enceladus, he says.

“It’s a sort of good news/bad news story,” Nimmo told ��è����Ƶ. “The bad news is we don’t think there’s water within a few tens of metres of the surface (where spacecraft could conceivably probe), which is a shame because it would be nice to get samples. The good news is we think there is water down there (or) we wouldn’t get enough motion to create fractures of the surface.”

The team calculates that the ice shears or slides by about 0.5 metres every time Enceladus completes an orbit around Saturn, which it does every 1.37 days.

Squeeze and stretch

The shearing motion is caused by so-called tidal forces. These occur because Saturn’s gravity does not tug evenly on the moon – it is stronger on the side of the moon closer to Saturn.

Because Enceladus’s orbit is not perfectly circular, Saturn’s gravity also pulls on the moon more strongly when the moon flies closer to the planet. The tides squeeze and stretch the moon and create movement along the tiger stripes.

A second team, led by Terry Hurford at NASA’s Goddard Space Flight Center in Greenbelt, Maryland, US, has calculated that these same tidal forces could be responsible for the tiger stripes opening and closing in addition to the slipping action.

Stresses during half of the 1.37-day orbit cause the cracks to open like zippers, exposing volatile compounds and allowing the ice to erupt in the form of water vapour plumes. During the other half of the orbit, the ice moves in the opposite direction and the cracks close.

Large slips

Hurford hopes to study data from a Cassini flyby of Enceladus on 24 April. At that time, some of the tiger stripes were expected to be opening up as the ice cracks pulled apart. Cassini may have been able to see whether there were plumes emanating from the stripes at this time.

But Andrew Dombard of the Applied Physics Laboratory at Johns Hopkins University in Laurel, Maryland, US, says the results from the two studies do not totally mesh. If the tiger stripes were open during half of their orbit, he says, it would reduce the frictional contact between the two sides of the rift and in turn limit the heating.

He also says that if the sides were under compression during the other half of the orbit, the built-up stress could produce sudden slipping motions along the cracks that could be much larger than the 0.5 metres calculated by Nimmo. He points out, however, that this action has not been observed by telescopes.

Cassini: Mission to Saturn – Learn more in our continually updated .

Journal reference: Nature (vol 447, p 239) and (vol 447, p 292)

The US military’s experimental satellite mission, Orbital Express, may be in peril. The lead spacecraft ASTRO has drifted several kilometres away from where it should be.

The $300 million mission is designed to test autonomous operations between two spacecraft so that one day, spacecraft may be able to repair or refuel satellites on their own.

During a test on 11 May, the mission’s two satellites, ASTRO and NextSat, undocked and separated to a distance of 10 metres as planned. Then, ASTRO (Autonomous Space Transfer and Robotic Orbiter) experienced a serious problem with its sensor flight computer. In response, its emergency system kicked in and ASTRO flew 120 metres away to avoid an inadvertent collision with NextSat.

Since then, ASTRO has coasted several kilometres away from NextSat. Amateur satellite watcher Ted Molczan observed ASTRO trailing NextSat by about 2 kilometres on 12 May. By the next day, he saw they had changed positions, with NextSat trailing ASTRO by 5 kilometres.

“Both vehicles are safe,” says Jan Walker, spokesperson for the Defense Advanced Research Projects Agency, which is running the mission.

Unfortunately, although ASTRO could see NextSat, it had difficulty tracking it – and thus lost the ability to navigate relative to NextSat – since they were farther apart than planned for this test.

Unknown cause

ASTRO was supposed to have moved just 30 metres away from NextSat before returning and docking with the satellite. Watch a Boeing video showing what in this test.

Now, ASTRO is running on a back-up flight computer, though ground controllers have not determined a root cause of the problem.

This was not the first undocked operation for the duo. On 5 May, ASTRO flew 10 metres away from NextSat and was able to return safely (see US achieves autonomous docking in space).

Over the course of its 90-day mission, it is supposed to run through a series of these scenarios, intentionally increasing its distance between the two spacecraft. Its final tests are supposed to have ASTRO fly up to 7 kilometres away from NextSat.

Mission managers hope to fly ASTRO back to NextSat within several days, though they admit the manoeuvre will be challenging.

Tense moments

“We are performing an approach and remate from long range weeks ahead of schedule,” Walker told ��è����Ƶ. “We had planned to take a ‘crawl, walk, run’ approach to tests, but we are within the ranges anticipated for our long range approach and remate activities.”

Before this problem, ASTRO had already experienced some tense moments in its mission. Shortly after its launch, it had problems pointing in the right direction because some of its hardware was accidentally installed upside down (see ‘Mechanic’ satellite suffers guidance system glitch).

A previous US attempt at autonomous satellite operations took place in April 2005, with the $110 million DART (Demonstration of Autonomous Rendezvous Technology) mission. This failed, however, when the spacecraft crashed into its target satellite after suffering serious problems with its navigation systems (see Spacecraft collision due to catalogue of errors).

The rocket company Armadillo Aerospace has just completed a 3-minute hover test of its vehicle Pixel, positioning itself to win the most challenging level of the $2 million Lunar Lander Centennial Challenge in October 2007. But as the only entrant in the 2006 competition, this team knows anything can happen on ‘game day’.

As the US prepares to return humans to the Moon, ambitious young companies are trying to build their own versions of a lunar lander that could land and take off vertically from the Moon.

The Northrop Grumman Lunar Lander Challenge, a NASA-backed contest, aims to spur the development of the required technologies. The competition has two levels. Level 1 requires two 90-second flights, and Level 2 requires two flights of 180 seconds each.

Armadillo Aerospace, co-founded by Doom video game creator John Carmack, made three attempts to win the shorter, Level 1 contest in 2006, but ultimately did not walk away with the top prize, worth $350,000 prize.

But on Saturday, the company flew its Pixel lander for 192 seconds at the Grayson County Airport in Denison, Texas, US. This is its longest flight so far – and crucially, makes it a serious contender for the lander contest’s $1.5 million Level 2 prize. Watch a (.mpg), where heated concrete appears to pop off the launch pad at times.

Rough terrain

Pixel quickly made another flight that lasted more than 90 seconds.

Armadillo has attempted a 180-second test flight four or five times since the lander challenge in 2006. But they have always been hampered by engine malfunctions, guidance issues or other glitches.

Still, the 192-second flight does not necessarily mean Armadillo will win the 2007 competition, which is to be held at the Wirefly X Prize Cup in Las Cruces, New Mexico, US, from 26 through 28 October.

That is because both levels of the contest require rockets to launch and then hover for the required time at an altitude of 50 metres while moving to a second launch pad 100 metres away. Then they must land vertically like a lunar lander, refuel and repeat the process back to the first pad.

For the contest’s Level 2, the rocket must also land on simulated lunar terrain, which includes craters and boulders.

Tethered test

During Saturday’s 192-second test flight, Pixel was roughly 3 to 4 metres off the ground – it would have to fly about 15 times higher for the contest. And on Saturday, tethers connected the vehicle to a crane so it would be supported before dropping to the ground.

That means the recent flight was not a good test of its landing gear, which played a key role in the failed 2006 challenge attempts. At that time, this gear was damaged during a rough landing.

Since then, the company has redesigned the landing gear. It is now made of foam wrapped in leather, which gives Pixel some added cushioning on impact.

They have also redesigned the engine several times to keep the engine chamber from overheating during the longer flights.

Team member Phil Eaton says he thinks they can make both 180-second flights to win the Level 2 challenge this year, not to mention the Level 1 prize.

‘In the bag’

“We made it a very specific goal to make sure that we had the Level 1 prize in the bag before we continued to do any additional work on the Level 2 prize,” Eaton told ��è����Ƶ.

In the spirit of the competition, he says they have also been offering advice to their competitors. “We want them to be as successful as possible while still allowing us to win,” Eaton says.

The contest is one of NASA’s Centennial Challenges, which are designed to spur private innovation in spaceflight. Although some of the contests have been going since 2005, NASA had never given out any prize money until the Astronaut Glove Challenge earlier in May, when Peter Homer won $200,000 for his simple glove design (see New spacesuit glove beats NASA’s, hands down).

A trio of mini-satellites has failed in their attempt to deploy a kilometre-long tether in space.

The setback means the low-cost Multi-Application Survivable Tether (MAST) experiment, launched on 17 April, may not achieve its goal of testing the survivability of a thin, braided tether in space.

Over the past week, mission managers determined that the tether-deploying element, known as Ted, had properly separated from the tether inspector, a tiny satellite called Gadget. But a glitch in the restraint system kept Ted from pushing away hard enough to keep unreeling the tether from its spool. So the tether deployed just a few metres, rather than a full kilometre.

Robert Hoyt, chief executive officer of Tethers Unlimited, which designed the picosatellites, says mission managers suspect they know what caused the glitch, but the company is not ready to disclose this to the public yet. “I don’t think we’ll ever know for sure,” he says.

Space tethers could one day be used to fling satellites into different orbits, thus saving satellite companies money on fuel.

Or tethers could enable clusters of satellites to fly in formation and prevent them from drifting away from one another over time. Such an application might be useful in interferometry, where images from several telescopes, spaced some distance apart, are combined to give greater resolution.

Some data

Despite the setback, the MAST team at Tethers Unlimited, a company in Bothell, Washington, US, still may be able to get other data from Gadget to learn how a short tether behaves in microgravity.

MAST team members discussed having Gadget crawl down the tether to Ted to try to restart the deployment, but they decided that option was too risky. “If we were to have Gadget start to crawl, there is the possibility of the satellites banging together, which would be very likely to damage solar cells and other systems,” Hoyt says.

This was not the first setback for the mission, which costs less than $1 million. After launch, the satellite team could not get a signal from Ted (see No signal yet heard from tether-deploying satellite). But they said that this should not have affected Ted’s ability to deploy the tether.

Longest tether

Then, sky watchers who had been on the lookout for the deployed tether and satellites from the ground had not seen anything when MAST was scheduled to appear overhead. “That’s one confirmation that the tether is not deployed to a very long length,” Hoyt told ��è����Ƶ.

In other space tether news, the longest planned space tether just got a little closer to launch. The satellite, a project of 500 students in Europe known as Young Engineers Satellite 2 (YES2), was shipped to its launch site in Russia from the Netherlands on 10 May.

YES2, a project of the European Space Agency, is scheduled to launch in September. If everything goes as planned, the satellite will unroll a 30-kilometre-long tether that is a mere 0.5 millimetres thick. The end of the tether will be attached to a small round capsule called Fotino that will eventually re-enter Earth’s atmosphere and attempt to land.

If it seems Canadians weigh less than their American neighbours, they do – but not for the reasons you might think. A large swath of Canada actually boasts lower gravity than its surroundings.

Researchers have puzzled for years over whether this was due to the crust there rebounding slowly after the end of the last ice age or a deeper issue involving convection in the Earth’s mantle – or some combination of the two.

Now, ultra-precise measurements taken over four years by a pair of satellites known as GRACE (Gravity Recovery and Climate Experiment) reveal that each effect is equally responsible for Canada’s low gravity. The work could shed light on how continents form and evolve over time.

GRACE, a joint mission of NASA and the German Aerospace Center, was launched into space in 2002. The two spacecraft fly 500 kilometres above the Earth, 220 kilometres apart. Using a microwave ranging system, the two spacecraft can measure distance differences between them as tiny as a micron.

That allows them to measure tiny changes in the distribution of mass – and hence gravity – on the Earth. For example, if the leading spacecraft were to encounter an area with more gravity, it would be pulled ever-so-slightly closer to Earth than the trailing spacecraft, and that distance can be measured.

Crushing weight

A team led by geophysicist Mark Tamisiea, who performed the work while at the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts, US, has now used GRACE to study the gravitational low over Canada’s Hudson Bay. ��è����Ƶs first noticed this low in the 1960s, when the planet’s first global gravity fields were mapped.

At first, researchers suspected it was due to an ice sheet called Laurentide that blanketed a sizeable chunk of North America during the last ice age. In places, the sheet was more than 3 kilometres thick, and it depressed the Earth’s crust beneath it.

When the ice age ended about 20,000 years ago, the ice rapidly melted. But the crust has been springing back much more slowly, and it is rebounding today by about 12 millimetres per year.

But in the last decade or so, scientists have begun to suspect that convection in the Earth’s mantle, a layer of hot, flowing rock beneath the crust, also plays a role.

The sludge-like mantle rises and falls in plumes as it is heated from below and cooled from above. The mantle can drag the overlying tectonic plates with it as it moves.

What lies beneath

GRACE cannot directly detect that movement since it is so slow. But scientists inferred the gravitational contribution of convection by subtracting the post-glacier effect from the region’s overall gravity signal.

“It’s a very good piece of evidence that allows us to look beneath the surface of the Earth,” says team member James Davis of the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts, US. “It would be very hard for someone to say it’s 100% post-glacial rebound.”

Even after the Earth’s crust rebounds completely from the glacier melting, there may still be a gravitational low over the area due to mantle convection. That would suggest that even parts of a continent away from the tectonic plate boundaries are affected by mantle convection.

Climate change

Ultimately, scientists hope to use such data to learn how continents form and evolve over time. “With this information, people could infer better whether the North America plate is actually predominantly stable,” says C K Shum, an expert on the Earth’s gravity at Ohio State University in Columbus, US, who was not a part of this study.

Using this same set of data, researchers have also found that there were two ice domes in the ice sheet on either side of Hudson Bay, given the features in the gravity field that have been left behind. The way the thickness of any ice cover changes at different periods could be used by climate modellers to understand past climate change.

“Where the ice formed, and how thick it was, is a clue to temperature, precipitation, and other weather/climate (for example, the jet stream) changes in the distant past,” says team member Jerry Mitrovica of the University of Toronto in Canada.

Journal reference: Science (vol 316, p 881)