Dark matter seems to be receding further into the shadows. Last year, researchers thought they may have spotted its signature when a balloon-borne experiment called ATIC detected a bizarre spike in the number of high-energy electrons streaming in from space.
But now, NASA鈥檚 space telescope finds no such spike 鈥� only subtle hints of a slight increase, suggesting that dark matter is not leaving any obvious trace in the charged particles detected from space.
Nobody knows exactly what dark matter is, but the leading theoretical model posits that it is made of up of particles called WIMPs (weakly interacting massive particles). When two WIMPS collide, the theory says, they annihilate, producing radiation and a cascade of particles, including electrons.
Advertisement
So researchers were excited last November, when a team studying data from two ATIC balloon flights over Antarctica reported finding many more electrons than expected at high energies around 600 gigaelectronvolts. Less exotic objects, such as pulsars and supernova remnants, also accelerate charged particles to high energies, so the ATIC data could potentially be explained by such garden-variety fare.
But the abundance and high energies of the 鈥榚xtra鈥� electrons detected, coupled with another unexpected cosmic ray result measured a few months earlier by a satellite called PAMELA, raised the tantalising possibility that dark matter 鈥� perhaps of an exotic type 鈥� might be responsible.
鈥淧article physicists have not had much to get excited about in the last 10 years 鈥� they were all ready for the Large Hadron Collider and then had a big setback鈥� when it broke down, says , a dark matter theorist at the Harvard-Smithsonian Center for Astrophysics. 鈥淭hen PAMELA and ATIC came along with extra high-energy signals that could not be easily explained, and it was fun to think about.鈥�
Now, however, astrophysicists using the Fermi telescope say they don鈥檛 see a dramatic spike in the number of high-energy electrons in space. 鈥淥ur energy spectrum doesn鈥檛 have prominent features,鈥� says Alexander Moiseev, a Fermi team leader at NASA鈥檚 Goddard Space Flight Center in Greenbelt, Maryland.
Pros and cons
What could explain the discrepancy? For one thing, the two experiments have different strengths and weaknesses.
ATIC flew in Earth鈥檚 atmosphere, which can create extra 鈥渘oise鈥� in its signal, while Fermi is a space mission. ATIC鈥檚 balloon flights also lasted for no more than three weeks, while Fermi is constantly taking data from orbit.
Indeed, the Fermi team analysed more than 4 million high-energy electrons detected with the telescope over the course of about six months to arrive at their result, collecting hundreds of times more data at these energies than any previous measurement. 鈥淲e are practically free of statistical errors,鈥� Moiseev told New 杏吧原创.
But ATIC has a thicker calorimeter, an instrument at the bottom of its detector that incoming space particles strike, generating showers of other particles. 鈥淭he deeper or thicker that calorimeter is, the less of that shower energy sneaks out the bottom,鈥� says ATIC team leader of Louisiana State University in Baton Rouge.
鈥淭hey can contain 68% of the energy, and we contain 85% of the energy,鈥� Wefel told New 杏吧原创. As a result, he says there is more uncertainty in Fermi鈥檚 measurement of the energy of incoming particles, which could broaden out any dramatic spikes like the one seen by ATIC.
鈥淭he reason they鈥檙e not seeing that peak structure is because they have much poorer energy resolution in their instrument,鈥� says Wefel. He adds that his team has analysed a third balloon flight since the original ATIC announcement in November and finds the same sharp peak as before.
Instrumental effect?
The Fermi team acknowledges that it has a thinner calorimeter but says its detector is better in other ways 鈥� it boasts an instrument that tracks the path of incoming particles, for example 鈥� something that ATIC does not have. It has also run detailed computer algorithms that show its energy resolution is sharp enough to be able to see a spike in energetic electrons. 鈥淲e would see an ATIC-like bump with huge confidence if it were there,鈥� maintains Moiseev.
, a physicist at the University of Michigan who is not affiliated with either team, agrees. The bump seen by ATIC 鈥渨as probably an instrumental effect they hadn鈥檛 compensated for鈥�, he says.
Both experiments have to grapple with the same basic challenge, Tarle explains 鈥� distinguishing between electrons and the much more abundant protons that pass through their detectors from space. Since neither experiment uses a magnetic field that could tell the two kinds of charged particles apart, the teams must try to do this by analysing the characteristics of the particle showers in their detectors.
鈥淚t鈥檚 hard to make these measurements 鈥� very hard,鈥� says Tarle. But he says Fermi鈥檚 calorimeter is better suited for the analysis than ATIC鈥檚. It is made of atoms that have a higher number of protons, which do not readily interact with protons coming in from space. That causes 鈥渓ess contamination in the electron data鈥�, he says.
Dark matter
So if the ATIC bump isn鈥檛 real, what does that mean for dark matter?
Tarle says it means that high-energy electron detectors such as ATIC and Fermi do not show any evidence for dark matter. 鈥淭here鈥檚 nothing in their data that could indicate new physics,鈥� he says.
But other researchers say Fermi鈥檚 data does show what may be a subtle sign of dark matter. If they look at the data in the most conservative way, Fermi team members do not see this potential signature 鈥� they say the electron energy spectrum they measure is smooth, without any wiggles that might indicate 鈥榚xtra鈥� electrons.
If they are not as conservative, however, Fermi team members say they see a slight bump in the number of electrons at higher energies 鈥� though nothing as dramatic as ATIC鈥檚.
That gentle bump, they say, might be due to a slight theoretical underestimation of how many high-energy cosmic rays are produced in objects such as pulsars 鈥� an idea Tarle favours.
鈥淭he most likely explanation of the excess electrons at high energy seen by Fermi is that the theoretical estimates are wrong,鈥� Tarle says. 鈥淭here is no reason to believe that these theoretical predictions based on lower energy data are valid in the high-energy regime of Fermi.鈥�
鈥楬ard to fit鈥�
Alternatively, it might be due to one or more nearby sources that are pumping out energetic electrons. The sources are thought to be nearby because high-energy electrons lose energy as they travel through space, so for them to arrive at the energies that Fermi detects, they must have come from somewhere within about 3000 light years of Earth.
The nearby sources could be pulsars, but 鈥渄ark matter is not ruled out鈥� as a possible source, says Moiseev.
Finkbeiner agrees. Last year, he and colleagues came up with a new model of dark matter that could account for both the PAMELA and ATIC signals.
After the Fermi team released its results at a earlier this week, Finkbeiner said his inbox was flooded with emails saying, 鈥淪o, annihilating dark matter is dead, right?鈥� he says. 鈥淣othing could be further from the truth.鈥�
鈥淚t was always a little bit hard to fit the ATIC bump,鈥� he says, explaining that such a sharp spike hints that dark matter might be annihilating straight to electrons 鈥� a process that is theoretically forbidden.
鈥楲ess information鈥�
His and other dark matter models instead argue that annihilating dark matter particles create intermediate particles 鈥� such as pions 鈥� before producing electrons.
鈥淚t鈥檚 hard to make a sharp feature but easy to make a broad, smooth feature鈥� like the one Fermi may be seeing, he says, adding that the same is true for electrons produced in astrophysical sources such as pulsars.
鈥淚n a way, it鈥檚 a relief we don鈥檛 have to make the ATIC bump, but if ATIC is real, it would really be telling us something,鈥� Finkbeiner told New 杏吧原创. 鈥淲e鈥檙e not likely to learn as much about dark matter from [Fermi鈥檚 electron spectrum] 鈥� basically, we have less information than we had before.鈥�
Tricky observation
If further observations with Fermi suggest there is not even a gentle rise in the number of high-energy electrons it detects, that will make any annihilating dark matter difficult to observe 鈥� but it will necessarily not rule it out, says Finkbeiner.
鈥淏efore the ATIC and PAMELA results, the expected annihilation signal for the leading dark matter candidate, the WIMP, was much smaller, so failure to find a signal with Fermi does not in any way rule out conventional WIMP annihilation,鈥� he says.
鈥淥f course, there could be no signal at all: dark matter could just sit there and gravitate and do absolutely nothing else,鈥� he adds. 鈥淭hat鈥檚 kind of the most boring scenario: we can never learn what kind of particle it is.鈥�
Tracing the source
The Fermi team hopes to shed light on the issue by continuing to collect electron data from all over the sky. It鈥檚 difficult to trace the source of electrons that fall into its detector because the charged particles are diverted by magnetic fields in space. But if Fermi detects even a slight excess of electrons in one region of the sky, it might point to their source, says Moiseev.
Fermi is also hunting for possible signs of dark matter in the distribution of gamma-ray photons in the sky. Gamma rays are thought to be produced by annihilating dark matter and unlike electrons, are not affected by intervening magnetic fields (see Where will new Fermi telescope find dark matter?).
Future experiments might also provide a cross-check of both ATIC and Fermi. One, called the Alpha Magnetic Spectrometer, may fly to the International Space Station before the shuttles are retired in 2010. It uses a magnetic field to separate charged particles and has a calorimeter a little thicker than Fermi鈥檚.
Journal reference:
