The brightening and subsequent dimming of the background star due to microlensing does not always follow a smooth bell curve, however. In about a dozen cases identified by OGLE and by the similar Microlensing Observations in Astrophysics (MOA) experiment based at Mount John University Observatory in New Zealand, irregularities in the lensing signal point to a planet orbiting the foreground star and distorting the symmetry of the lens. The duration of a deviation from the bell curve indicates the suspected planet's mass.
Microlensing has its downside—the planetary signals are ephemeral, lasting only as long as the background star and the planet-hosting star remain in alignment (typically about a month). But it has one critical advantage over other planet-hunting techniques: it is sensitive to bodies not especially close to their parent stars. More prolific planet-search methods, including the technique employed by NASA's Kepler spacecraft, which detects periodic variations in starlight caused by orbiting planets eclipsing their stars, have the most success detecting planets that orbit very close to their host stars and hence complete an orbit very quickly.
"Microlensing can probe planets of all masses for a very large range of orbital separations," from about 0.5 times to 10 times the Earth–sun distance, Cassan says. He notes that the abundance estimates can only increase with exploration of a larger range of orbital distances and planetary masses. "Our results are given for masses between five Earths and 10 Jupiter masses," Cassan says. "If there are other planets farther or closer in, the average number of planets per star would increase accordingly."
The conclusion that smaller planets occur more often than bigger ones reinforces what Kepler has shown for planets that orbit close to their stars. The spacecraft is designed to locate worlds similar to our own—small, rocky planets at temperate, Earth-like distances from their host stars. That hunt is still underway, but early results from the mission have revealed that smallish planets—those just a bit bigger than Earth—are common in the hotter, close-in orbits to which Kepler is already sensitive.
"Kepler has already been finding that small planets are actually quite ubiquitous around stars," Gaudi says. "That bodes well for our goal of eventually finding an Earth-size planet in the habitable zone. All signs are pointing to the low-mass planets being common, so I think there's a good chance that we'll find a system like that in the coming years."