Harvard astronomers and a team of scientists from NASA’s Goddard Space Flight Center (GSFC) announced at a press conference on Tuesday that they measured thermal radiation from two planets orbiting stars outside our solar system, marking the first successful direct observation of extrasolar planets.

Over the last 10 years, astronomers have detected more than 130 planets around other stars, but the detections have always been indirect, according to Assistant Professor of Astronomy David Charbonneau, who led the Harvard team.

Previously, astronomers inferred the presence of planets from their effect on their central stars’ orbit, he said.

“Until yesterday, no one had talked about measuring [light from the planet itself] directly,” said Charbonneau, who researches at the Harvard-Smithsonian Center for Astrophysics, based in Cambridge.

Charbonneau said his previous attempts to obtain infrared data from a ground telescope failed because thermal radiation from the earth’s atmosphere and the telescope overwhelmed the signal from the planet.

The researchers obtained the results from the two-year-old Spitzer Space Telescope, which observes infrared wavelengths.

“Spitzer is the most sensible and stable infrared telescope out there,” Charbonneau said.

Space, he said, is a much “cleaner” environment for astrophysical observation because infrared light from extrasolar planets is easier to detect when frigid temperatures lessen interfering emissions from other sources.

Spitzer provided researchers with direct observations about the temperature and color of two planets—TrES-1 and HD 209458b—and also dispelled a long-accepted explanation for the size of one of the planets.

Harvard’s team observed light from TrES-1, while GSFC’s team observed light from HD 209458b, but Charbonneau said the similarity of the results makes them even more convincing.

“The exciting part was that not one, but two planets had been detected by this technique,” he said.

To isolate TrES-1’s infrared emissions, Charbonneau said his team used Spitzer to obtain data when the planet was in transit, meaning it was passing in front of its star, and during the secondary eclipse, when the planet was behind the star. The light from the star was then subtracted from the light from both the star and the planet. The data, collected for six hours on Halloween, allowed Charbonneau and his colleagues to determine that there was a signal from TrES-1.

“We were really ecstatic to finally see this long sought-after signal,” Charbonneau said. “It’s the first time this technique has been used successfully.”

With the information they collected, the two teams directly observed the temperatures on the two planets and examined each planet’s brightness in different colors of light, allowing them to discern the presence of molecules such as carbon monoxide.

In addition, Charbonneau said the GSFC team discovered that HD 209458b, which is 70 percent of Jupiter’s mass but 35 percent larger, orbits in a circular path. Previously, he said, astronomers believed that the planet’s surprising largeness could be explained by an elliptical orbit, but said this theory no longer holds.

Charbonneau said he hopes to make infrared observations on smaller, rockier planets in the future, which he said is exciting because these planets are similar to Earth.

“Our big hope is...to compare planets and see how they differ from properties of planets in our own solar system,” he said.

Professor of Astronomy Dimitar D. Sasselov said he foresees further research in this area throughout the coming year.

“We’ve made the discovery and now it’s time for exploration,” he said.

­—Staff writer Lulu Zhou can be reached at luluzhou@fas.harvard.edu.