A Planet That Wasn’t: Uncovering the True Nature of 42 Draconis b

In 2009, astronomers believed they had found a planet orbiting the K giant star 42 Draconis, a star several times the size of the Sun. Using radial velocity (RV) measurements—tiny shifts in the star’s light caused by movement—they detected a periodic signal that pointed to a planet nearly four times the mass of Jupiter, orbiting every 479 days. K giant stars like 42 Draconis are often targeted in exoplanet searches because their slow rotation and cooler temperatures make it easier to detect such wobbles. But giant stars also introduce complications, such as pulsations and surface activity, that can create signals mimicking those of planets.

Revisiting the Evidence with Long-Term Observations

In this new study, led by Artie Hatzes, the team added more than a decade’s worth of RV measurements to the original dataset. The early data supported the idea of a planet, with a strong and steady signal. But the new measurements told a different story. The strength of the 479-day signal had dropped from 110 m/s to just 27 m/s. If the wobble were due to a planet, it should remain constant over time. This fading amplitude was one of the first signs that something more complex was going on.

The Discovery of Multiple Periods

Digging deeper, the team performed a frequency analysis of the full dataset, which revealed a second RV signal at 530 days. When two similar signals are present, they can interfere and create a "beating" effect, where the combined signal appears to rise and fall in strength over time. This could easily be mistaken for a single, fading planetary signal. Because the two periods are so close, if both were caused by planets, they would orbit dangerously near each other—an arrangement that would be gravitationally unstable. This provided further evidence that the signals likely didn’t come from planets.

Infrared Photometry and the Final Blow to the Planet Hypothesis

To test whether the RV signals were truly caused by a planet, the researchers turned to data from the COBE/DIRBE space mission, which measured the star’s brightness in infrared light. The DIRBE data revealed two significant brightness variations—one matching the 479-day RV period and another at 170 days. This means that the star itself was changing in brightness on the same timescales as the supposed planet’s orbit. Since planets don't cause stars to brighten and dim periodically, this was a critical piece of evidence showing the signal wasn’t planetary in origin.

Looking for Other Signs of Stellar Activity

The authors also looked for other signs that the star might be responsible for the RV changes. They analyzed the shape of spectral lines from the star's light—features that can shift due to surface activity like spots or convection. The measurements showed variability, but not at the 479-day period. Older data from the Hipparcos mission was also reexamined, showing possible variations at 690 days, though with low confidence. Together, these findings support the idea that multiple types of stellar activity are present, none of which match the steady pull expected from a planet.

Short-Term Variability and Stellar Oscillations

Even on shorter timescales, 42 Draconis showed interesting behavior. Over several nights, the team recorded RV variations with periods of less than two days and amplitudes up to 80 m/s. These match theoretical predictions for p-mode oscillations—vibrations that occur within stars. These short-term oscillations are known to occur in giant stars and add further evidence that the variations are caused by the star itself rather than an orbiting companion.

Conclusion: A Case Study in Caution

42 Draconis b was once a textbook case of a confirmed exoplanet. It passed all the standard tests, and its discovery was even recognized by the International Astronomical Union, who named it "Orbitar." But this new analysis shows that all the signs pointed in the wrong direction. The signal wasn’t from a planet, but from the complex behavior of the star itself—possibly long-period stellar oscillations. This raises concerns about other claimed exoplanets around giant stars. How many are real, and how many are illusions created by misunderstood stellar variability?

Implications for the Future of Planet Detection

This study highlights the importance of long-term monitoring and the need for multiple, independent checks before declaring a planet discovery. Especially for K giants, signals that appear planetary may instead come from pulsations, rotation, or other intrinsic processes. As astronomers refine their techniques and tools like the Gaia satellite add new data, they’ll be better equipped to separate real planets from stellar impostors. In the meantime, 42 Draconis serves as both a cautionary tale and a reminder of the self-correcting nature of science.

Source: Hatzes