Periodicities in radio emissions from the Jupiter's magnetosphere and consequences for radio emissions from star-exoplanet systems

TL;DR

This paper tackles the challenge of detecting periodic radio emissions from exoplanets and star–planet interactions under unevenly spaced observations. Using the Lomb-Scargle periodogram, it is validated by simulations and by approximately $1400$ hours of Jupiter data from the NenuFAR telescope to show recovery of true periodicities and the emergence of beat and harmonic artifacts that inform signal detection. Key Jovian periodicities are identified, including $T_ ext{Jupiter} \\approx 9.93\,\mathrm{h}$, $T_ ext{beat IJ} \\approx 12.96\,\mathrm{h}$, $T_ ext{Io/2} \\approx 21.23\,\mathrm{h}$, and $T_ ext{Day} \\approx 23.93\,\mathrm{h}$, with windowing contributing additional peaks that can reinforce weak signals. The work demonstrates the power of LS analysis for exoplanetary radio searches and provides practical guidance on data handling, windowing, and resolution to enable robust inferences about magnetic topology and star–planet interactions.

Abstract

The search for radio signals from exoplanets or star-planet interactions is a topic of major scientific interest, as it is likely the best way to detect and measure a planetary magnetic field and, therefore, to probe the inner structure of exoplanets. However, detecting these radio emissions is challenging, since they are anisotropic by nature, sporadic, and of low intensity because of their great distances, and because the sky cannot be monitored continuously. The aim of this article is to demonstrate the relevance of using statistical tools to detect periodic radio signals in unevenly spaced observations, and identify the implications of the measured period. The identification of periodic radio signals is achieved here by a Lomb-Scargle analysis. We first apply the technique to simulated astrophysical observations with controlled simulated noise. This allows us to characterize the origin of spurious detection peaks in the resulting periodograms, as well as to identify peaks corresponding to real periods in the studied system, and to harmonic or beat periods. We then validate this method with a real signal, using approximately 1400 hours of data from observations of Jupiter's radio emissions by the NenuFAR radio telescope over more than six years, to detect the periodicities of Jovian radio emissions (auroral and induced by the Galilean moons). We demonstrate with the simulation that the LombScargle periodogram allows us to correctly identify periodic radio signals, even in a diluted signal. On real measurements, it correctly detects the rotation period of the strong signal produced by Jupiter and the beat period of the emission triggered by the interaction between Jupiter and its Galilean moon Io, ...

Figures (11)

• Figure 1: (left) Randomly spaced sinusoidal wave with period T$_\mathrm{sine} = 12.90$ h. Only a few days over the entire 5--year interval are shown, in order to see the sinusoidal signal (right) Corresponding Lomb--Scargle Periodogram. $95$% confidence level (based on randomization test) is indicated for the Lomb--Scargle periodogram. It is very low in this case as only signal is given (no noise).
• Figure 2: (left) Semi-regularly spaced sinusoidal wave with period T$_\mathrm{sine} = 12.90$ h. The samples are gathered into $125$ intervals of $8$ hours, spaced by $\mathrm{N} \times 23.93$ hours (with $\mathrm{N} = 1, 2, 3, 4, ...$). (right) Corresponding Lomb--Scargle Periodogram. The $95$% confidence level is also indicated. It is also very low in this case as only a signal is given (no noise).
• Figure 3: Same as Figure \ref{['fig:regularly_spaced_sinus']}, but with all values of the sinusoid at 1, in order to show the effect of the windowing on the Lomb--Scargle periodogram. The $95$% confidence level is also indicated. As all values are saturated to 1, randomly shuffling the values does not change the Lomb--Scargle analysis and the $95$% confidence level is at the values of the highest peaks (due to the sidereal periodicity at $\mathrm{N} \times 23.93$ hours).
• Figure 4: (left panel) Semi-regularly spaced sinusoidal wave with period T$_\mathrm{sine} = 12.90$ h (light blue), normal distribution with a standard deviation $\sigma = K \times \sigma_{sin}$ with $K=15$ (light grey), and combined signal with higher signal to noise ratio (red points). The samples are gathered into $125$ intervals of $8$ hours, spaced by $\mathrm{N} \times 23.93$ hours. (right panel) Corresponding Lomb--Scargle Periodogram. $95$% confidence level is also shown in the Lomb--Scargle periodogram.
• Figure 5: Typical NenuFAR "time (UTC) versus frequency (in MHz) spectrogram" of Jupiter signal. (Top panel) Stokes I with data preprocessing applied (see text). (Bottom panel) Ratio between Stokes $V$ and Stokes $I$ showing the degree of circular polarization.