Optimising the global detection of solar-like oscillations. Tuning the frequency range for asteroseismic detection predictions and searches

Abstract

A well-established method exists for predicting the detectability of solar-like oscillations and has been widely used to support target selection for space-based photometric missions. The method evaluates the probability of an asteroseismic detection from the expected global signal-to-noise ratio (SNR) of the oscillation signal relative to the broadband background from shot noise and granulation. Stellar parameters are used to estimate the oscillation and granulation signals, while instrumental properties and apparent stellar brightness determine the expected shot noise. We investigate whether there is an optimal choice for the frequency range, $W$, over which the global SNR is calculated. The oscillation power is assumed to follow a Gaussian-like envelope with full width at half maximum $Γ_{\rm env}$ centred on the frequency of maximum oscillation power. It has commonly been assumed that $W \simeq 2Γ_{\rm env}$ when predicting detections. We compute numerical predictions of the global SNR and corresponding detection probabilities for a range of stellar and observational parameters, adopting widths $W=αΓ_{\rm env}$ where $α$ is a multiplicative factor. We also examine the impact of this choice on detection yields across a population of targets using bright solar-like oscillators observed by TESS as a representative sample. We find that the commonly adopted value $α\simeq 2$ is suboptimal and that $α\simeq 1.2$ maximises the detection probability. This choice can also significantly affect predicted detection yields for stellar samples. We therefore recommend adopting $W \simeq 1.2Γ_{\rm env}$ both when computing detection probabilities and when searching for oscillations in real data via tests of excess mode power, as it optimises the probability of robust detections.

Figures (3)

• Figure 1: Left: Variation in SNR values with $\alpha$ for different assumed values of $\rm SNR_{tot}(2) \in[0.02,0.10]$ (see plot annotation), with the vertical dotted line marking the standard value of $\alpha=2$ and the vertical dashed line marking the optimal value of $\alpha\simeq1.2$. The corresponding horizontal dotted and dashed lines mark the intersection with the red curve, having $\rm SNR_{tot}(2)=0.050$. Right: Corresponding detection probabilities, $p_{\rm final}$, for $\nu_{\rm max} = 2000\,\rm \mu Hz$ and $T=1$ month.
• Figure 2: Variation in the change $\Delta p_{\rm final}$ between $\alpha=1.2$ and 2.0 with $\nu_{\rm max}$, $p_{\rm final}$ (left column) and $\rm SNR_{\rm tot}(2)$ (right column). The top row shows results for $T=1$ month, and the bottom row shows them for $T=6$ months.
• Figure 3: Results for predictions using the bright TESS stars from the TESS Luminaries Sample Lund2025, assuming a single sector of TESS observations ($T \simeq 1$ month), data with a cadence of $120$ sec, a false-alarm probability of $5\%$, and input values of $T_{\rm eff}$, $\log g$, and $R$ from the Gaia DR3 gspphot.