Search for Long-Transient Gravitational Waves from Supernova SN2023ixf using GFH-v2 Pipeline

Technical Summary: Search for Long-Transient Gravitational Waves from SN 2023ixf using GFH-v2

Problem and Motivation
Rapidly rotating newborn magnetars formed in core-collapse supernovae (CCSNe) are predicted to emit long-transient continuous gravitational waves (tCWs). Unlike short-duration bursts associated with the collapse itself, these signals arise from the spin-down of a newly formed, non-axisymmetric neutron star or magnetar. The emission is driven by the loss of rotational energy, producing a signal with a frequency that decreases according to a power law and a strain amplitude that decays over time. Detecting these signals would provide direct insight into the birth, internal structure, and spin evolution of neutron stars.

SN 2023ixf, a Type II core-collapse supernova discovered in the nearby galaxy M101 (distance $\approx$ 6.85 Mpc), presents a unique opportunity for such a search. Its proximity and the precise timing of its electromagnetic detection (within a $\sim$2-hour window) allow for a directed search. Crucially, the event's epoch overlaps with LIGO's Engineering Run 15 (ER15), providing a dataset of coincident data from the Hanford (H1) and Livingston (L1) detectors immediately preceding the fourth observing run (O4).

Methodology
The analysis employs the GFH-v2 pipeline, a hierarchical semi-coherent search method based on the Generalized Frequency Hough transform, specifically designed for signals with power-law frequency evolution.

• Signal Model: The search targets a rigid, non-precessing neutron star spinning down primarily via gravitational-wave emission. This corresponds to a braking index $n=5$. The gravitational-wave frequency $f_{gw}$ evolves as $\dot{f}_{gw} = -k f_{gw}^n$, and the strain amplitude $h_0(t)$ decays as $t^{-1/2}$. The search assumes canonical neutron star parameters ($M=1.4 M_\odot$, $R=12$ km) and a fixed sky position based on electromagnetic localization.
• Data: The analysis uses coincident science segments from H1 and L1 during ER15, totaling approximately 48,619 seconds (0.56 days) of data within a conservative on-source window. The data quality was verified to be comparable to O4a sensitivity in the 600–2000 Hz range.
• Search Configuration: The parameter space covers initial frequencies ($f_0$) from 600 to 2000 Hz (step 10 Hz) and ellipticities ($\epsilon$) from $3 \times 10^{-4}$ to $3 \times 10^{-3}$ (divided into five sub-intervals). This results in 700 independent search configurations.
• Pipeline Steps:
  1. Data Preparation: Short Fourier Transform Database (SFDB) data are band-passed and transformed into complex analytic time series.
  2. Peakmap Generation: Local maxima in the equalized power spectrum are selected to form a time-frequency peakmap.
  3. Hough Transform: The peakmap is mapped from $(t, f)$ to transformed coordinates $(t, x)$ where $x = 1/f^{n-1}$. In this space, the power-law evolution becomes linear. A Hough transform maps this to the $(x_0, k)$ parameter space (initial frequency and spin-down parameter).
  4. Candidate Selection: Candidates are identified as bins with excess counts relative to the local background. A critical ratio (CR) is calculated using locally estimated noise statistics (median and dispersion within 10 Hz intervals) to account for non-stationary noise.
  5. Coincidence: Candidates from H1 and L1 are compared in the $(x_0, k)$ parameter space. A pair is considered coincident if the distance metric $d_{coin} \le 2$.
  6. Follow-up: Coincident candidates with $CR \ge 3.5$ undergo a heterodyne follow-up. The expected phase evolution is removed from the data, and the signal is re-analyzed with longer coherence times ($3\times$ and $6\times$ the original) to check for an increase in significance, which would indicate a real astrophysical signal.

Key Results

• Candidate Selection: After applying the coincidence requirement and the CR threshold ($CR \ge 3.5$), 131 coincident candidates remained.
• Follow-up Outcome: None of the 131 candidates exhibited the expected increase in critical ratio during the heterodyne follow-up stages. Consequently, no candidates were promoted to the second follow-up stage.
• Detection: The analysis found no evidence for long-transient gravitational wave emission associated with SN 2023ixf in the ER15 data.
• Upper Limits: Upper limits on the maximum detectable distance were established as a function of initial frequency and ellipticity. For the highest ellipticity interval ($\epsilon \in [2.5 \times 10^{-3}, 3.0 \times 10^{-3}]$), the 90% efficiency upper limits range from approximately 1 to 2.5 Mpc across most of the analyzed frequency band. For lower ellipticities, limits were either not determinable or restricted to smaller distances due to the limited data duration and lower signal-to-noise ratios.

Significance and Claims
The authors state that while the derived upper limits (1–2.5 Mpc) are below the actual distance to M101 ($\approx$ 6.85 Mpc), the work represents the first application of the GFH-v2 pipeline to a directed search for a nearby core-collapse supernova. The study successfully characterizes the pipeline's performance on real, commissioning-era detector data (ER15). It demonstrates that the GFH-v2 method can be effectively deployed to search for tCWs from magnetar-forming events, establishing a framework for future searches using more sensitive datasets from upcoming observing runs (O5) and next-generation detectors. The paper concludes that the search provides a direct sensitivity estimate for long-transient searches in ER15 data and serves as a benchmark for future efforts.