Enhancing online estimation of CBC parameters with the low-latency MBTA analysis

TL;DR

This work presents a low-latency MBTA-based procedure to online-estimate CBC parameters by using an SNR optimizer that densifies templates around the initial detection in a metric-guided, dimensionless chirp-time space. It combines fast template placement, single-band filtering, and hierarchical processing to deliver online detector-frame and source-frame parameter estimates, along with p_astro classifications and Bayestar-based sky localization. Validation on injections and O4a-events shows modest SNR gains (≈2–3%), reduced sky areas, and broadly consistent parameter posteriors with LVK results, while improving NSBH classification and enabling rapid EM-follow-up guidance. The online implementation (O4c) demonstrates feasible latency, with two iterative passes typically completing within about one minute per event, illustrating practical applicability for real-time multi-messenger campaigns.

Abstract

In this paper, we describe the procedure implemented in the Multi-Band Template Analysis (MBTA) search pipeline to produce online posterior distributions of compact binary coalescence (CBC) gravitational-wave parameters. This procedure relies on an SNR optimizer technique, which consists of filtering dense local template banks. We present how these banks are constructed using information from the initial detection and detail how the results of the filtering are used to estimate source parameters and provide posterior distributions. We demonstrate the performance of our procedure on simulations and compare our source parameter estimates with the results from the first part of the fourth observing run (O4a) recently released by the LIGO-Virgo-KAGRA (LVK) collaboration.

Figures (13)

• Figure 1: Illustration of the SNR optimizer bank construction process for GW230529_181500 Abac_2024. Each blue point represents a template. The purple stars indicate the bank center---i.e. the parameters of the most probable MBTA trigger for this event after accounting for astrophysical priors described in Section \ref{['sec:03']}. The parameters recovered online by the highest $\mathrm{SNR}$ trigger from the MBTA pipeline are highlighted by red triangles. All values corresponding to non-physical parameters are grayed out in the top and middle rows, and excluded from the bottom row. The top row shows the 2D projections of the universal bank in the coordinate system defined by Equation \ref{['A01-eq:new_var']}. In this case, the templates are uniformly distributed in this space. The middle row displays the same samples transformed into dimensionless chirp time coordinates using Equation \ref{['A01-eq:U_to_theta']} (with $\boldsymbol{S}=(1, 1, 2)$). The bottom row presents the final set of templates in a more commonly used parameter space: detector-frame chirp mass $m_{\text{chirp}}\xspace^d$ (in solar masses), symmetric mass ratio $\eta$ and effective spin $\chi_\mathrm{eff}\xspace$.
• Figure 2: Prior probability of the templates used in the first iteration of the SNR optimizer, initiated by the MBTA trigger for GW230529_181500 Abac_2024, used alongside recovered SNRs to infer source parameters. The distributions are shown for the detector-frame chirp mass ($m_{\text{chirp}}\xspace^d$), the symmetric mass ratio ($\eta$), and the effective spin ($\chi_\mathrm{eff}\xspace$). The plots display the marginal 1D (diagonal) and 2D (off-diagonal) distributions. Black lines indicate the distributions of the templates used for filtering. Blue lines and dots show the same distributions reweighted according to Equation \ref{['A02-eq:prior_reweighting']}.
• Figure 3: Visualization of the parameter inference obtained after the second iteration of the MBTA SNR optimizer on data associated with GW230529_181500. The plot shows the 1D (diagonal) and 2D (off-diagonal) marginal posterior distributions for the component masses ($m_1, m_2$) and effective spin ($\chi_\mathrm{eff}$). Red lines indicate the detector-frame parameters of the online most significant MBTA trigger. Black lines give the median (solid) and $90\%$ symmetric credible intervals (dashed) of the source parameters obtained by the MBTA SNR optimizer. Green lines indicate the values inferred by the LVK collaboration in Abac_2024.
• Figure 4: Distribution of the network $\mathrm{SNR}$ ratio between the SNR optimizer output and the initial trigger from the search, evaluated on the two subsets of O4a LVK common injections. The $90\%$ symmetric credible intervals (CI) of the median are estimated using a bootstrap method.
• Figure 5: Distribution of the $90\%$ credible region ratio between the second SNR optimizer iteration highest-SNR trigger and the initial trigger, evaluated on a subset of O4a LVK common injections (pop-1). The $90\%$ symmetric credible intervals (CI) of the median are estimated using a bootstrap method.