2 like evetns [40]. They show future prospects of detection II. PARAMETER ESTIMATION METHODS of deviations from GR assuming boson stars. We note, however, that the above works do not consider both ef- A. ECO features fects simultaneously but focus on only one of them1. The aim of this work is to give model-independent tests There are several features of ECOs that differ from BH of strong-field gravity regimes from the measurements of (see Refs. [4] for review). In this paper, we focus on the tidal deformability and SIQM via GWs from compact bi- tidal deformability and SIQM. nary inspirals. One motivation to think about BH mim- ickers is to modify the BHs in GR to be compatible with the stringy paradigm as to BH information loss. If we 1. Tidal deformability assume something like a firewall, only the small region near the horizon might be modified, or in other words In the compact binary inspiral, at the leading order, practically only the absorbing boundary condition across the tidally induced quadrupole moment tensor Qij,Tidal the horizon might be modified. This change of boundary is proportional to the companion’s tidal field Eij as condition may also result in the modification of tidal de- Qij,Tidal = −λEij. The information about the EOS (or formability, unless the modification is restricted to a re- structure) can be quantified by the tidal deformability ally tiny region in the vicinity of the horizon, e.g., within parameter λ [6]. The leading order tidal contribution the Planck distance from the horizon. (Thus, we think to the GW phase evolution (relative 5PN order) arises non-trivial tidal deformability will not necessarily imply through the symmetric contribution of tidal deformation, smaller compactness.) the binary tidal deformability [6, 7, 47]

In this paper, we reanalyze the data around six 4 4 16 (m1 + 12m2)m Λ1 + (m2 + 12m1)m Λ2 low-mass events identified as BBH; GW151226, Λ˜ = 1 2 , (1) 13 (m + m )5 GW170608, GW190707 093326 (hereafter 1 2 GW190707), GW190720 00836 (hereafter GW190720), which is a mass-weighted linear combination of the both GW190728 064510 (hereafter GW190728), and component tidal parameters, where m1,2 is the compo- GW190924 021846 (hereafter GW190924), using an nent mass and Λ1,2 is the dimensionless tidal deformabil- 5 inspiral-only waveform model with both tidal and SIQM ity parameter of each object defined as Λ1,2 = λ1,2/m1,2. terms, and present constraints on the binary tidal For the waveform models used in this paper, the tidal deformability and SIQM at the same time. We focus effects to the gravitational-wave phase are dominated on the inspiral regime because post-inspiral regimes of by the symmetric contributions, Λ˜ terms, and the an- binary ECOs are not modeled well. Since the inspiral tisymmetric contributions, δΛ˜ terms, are always sub- regime can be accurately described by post-Newtonian dominant [48, 49]. The tidal deformability can char- (PN) formula [5, 42, 43], we use PN inspiral-only acterize the compact objects. Λ for BHs in classical waveform model. GR vanishes as shown for Schwarzschild BH [50, 51] The remainder of this paper is organized as follows. In and for Kerr BH [52–54]. For NSs, Λ is a few hun- Sec. II, we explain the methods of Bayesian parameter dred, depending on EOS [8, 47, 55–57]. Upper bound on ˜ estimation for GWs including waveform models used to the binary tidal deformability Λ by GW170817 is about analyze. In Sec. III, we present results of our analysis 900 [13, 58] (see also [16, 17] for reanalysis). ECOs differ of GWTC-2 events by using TF2g Tidal SIQM waveform from BHs in classical GR in having nonzero tidal de- model. Section IV is devoted to a summary and conclu- formabilities [38, 39, 59]. It is intriguing that Λ is nega- sion. In Appendix A, we show the results for seven events tive for gravastars [60, 61]. added GW190814 by using the TF2 Tidal SIQM wave- form model. Here, TF2 is the abbreviation of TaylorF2, which is the PN waveform model for point-particle and 2. SIQM spin effects [42, 44, 45] and TF2g is an extended waveform model of TF2 obtained by Taylor-expanding the effective- For a compact object with mass m and the dimen- one-body formula [46]. sionless spin parallel to the orbital angular momentum, χ = S/m2, where S is the magnitude of the spin angular We employ the units c = G = 1, where c and G are momentum of the aligned component, the spin-induced the speed of light and the gravitational constant, respec- quadrupole moment scalar is given by [10] tively. 2 3 QSpin = −(1 + δκ)χ m , (2) where δκ denotes deviations from the Kerr BHs in GR. The symmetric combination of the deviation parameters of the respective objects, δκ1,2, is defined as δκs = (δκ1 + 1 The analysis in Ref. [41] have considered both effects at one time δκ2)/2. The SIQM can also characterize the compact for Fisher information matrix analysis. objects. For Kerr BH, we have δκ = 0 [10]. For spinning 3

NS, we have δκ ∼ 2 − 20 [62–64]. ECOs differ from We use an extended model of TaylorF2 [42, 44, 45], BHs, having δκ 6= 0: δκ ∼ 10 − 150 for spinning boson TF2g, as ΨBBH(f), which consists of point-particle and stars [65–68]. Interestingly, δκ for gravastar can take spin parts. For TaylorF2 the 3.5PN-order formulas is negative values [61]. employed for the point-particle part of the phase as sum- marized in [45, 69]. For TF2g, the phase of the point- particle part is extended to the quasi-5.5PN-order, which B. Waveform models for inspiraling exotic compact is derived by the Taylor expansion of the effective-one- objects body formula taking into account the notion in the test particle limit [46]. We set the uncalculated terms at 4PN- We use the post-Newtonian (PN) waveform model, order and beyond to zero. The added higher PN-order which can accurately describe the early inspiral regime terms enable to reduce the tidal deformability biasing. for compact binary coalescences [5, 42]. The frequency- The waveform models for both BBH and binary ECO domain gravitational waveform for binary ECOs can be used in our parameter estimation analyses assume that written as the spins of component objects are aligned with the or- bital angular momentum, and incorporate 3.5PN-order ˜ iΨBBH(f)+ΨSIQM(f)+ΨTidal(f) hECO(f) = A(f)e , (3) formula in couplings between the orbital angular momen- tum and the component spins [70], 3PN-order formula in where A(f) is the amplitude of the GW signal and the point-mass spin-spin, and self-spin interactions [71, 72] phase which consists of the BBH phase ΨBBH(f) and as summarized in [69]. The different PN-order terms for the additional SIQM effect ΨSIQM(f) and tidal effect spin effects could help to break degeneracy between pa- ΨTidal(f). We consider the amplitude formula up to rameters. the 3PN-order as summarized in [69], where the point We also use PN formula for the tidal effects. We em- particle and the spin effects are included for both BBH ploy the 2.5PN-order (relative 5+2.5PN-order) tidal-part and ECO hypotheses and the leading order term is ap- of the phase. Recently, Henry, Faye and Blanchet [73] −1 det 5/6 −7/6 proximately written as ∼ dL (M ) f where dL have derived the complete form up to 5+2.5PN-order for is the luminosity distance to the source, and Mdet = the mass quadrupole, current quadrupole, and mass oc- 3/5 1/5 (1 + z)(m1m2) /(m1 + m2) is the detector-frame tupole contributing to the GW tidal phase. We rewrite it (redshifted) chirp mass, which gives the leading-order only for the mass quadrupole interactions as a function evolution of the binary amplitude and phase, and z is of the dimensionless tidal deformability of each object 5 2 the source redshift. defined as Λ1,2 = λ1,2/m1,2 by ,

2 3 X  5 Ψ (f) = x5/2 Λ X4 −24(12 − 11X ) − (3179 − 919X − 2286X2 + 260X3 )x HFB 128η A A A 28 A A A A=1 3/2 +24π(12 − 11XA)x 193986935 14415613 57859 209495 965  −5 − X − X2 − X3 + X4 − 4X5 x2 571536 381024 A 378 A 1512 A 54 A A π i + (27719 − 22415X + 7598X2 − 10520X3 )x5/2 , (4) 28 A A A

2/3 where x = [πMtot(1 + z)f] is the dimensionless PN current quadrupole and mass octupole have been de- expansion parameter, Mtot = m1 + m2 is the total mass, rived [73] and they provide the correct coefficient of rel- 2 η = m1m2/(m1 + m2) is the symmetric mass ratio, and ative 5+2.5PN-order term. Although it is not imple- XA = mA/Mtot, A=1,2. Here, we do not ignore the anti- mented in our analysis, the difference between the correct symmetric contribution δΛ˜ terms, while they are always coefficient and the one we used is numerically very small subdominant on the tidal effects to the GW phase, com- and thus should not change much our results. pared with the symmetric contribution Λ˜ terms [48, 49]. The leading order effect due to the SIQM appears as a We set the uncalculated terms at relative 5+2PN-order part of spin-spin interactions in the PN phase at relative to zero. We note that recently the complete GW phases 2 PN order [10], and the leading order additional term up to relative 5+2.5PN-order for the mass quadrupole, for binary ECO is described as 2 2 2 2 75 δκ1m1χ1 + δκ2m2χ2 −1/2 ΨSIQM(f) = x . (5) 64 m1m2 2 We will extend the our analysis by adding the current quadrupole In our analysis, we also incorporate relative 3PN correc- and mass octupole interactions. tions to the GW phase due to the SIQM effects as de- 4

ECO scribed in [33, 71, 74]. The expression for the SIQM parts where BFBBH,i is the Bayes factor for individual events. in terms of δκs and δκa are described in [33], where δκa The one-dimensional posterior for a specific parameter is are the anti-symmetric combination of the component obtained by marginalizing the multi-dimensional poste- SIQMs defined as δκa = (δκ1 − δκ2)/2. rior over the other parameters. In summary, our template models are the TF2g, We compute posterior probability distribution func- TF2g Tidal, TF2g SIQM, and TF2g Tidal SIQM wave- tions (PDFs) by using Bayesian stochastic sampling form models, which are, respectively, the reference BBH based on nested sampling algorithm [81, 82]. Specifi- model, the ones with only the tidal terms, with only cally, we use the parameter estimation software, LALIn- the SIQM terms, and with both the tidal and the SIQM ference [76], which is one of the software of LIGO Al- terms. gorithm Library (LAL) software suite [83]. We select low mass mergers in GWTC-2 [2] which have higher fre- quency cutoff (fhigh & 120 Hz) and larger inspiral signal- C. Bayesian inference to-noise ratios (SNRs) (ρinspiral & 9) (see Table V in the paper on tests of GR by the LVC [3]). We take the low We employ Bayesian inference for GW parameter esti- frequency cutoff flow = 20 Hz for all events but 30 Hz mation and model selection (see Refs. [75, 76] for review) for Hanford data for GW170608 by following the papers between binary ECO and BBH. Given a data d, which which reported the detection [84] and the high frequency contains the signal and the noise, according to Bayes’ the- cutoff fhigh = 150 Hz for GW151226, fhigh = 180 Hz orem, the posterior distribution of the signal parameters for GW170608, fhigh = 160 Hz for GW190707, fhigh = θ that the waveform h˜(θ) depends on is given by 125 Hz for GW190720, fhigh = 160 Hz for GW190728, and fhigh = 175 Hz for GW190924, which are determined L(d|θ)π(θ) p(θ|d) = , (6) to be restricted to the inspiral regime. Z where L(d|θ) is the likelihood function of the data for given parameters θ, π(θ) is the prior distribution for θ, D. Source parameters and their priors and Z is the evidence. By assuming stationarity and Gaussianity for the detector noise, the likelihood function The source parameters and their prior probability dis- is evaluated as, tributions are basically chosen to follow those adopted in the paper on GWTC-2 by the LVC [2] and our recent  hd − h(θ)|d − h(θ)i L(d|θ) ∝ exp − , (7) work for GW analysis of BNSs [16, 17]. We mention the 2 specific choices adopted below. For BBH hypothesis, the parameters are the compo- where the noise-weighted inner product h·|·i is defined by nent masses m1,2, where we assume m1 ≥ m2, the orbit- Z fhigh a˜∗(f)˜b(f) aligned dimensionless spin components of the objects ha|bi := 4Re df , (8) χ1,2, the luminosity distance to the source dL, the bi- f Sn(f) low nary inclination angle θJN, which is the angle between the total angular momentum and the line of sight, the using the noise power spectrum density Sn(f). We use polarization angle ψ, and the coalescence time t , the Sn(f) obtained with BayesLine algorithm [77–79]. The c phase at the coalescence time φ . For binary ECO hy- lower limit of the integration flow is the seismic cut-off c pothesis, we add the binary tidal deformability Λ,˜ and frequency and the higher limit fhigh is the cutoff fre- quency of waveforms. To restrict the analysis to the in- the SIQMs δκ1,2. spiral regimes of the signals, we set the upper frequency We employ a uniform prior on the detector-frame com- ponent masses mdet in the range [1.0, 60.0]M . We as- cutoff fhigh to be referred to the one used in [3, 80]. 1,2 The evidence is obtained as the likelihood marginalized sume a uniform prior on the spin magnitudes χ1,2 in over the prior volume, the range [−0.99, 0.99]. We assume a uniform prior on both the binary tidal deformability Λ˜ and the asym- Z ˜ Z = dθL(d|θ)π(θ). (9) metric contribution δΛ in the range [−3000, 3000] and a uniform prior on the SIQMs for individual objects To perform model selection between the binary ECO and δκ1,2 in the range [−200, 200]. While in the analysis BBH hypotheses, we compute the ratio between two dif- of the paper on tests of GR by the LVC, they restrict ferent evidences, called the Bayes factor, δκa = (δκ1 − δκ2)/2 = 0, implying δκ1 = δκ2 = δκs [3], we do not assume so. ECO ZECO BFBBH = . (10) ZBBH III. RESULTS The combined Bayes factor is defined as

ECO X ECO We analyze the BBH events with the low-mass (or log10 BFBBH,total = log10 BFBBH,i, (11) i long inspiral regime) and higher SNR for inspiral regime 5 among GWTC-2 events [2, 85]. Here, we use the pub- 0.006 lic data on the Gravitational Wave Open Science Cen- TF2g_SIQM ter (https://www.gw-openscience.org) released by the TF2g_Tidal_SIQM LVC. We analyze each event using inspiral-only template 0.005 original prior TF2g Tidal SIQM waveform model. First, we present the results for the largest SNR event GW170608 in detail. Next, we show the results for other events and model 0.004 selection by the Bayes factor combining six events.

0.003

0.0010 PDF TF2g_Tidal TF2g_Tidal_SIQM 0.002 0.0008

0.001

0.0006 0.000 150 100 50 0 50 100 150 PDF s 0.0004

FIG. 2. Marginalized posterior PDFs of δκs for GW170608 using the TF2g SIQM (green, dotted) and 0.0002 TF2g Tidal SIQM (red, solid) waveform models. We set fhigh = 180 Hz. They are weighted by dividing the origi- nal prior: uniform on δκ1,2. The original prior is also shown by solid yellow curve. The peak of δκs is shifted toward zero 0.0000 by adding the tidal terms. 2000 1000 0 1000 2000

FIG. 1. Marginalized posterior PDFs of Λ˜ for low- mass event GW170608 with the TF2g Tidal (blue, dashed) and TF2g Tidal SIQM (red, solid) waveform models. We set fhigh = 180 Hz. Adding the SIQM terms do not affect the constraint on the tidal deformability Λ.˜

TABLE I. The logarithm of the Bayes factors for a sig- nal compared to Gaussian noise log10 BFs/n and SNRs for GW170608 using the TF2g, TF2g Tidal, TF2g SIQM, and TF2g Tidal SIQM waveform models. The tidal and SIQM terms do not affect the Bayes factor and SNR.

log10 BFs/n SNR TF2g (BBH in GR) 71.3 14.7 TF2g Tidal 69.8 14.7 TF2g SIQM 70.4 14.7 TF2g Tidal SIQM 69.3 14.7

FIG. 3. Corner plot on the Λ-˜ δκs plane for GW170608 A. Estimating tidal deformability and SIQMs for using the TF2g Tidal SIQM waveform model by setting fhigh = GW170608 180 Hz. The contours correspond to 50% and 90% credible regions. The constraints show that GW170608 is consistent ˜ We show the results of low mass presumed BBH event with BBH in GR (Λ = δκs = 0). GW170608. We present posteriors of binary tidal de- formability and the SIQM for GW170608. Figure 1 shows 6

det posteriors of Λ˜ for GW170608. We set fhigh = 180 Hz. also biases the estimate of M . Comparing the poste- The posterior distribution of Λ˜ for the TF2g Tidal wave- riors of χeff for the TF2g Tidal (blue dot-dashed) with form model (blue dot-dashed) is consistent with the one TF2g Tidal SIQM (red solid) waveform models, adding for the TF2g Tidal SIQM waveform model (red solid). the SIQM terms reduce the statistical error of χeff . Adding the SIQM terms do not affect the constraint on the tidal deformability Λ.˜ Figure 2 shows posteriors of δκs for GW170608 us- C. Constraints on tidal deformability and SIQMs ing the TF2g SIQM and TF2g Tidal SIQM waveform mod- for GWTC-2 events els. Here, they are weighted by dividing the original prior: uniform on δκ1,2, to effectively impose a uniform We also analyzed other three low mass events: prior on δκ . We find that δκ is poorly constrained s s GW151226 with fhigh = 150 Hz, GW190707 with for GW170608 for both templates, which is consistent fhigh = 160 Hz, GW190720 with fhigh = 125 Hz, with the results shown for the templates with only the GW190728 with fhigh = 160 Hz, and GW190924 with SIQMs in [3, 36]. Equation (2) shows that Q = 0 when i fhigh = 175 Hz, using the TF2g Tidal SIQM waveform χi = 0, independent of the value of δκi, and δκi is not model. We present posteriors of binary tidal deforma- constrained unless at least one of the magnitude of the bility and the SIQMs for GW151226, GW170608, component spin χi exclude zero. It is natural that our GW190707, GW190720, GW190728, and GW190924. constraints on δκs are poorer than LVC’s one since we do The left panel of Fig. 5 shows posteriors of Λ˜ for six events not assume δκa = 0. Comparing the posteriors for the using the TF2g Tidal SIQM waveform model. This figure TF2g SIQM (green, dotted) with TF2g Tidal SIQM red, shows the posterior of δκs for the uniform prior on δκ1,2. solid) waveform models, adding the tidal terms shift the The 90% symmetric credible ranges of Λ˜ are summarized peak of δκs toward zero. ˜ in Table II, which are [−1441, 649] for GW151226, Figure 3 shows corner plot on Λ-δκs plane for [−1265, 565] for GW170608, [−1265, 565] for GW170608 using the TF2g Tidal SIQM waveform model, GW170608, [−590, 1661] for GW190707, [−1445, 1768] which shows that the observed data is consistent with ˜ for GW190720, [−1432, 1078] for GW190728, and BBH in GR (Λ = δκs = 0). This figure shows the poste- [−2041, 1118] for GW190924. rior of δκ for the uniform prior on δκ . We find weak s 1,2 The right panel of Fig. 5 shows the posterior distribu- negative correlation between Λ˜ and δκs. tion on Λ-˜ δκs plane. For all events, there exists nega- Table I shows the logarithm of the Bayes factor for a tive correlation between Λ˜ and δκs. We find that δκs is signal compared to Gaussian noise log10 BFs/n. Adding poorly constrained for all events we analyzed, which is the tidal and SIQM terms do not drastically change the consistent with the results in [3, 36]. We do not show the Bayes factor log BF and SNRs. 10 s/n credible intervals for δκs, since all the posterior PDFs present tails reaching the edge of the prior when they are divided by the original prior (uniform on δκ1,2) as shown B. Biases of binary parameters due to tidal and in Fig. 2 for GW170608 and in [3, 36] for other cases. SIQM terms for GW170608 It is natural that our constraints on δκs are poorer than LVC’s ones, since we do not assume δκa = 0, while LVC Figure 4 shows the marginalized posterior PDFs of did. parameters other than the tidal deformability and the SIQM obtained by using the TF2g model (orange, solid), TF2g Tidal (blue, dot-dashed), TF2g SIQM (green, dot- D. Model selection between ECO and BBH ted), and TF2g Tidal SIQM (red, dashed) waveform mod- els. We present the distribution of the mass parameters: the mass ratio q = m /m , the primary mass m , the Bayes factor between the binary ECO and BBH wave- 2 1 1 form model, BFECO, quantifies the statistical significance secondary mass m2, the chirp mass M, the detector- BBH det of one hypothesis over the other. The binary ECO hy- frame chirp mass M and the total mass Mtot, the pothesis (the TF2g Tidal SIQM waveform model) is dis- luminosity distance dL, the binary inclination θJN, and favored compared to BBH (the TF2g waveform model) the effective inspiral spin χeff , which gives the leading- 3 order spin effect on the binary phase evolution [86, 87] for all events as shown Table II . The combined Bayes factor of six events estimated by defined as χeff = (m1χ1 +m2χ2)/Mtot. The source-frame ECO chirp mass is derived by assuming the Hubble constant using Eq. (11), log10 BFBBH,total = −10.4, shows that −1 −1 BBH hypothesis is preferred to the binary ECO hypoth- H0 = 69 km s Mpc (a default value in LALInference esis. adopted from Planck 2013 results [88]). q, m1, m2, M, Mtot, dL, and θJN are not affected by adding the tidal and SIQM terms to TF2g baseline model. Adding either the tidal or the SIQM terms shift the peak of χeff toward 3 zero, as also shown in Fig. 5 of [36] for SIQM, which log10 BF > 1.5 is often interpreted as a strong preference for one model over another, and log BF > 2 as decisive evidence [89]. is because measurement of χeff becomes difficult. This 10 7

0.3 TF2g fhigh = 180 Hz 0.3 1.5 TF2g_Tidal 0.2 TF2g_SIQM TF2g_Tidal_SIQM 0.2 1.0 PDF 0.1 0.5 0.1

0.0 0.0 0.0 0.2 0.6 1.0 10 15 20 4 6 8 10 q m1(M ) m2(M )

10.0 3 0.6 7.5

2 0.4 5.0 PDF

1 2.5 0.2

0 0.0 0.0 7.5 8.0 8.5 8.25 8.50 8.75 20 25 det (M ) (M ) Mtot(M )

0.006 1.00 6

0.75 0.004 4 0.50 PDF 0.002 2 0.25

0.000 0.00 0 100 300 500 0 90 180 0.5 0.0 0.5 dL(Mpc) JN(deg) eff

FIG. 4. Marginalized posterior PDFs of various parameters for GW170608 derived by using the TF2g (orange, solid), TF2g Tidal (blue, dot-dashed), TF2g SIQM (green, dotted), and TF2g Tidal SIQM (red, dashed) waveform models, by setting fhigh = 180 Hz. The top-left, top-middle, top-right, middle-left, center, middle-right, bottom-left, bottom-middle, and bottom- right panels show the mass ratio q, the primary mass m1, the secondary mass m2, the source-frame chirp mass M, the det detector-frame chirp mass M , the total mass Mtot, the luminosity distance to the source dL, the inclination angle θJN, and the effective spin parameter χeff , respectively. Adding either the tidal or the SIQM terms shift peak of χeff toward zero, due to difficulty to measure χeff . 8

0.0008 GW151226 GW170608 GW190707 0.0007 GW190720 GW190728 GW190924 0.0006

0.0005

PDF 0.0004

0.0003

0.0002

0.0001

0.0000 3000 2000 1000 0 1000 2000 3000

FIG. 5. Left: The same as Fig. 1 but for GW151226 (red dashed), GW170608 (green solid), GW190707 (blue dot- dashed), GW190720 (cyan dash-dotdotted), GW190728 (magenta loosely dashed), GW190924 (orange dotted), using the TF2g Tidal SIQM waveform model by setting fhigh = 150, 180, 160, 125, 160, and 175 Hz, respectively. These constraints show that all events are consistent with BBH in GR (Λ˜ = 0). Right: The same as Fig. 3 but for GW151226 (red), GW170608 (green), GW190707 (blue), GW190720 (cyan), GW190728 (magenta), GW190924 (orange) using the TF2g Tidal SIQM wave- form model for fhigh = 150, 180, 160, 125, 160, and 175 Hz, respectively. These constraints show that all events are consistent with BBH in GR (Λ˜ = δκs = 0). 9

hospitality of Chris’s group, in particular Khun Sang ˜ TABLE II. The 90% symmetric credible ranges of Λ, the log- Phukon, Archisman Gosh, and Tim Dietrich, during his arithm of the Bayes factors between binary ECO and BBH, stay at Nikhef. We thank Aditya Vijaykumar, Nathan K. i.e., log BFECO, and SNRs for GW151226, GW170608, 10 BBH Johnson-McDaniel, Rahul Kashyap, Arunava Mukher- GW190707, GW190720, GW190728, GW190924, and the re- sult combining six events using the TF2g Tidal SIQM wave- jee, and Parameswaran Ajith for sharing and discussing their results on a similar study [91]. We would like form model with respective fhigh. For individual events, ECO to thank Soichiro Morisaki, Kyohei Kawaguchi, and log10 BFBBH are negative, thus favoring the BBH in GR compared to binary ECO. For the combined case, Hideyuki Tagoshi for fruitful discussions. We thank ECO log10 BFBBH,total is also negative, thus disfavoring binary Quentin Henry and Luc Blanchet for pointing out the ECO. complete GW tidal phase they have derived and con- firming the correction that the expression that we have ˜ ECO Event fhigh [Hz] Λ log10 BFBBH SNR rewritten with their value for the mass quadrupole GW151226 150 [−1441, 649] -0.45 10.7 part as a function of Λ1,2, especially the 2 and 2.5 GW170608 180 [−1265, 565] -2.1 14.7 PN-order terms. T. Narikawa was supported in part GW190707 160 [−590, 1661] -2.1 11.2 by a Grant-in-Aid for JSPS Research Fellows. This GW190720 125 [−1445, 1768] -1.8 9.3 work is supported by Japanese Society for the Pro- GW190728 160 [−1432, 1078] -2.0 12.1 motion of Science (JSPS) KAKENHI Grant Numbers GW190924 175 [−2041, 1118] -2.0 11.4 JP21K03548, JP17H06361, JP17H06358, JP17H06357, Combined - - -10.4 - and JP20K03928. We would also like to thank Comput- ing Infrastructure ORION in Osaka City University. We are also grateful to the LIGO-Virgo collaboration for the public release of gravitational-wave data of GW151226, IV. SUMMARY AND CONCLUSION GW170608, GW190707 093326, and GW190924 021846. This research has made use of data, software, and web We implemented the tidal and SIQM terms in tools obtained from the Gravitational Wave Open Sci- the aligned-spin PN inspiral waveform model TF2g, ence Center (https://www.gw-openscience.org), a service which we call TF2g Tidal SIQM. We analyzed six of LIGO Laboratory, the LIGO Scientific Collaboration low-mass events GW151226, GW170608, GW190707, and the Virgo Collaboration. LIGO is funded by the GW190720, GW190728, and GW190924 using the U. S. National Science Foundation. Virgo is funded by TF2g Tidal SIQM waveform model as templates. The ob- the French Centre National de la Recherche Scientifique tained results are the first constraints on the tidal de- (CNRS), the Italian Istituto Nazionale di Fisica Nucle- formability of the events classified as BBH in GWTC- are (INFN), and the Dutch Nikhef, with contributions by 2 events, motivated by ECO hypotheses. We found Polish and Hungarian institutes. that all events that we have analyzed are consistent with BBH mergers in GR. The logarithmic Bayes fac- ECO tor log10 BFBBH for individual events are less than -1.5 Appendix A: Results by using TF2 waveform model except for GW151226, thus favoring the BBH in GR com- pared to binary ECO by Bayesian model selection. The While we show the results by using combined logarithmic Bayes factor between binary ECO TF2g Tidal SIQM waveform model in Sec. III, we and BBH is log BFECO = −10.1, which means that 10 BBH,total show the results for seven events added GW190814 the ECO or non-GR model (with Tidal and SIQM param- by using TF2 Tidal SIQM waveform model in this Ap- eters) is also disfavored compared to BBH in GR. pendix. For highly unequal mass ratio events GW190814, In this paper, we used the inspiral-only waveform TF2g waveform model do not work well due to setting model as templates because there is no robust predic- the uncalculated terms at high PN-order to zero. We tion waveform of the merger-ringdown regimes of binary take the low frequency cutoff flow = 20 Hz for Hanford ECO merger. It might be interesting to use a toy model data and 30 Hz for Livingston data for GW190814 by for post-contact regimes of binary ECOs, which has been following the papers which reported the detection [90] recently derived [31]. Also, GW echoes could be used to and the high frequency cutoff f = 140 Hz. examine post-merger regime. Such extensions including high We present posteriors of binary tidal deformability the waveform after the inspiral regime would allow us to and the SIQMs for GW151226, GW170608, GW190707, analyze heavy-mass events and to put a different type of GW190720, GW190728, GW190814 and GW190924 by constraints on ECO hypothesis. using the TF2 Tidal SIQM waveform model. The left panel of Fig. 6 shows posteriors of Λ˜ for seven events using the TF2 Tidal SIQM wave- ACKNOWLEDGMENT form model. The right panel of Fig. 6 shows the posterior distribution on Λ-˜ δκs plane. The 90% T. Narikawa thanks Chris van den Broeck and Anu- symmetric credible ranges of Λ˜ are summarized in radha Samajdar for useful discussions, and he also thanks Table III, which are [−1656, 590] for GW151226, 10

[−1299, 520] for GW170608, [−791, 1487] for GW190707, [−1523, 1436] for GW190720, [−1445, 846] TABLE III. The same as Table II but for the TF2 Tidal SIQM waveform model and GW190814 is added. for GW190728, [−786, 1272] for GW190814, and For both the individual events and the combined case, the log [−2093, 939] for GW190924. Except for GW190814, ECO Bayes factor log10 BFBBH,total is negative, thus disfavoring bi- systematic uncertainty between TF2g Tidal SIQM and nary ECO. TF2 Tidal SIQM results remains subdominant to statis- tical uncertainty. ˜ ECO Event fhigh [Hz] Λ log10 BFBBH SNR GW151226 150 [−1656, 590] -0.66 10.7 GW170608 180 [−1299, 520] -1.8 14.7 We show the Bayes factor between the binary ECO GW190707 160 [−791, 1487] -2.0 11.2 ECO and BBH waveform model, BFBBH in Table III. The GW190720 125 [−1523, 1436] -1.3 9.3 binary ECO hypothesis (the TF2 Tidal SIQM waveform GW190728 160 [−1445, 846] -1.9 12.1 model) is disfavored compared to BBH (the TF2g wave- GW190814 140 [−786, 1272] -4.0 22.0 form model) for all events. The combined Bayes factor GW190924 175 [−2093, 939] -2.2 11.4 ECO Combined - - -14.0 - of seven events are log10 BFBBH,total = −14.0 shows that BBH hypothesis is preferred to the binary ECO hypoth- esis as shown for TF2g Tidal SIQM.

[1] B. P. Abbott et al. [LIGO Scientific and Virgo], “GWTC- [12] K. Yagi and N. Yunes, “I-Love-Q Relations in Neutron 1: A Gravitational-Wave Transient Catalog of Compact Stars and their Applications to Astrophysics, Gravita- Binary Mergers Observed by LIGO and Virgo during tional Waves and Fundamental Physics,” Phys. Rev. D the First and Second Observing Runs,” Phys. Rev. X 88, 023009 (2013) [arXiv:1303.1528 [gr-qc]]. 9, 031040 (2019) [arXiv:1811.12907 [astro-ph.HE]]. [13] B. P. Abbott et al. [LIGO Scientific and Virgo Col- [2] R. Abbott et al. [LIGO Scientific and Virgo], “GWTC- laborations], “Properties of the binary neutron star 2: Compact Binary Coalescences Observed by LIGO and merger GW170817,” Phys. Rev. X 9, 011001 (2019) Virgo During the First Half of the Third Observing Run,” [arXiv:1805.11579 [gr-qc]]. [arXiv:2010.14527 [gr-qc]]. [14] B. P. Abbott et al. [LIGO Scientific and Virgo], [3] R. Abbott et al. [LIGO Scientific and Virgo], “Tests of “GW170817: Measurements of neutron star radii and General Relativity with Binary Black Holes from the equation of state,” Phys. Rev. Lett. 121, 161101 (2018) second LIGO-Virgo Gravitational-Wave Transient Cat- [arXiv:1805.11581 [gr-qc]]. alog,” [arXiv:2010.14529 [gr-qc]]. [15] B. P. Abbott et al. [LIGO Scientific and Virgo], [4] V. Cardoso and P. Pani, “Testing the nature of dark com- “GW190425: Observation of a Compact Binary Coales- pact objects: a status report,” Living Rev. Rel. 22, 4 cence with Total Mass ∼ 3.4M ,” Astrophys. J. Lett. (2019) [arXiv:1904.05363 [gr-qc]]. 892, L3 (2020) [arXiv:2001.01761 [astro-ph.HE]]. [5] E. Poisson and C. Will, ”Gravity: Newtonian, Post- [16] T. Narikawa, N. Uchikata, K. Kawaguchi, K. Kiuchi, Newtonian, Relativistic” (Cambridge University Press, K. Kyutoku, M. Shibata and H. Tagoshi, “Discrepancy in Cambridge, England, 2014). tidal deformability of GW170817 between the Advanced [6] E. E. Flanagan and T. Hinderer, “Constraining neutron LIGO twin detectors,” Phys. Rev. Research. 1, 033055 star tidal Love numbers with gravitational wave detec- (2019) [arXiv:1812.06100 [astro-ph.HE]]. tors,” Phys. Rev. D 77, 021502 (2008) [arXiv:0709.1915 [17] T. Narikawa, N. Uchikata, K. Kawaguchi, K. Kiuchi, [astro-ph]]. K. Kyutoku, M. Shibata and H. Tagoshi, “Reanal- [7] T. Hinderer, “Tidal Love numbers of neutron stars,” As- ysis of the binary neutron star mergers GW170817 trophys. J. 677, 1216 (2008) [arXiv:0711.2420 [astro-ph]]. and GW190425 using numerical-relativity calibrated [8] T. Damour, A. Nagar and L. Villain, “Measurability of waveform models,” Phys. Rev. Res. 2, 043039 (2020) the tidal polarizability of neutron stars in late-inspiral [arXiv:1910.08971 [gr-qc]]. gravitational-wave signals”, Phys. Rev. D 85, 123007 [18] J. B. Hartle, “Tidal Friction in Slowly Rotating Black (2012) [arXiv:1203.4352 [gr-qc]]. Holes,” Phys. Rev. D 8, 1010-1024 (1973) [9] M. Agathos, J. Meidam, W. Del Pozzo, T. G. F. Li, [19] S. A. Hughes, “Evolution of circular, nonequatorial orbits M. Tompitak, J. Veitch, S. Vitale and C. Van Den of Kerr black holes due to gravitational wave emission. Broeck, “Constraining the neutron star equation of state II. Inspiral trajectories and gravitational wave forms,” with gravitational wave signals from coalescing binary Phys. Rev. D 64, 064004 (2001) [erratum: Phys. Rev. D neutron stars,” Phys. Rev. D 92, no.2, 023012 (2015) 88, 109902 (2013)] [arXiv:gr-qc/0104041 [gr-qc]]. [arXiv:1503.05405 [gr-qc]]. [20] A. Maselli, P. Pani, V. Cardoso, T. Abdelsalhin, [10] E. Poisson, “Gravitational waves from inspiraling com- L. Gualtieri and V. Ferrari, “Probing Planckian correc- pact binaries: The Quadrupole moment term,” Phys. tions at the horizon scale with LISA binaries,” Phys. Rev. Rev. D 57, 5287-5290 (1998) [arXiv:gr-qc/9709032 [gr- Lett. 120, 081101 (2018) [arXiv:1703.10612 [gr-qc]]. qc]]. [21] S. Isoyama and H. Nakano, “Post-Newtonian templates [11] K. Yagi and N. Yunes, “I-Love-Q,” Science 341, 365-368 for binary black-hole inspirals: the effect of the hori- (2013) [arXiv:1302.4499 [gr-qc]]. zon fluxes and the secular change in the black-hole 11

0.0008 GW190814 GW151226 GW170608 0.0007 GW190707 GW190720 GW190728 0.0006 GW190814 GW190924 0.0005

PDF 0.0004

0.0003

0.0002

0.0001

0.0000 3000 2000 1000 0 1000 2000 3000

FIG. 6. Left: The same as the left panel of Fig. 5 but for the TF2 Tidal SIQM waveform model and GW190814 (purple loosely dotted) is added. These constraints show that all events are consistent with BBH in GR (Λ˜ = 0). Right: The same as the right panel of Fig. 5 but for the TF2 Tidal SIQM waveform model and GW190814 (purple) is added. These constraints show that all events are consistent with BBH in GR (Λ˜ = δκs = 0).

masses and spins,” Class. Quant. Grav. 35, 024001 (2018) Rev. D 97, 124037 (2018) [arXiv:1712.09966 [gr-qc]]. [arXiv:1705.03869 [gr-qc]]. [30] N. Uchikata, H. Nakano, T. Narikawa, N. Sago, [22] S. Datta, R. Brito, S. Bose, P. Pani and S. A. Hughes, H. Tagoshi and T. Tanaka, “Searching for black hole “Tidal heating as a discriminator for horizons in extreme echoes from the LIGO-Virgo Catalog GWTC-1,” Phys. mass ratio inspirals,” Phys. Rev. D 101, 044004 (2020) Rev. D 100, 062006 (2019) [arXiv:1906.00838 [gr-qc]]. [arXiv:1910.07841 [gr-qc]]. [31] A. Toubiana, S. Babak, E. Barausse and L. Lehner, [23] S. Datta and S. Bose, “Probing the nature of cen- “Modeling gravitational waves from exotic com- tral objects in extreme-mass-ratio inspirals with grav- pact objects,” Phys. Rev. D 103, 064042 (2021) itational waves,” Phys. Rev. D 99, 084001 (2019) [arXiv:2011.12122 [gr-qc]]. [arXiv:1902.01723 [gr-qc]]. [32] S. Kastha, A. Gupta, K. G. Arun, B. S. Sathyaprakash [24] S. Datta, K. S. Phukon and S. Bose, “Recognizing black and C. Van Den Broeck, “Testing the multipole struc- holes in gravitational-wave observations: Telling apart ture of compact binaries using gravitational wave impostors in mass-gap binaries,” [arXiv:2004.05974 [gr- observations,” Phys. Rev. D 98, 124033 (2018) qc]]. [arXiv:1809.10465 [gr-qc]]. [25] N. Sago and T. Tanaka, “Oscillations in the EMRI grav- [33] N. V. Krishnendu, K. G. Arun and C. K. Mishra, itational wave phase correction as a probe of reflective “Testing the binary black hole nature of a compact bi- boundary of the central black hole,” [arXiv:2106.07123 nary coalescence,” Phys. Rev. Lett. 119, 091101 (2017) [gr-qc]]. [arXiv:1701.06318 [gr-qc]]. [26] K. Fransen, G. Koekoek, R. Tielemans and B. Vercnocke, [34] N. V. Krishnendu, C. K. Mishra and K. G. Arun, “Spin- “Modeling and detecting resonant tides of exotic compact induced deformations and tests of binary black hole na- objects,” [arXiv:2005.12286 [gr-qc]]. ture using third-generation detectors,” Phys. Rev. D 99, [27] Y. Asali, P. T. H. Pang, A. Samajdar and C. Van Den 064008 (2019) [arXiv:1811.00317 [gr-qc]]. Broeck, “Probing resonant excitations in exotic com- [35] N. V. Krishnendu and A. B. Yelikar, “Testing the pact objects via gravitational waves,” Phys. Rev. D 102, Kerr nature of supermassive and intermediate-mass black 024016 (2020) [arXiv:2004.05128 [gr-qc]]. hole binaries using spin-induced multipole moment mea- [28] J. Abedi, H. Dykaar and N. Afshordi, “Echoes from the surements,” Class. Quant. Grav. 37, 205019 (2020) Abyss: Tentative evidence for Planck-scale structure at [arXiv:1904.12712 [gr-qc]]. black hole horizons,” Phys. Rev. D 96, 082004 (2017) [36] N. V. Krishnendu, M. Saleem, A. Samajdar, K. G. Arun, [arXiv:1612.00266 [gr-qc]]. W. Del Pozzo and C. K. Mishra, “Constraints on [29] J. Westerweck, A. Nielsen, O. Fischer-Birnholtz, the binary black hole nature of GW151226 and M. Cabero, C. Capano, T. Dent, B. Krishnan, GW170608 from the measurement of spin-induced G. Meadors and A. H. Nitz, “Low significance of evidence quadrupole moments,” Phys. Rev. D 100, 104019 (2019) for black hole echoes in gravitational wave data,” Phys. [arXiv:1908.02247 [gr-qc]]. 12

[37] M. Wade, J. D. E. Creighton, E. Ochsner and qc]]. A. B. Nielsen, “Advanced LIGO’s ability to detect ap- [52] E. Poisson, “Tidal deformation of a slowly rotating parent violations of the cosmic censorship conjecture black hole,” Phys. Rev. D 91, no.4, 044004 (2015) and the no-hair theorem through compact binary coa- [arXiv:1411.4711 [gr-qc]]. lescence detections,” Phys. Rev. D 88, 083002 (2013) [53] P. Pani, L. Gualtieri, A. Maselli and V. Ferrari, “Tidal [arXiv:1306.3901 [gr-qc]]. deformations of a spinning compact object,” Phys. Rev. [38] V. Cardoso, E. Franzin, A. Maselli, P. Pani and G. Ra- D 92, 024010 (2015) [arXiv:1503.07365 [gr-qc]]. poso, “Testing strong-field gravity with tidal Love num- [54] P. Landry and E. Poisson, “Tidal deformation of a slowly bers,” Phys. Rev. D 95, 084014 (2017) [arXiv:1701.01116 rotating material body. External metric,” Phys. Rev. D [gr-qc]]. 91, 104018 (2015) [arXiv:1503.07366 [gr-qc]]. [39] N. Sennett, T. Hinderer, J. Steinhoff, A. Buonanno and [55] T. Hinderer, B. D. Lackey, R. N. Lang and J. S. Read, S. Ossokine, “Distinguishing Boson Stars from Black “Tidal deformability of neutron stars with realistic equa- Holes and Neutron Stars from Tidal Interactions in Inspi- tions of state and their gravitational wave signatures raling Binary Systems,” Phys. Rev. D 96, 024002 (2017) in binary inspiral,” Phys. Rev. D 81, 123016 (2010) [arXiv:1704.08651 [gr-qc]]. [arXiv:0911.3535 [astro-ph.HE]]. [40] N. K. Johnson-Mcdaniel, A. Mukherjee, R. Kashyap, [56] S. Postnikov, M. Prakash and J. M. Lattimer, “Tidal P. Ajith, W. Del Pozzo and S. Vitale, “Constraining black Love Numbers of Neutron and Self-Bound Quark Stars,” hole mimickers with gravitational wave observations,” Phys. Rev. D 82, 024016 (2010) [arXiv:1004.5098 [astro- Phys. Rev. D 102, 123010 (2020) [arXiv:1804.08026 [gr- ph.SR]]. qc]]. [57] J. M. Lattimer and M. Prakash, “The Equation of State [41] C. Pacilio, M. Vaglio, A. Maselli and P. Pani, of Hot, Dense Matter and Neutron Stars,” Phys. Rept. “Gravitational-wave detectors as particle-physics labora- 621, 127-164 (2016) [arXiv:1512.07820 [astro-ph.SR]]. tories: Constraining scalar interactions with a coherent [58] B. P. Abbott et al. [LIGO Scientific and Virgo Collabora- inspiral model of boson-star binaries,” Phys. Rev. D 102, tions], “GW170817: Observation of Gravitational Waves 083002 (2020) [arXiv:2007.05264 [gr-qc]]. from a Binary Neutron Star Inspiral”, Phys. Rev. Lett. [42] L. Blanchet, Gravitational Radiation from Post- 119, 161101 (2017) [arXiv:1710.05832 [gr-qc]]. Newtonian Sources and Inspiralling Compact Binaries, [59] V. Cardoso, M. Kimura, A. Maselli and L. Senatore, Living Rev. Rel. 17, 2 (2014) [arXiv:1310.1528 [gr-qc]]. “Black Holes in an Effective Field Theory Extension of [43] S. Isoyama, R. Sturani and H. Nakano, “Post-Newtonian General Relativity,” Phys. Rev. Lett. 121, 251105 (2018) templates for gravitational waves from compact binary [arXiv:1808.08962 [gr-qc]]. inspirals,” [arXiv:2012.01350 [gr-qc]]. [60] P. Pani, “I-Love-Q relations for gravastars and the ap- [44] S. V. Dhurandhar and B. S. Sathyaprakash, “Choice of proach to the black-hole limit,” Phys. Rev. D 92, no.12, filters for the detection of gravitational waves from co- 124030 (2015) [erratum: Phys. Rev. D 95, 049902 (2017)] alescing binaries. 2. Detection in colored noise,” Phys. [arXiv:1506.06050 [gr-qc]]. Rev. D 49, 1707-1722 (1994) [61] N. Uchikata, S. Yoshida and P. Pani, “Tidal deforma- [45] A. Buonanno, B. Iyer, E. Ochsner, Y. Pan and bility and I-Love-Q relations for gravastars with poly- B. S. Sathyaprakash, “Comparison of post-Newtonian tropic thin shells,” Phys. Rev. D 94, 064015 (2016) templates for compact binary inspiral signals in [arXiv:1607.03593 [gr-qc]]. gravitational-wave detectors”, Phys. Rev. D 80, 084043 [62] G. Pappas and T. A. Apostolatos, “Multipole Moments (2009) [arXiv:0907.0700 [gr-qc]]. of numerical spacetimes,” [arXiv:1211.6299 [gr-qc]]. [46] F. Messina, R. Dudi, A. Nagar and S. Bernuzzi, [63] G. Pappas and T. A. Apostolatos, “Revising the “Quasi-5.5PN TaylorF2 approximant for compact bina- multipole moments of numerical spacetimes, and its ries: point-mass phasing and impact on the tidal po- consequences,” Phys. Rev. Lett. 108, 231104 (2012) larizability inference,” Phys. Rev. D 99, 124051 (2019) [arXiv:1201.6067 [gr-qc]]. [arXiv:1904.09558 [gr-qc]]. [64] I. Harry and T. Hinderer, “Observing and measuring the [47] J. Vines, E. E. Flanagan and T. Hinderer, “Post-1- neutron-star equation-of-state in spinning binary neutron Newtonian tidal effects in the gravitational waveform star systems,” Class. Quant. Grav. 35, 145010 (2018) from binary inspirals,” Phys. Rev. D 83, 084051 (2011) [arXiv:1801.09972 [gr-qc]]. [arXiv:1101.1673 [gr-qc]]. [65] F. D. Ryan, “Spinning boson stars with large selfinterac- [48] M. Favata, “Systematic parameter errors in inspiraling tion,” Phys. Rev. D 55, 6081-6091 (1997) neutron star binaries,” Phys. Rev. Lett. 112, 101101 [66] F. D. Ryan, “Accuracy of estimating the multipole mo- (2014) [arXiv:1310.8288 [gr-qc]]. ments of a massive body from the gravitational waves of [49] L. Wade, J. D. E. Creighton, E. Ochsner, B. D. Lackey, a binary inspiral,” Phys. Rev. D 56, 1845-1855 (1997) B. F. Farr, T. B. Littenberg and V. Raymond, “System- [67] C. A. R. Herdeiro and E. Radu, “Kerr black holes atic and statistical errors in a bayesian approach to the with scalar hair,” Phys. Rev. Lett. 112, 221101 (2014) estimation of the neutron-star equation of state using ad- [arXiv:1403.2757 [gr-qc]]. vanced gravitational wave detectors”, Phys. Rev. D 89, [68] D. Baumann, H. S. Chia and R. A. Porto, “Probing Ul- 103012 (2014) [arXiv:1402.5156 [gr-qc]]. tralight Bosons with Binary Black Holes,” Phys. Rev. D [50] T. Binnington and E. Poisson, “Relativistic theory of 99, 044001 (2019) [arXiv:1804.03208 [gr-qc]]. tidal Love numbers,” Phys. Rev. D 80, 084018 (2009) [69] S. Khan, S. Husa, M. Hannam, F. Ohme, M. P¨urrer, [arXiv:0906.1366 [gr-qc]]. X. Jim´enez Forteza and A. Boh´e, “Frequency-domain [51] T. Damour and A. Nagar, “Effective One Body descrip- gravitational waves from nonprecessing black-hole bi- tion of tidal effects in inspiralling compact binaries,” naries. II. A phenomenological model for the ad- Phys. Rev. D 81, 084016 (2010) [arXiv:0911.5041 [gr- vanced detector era”, Phys. Rev. D 93, 044007 (2016) 13

[arXiv:1508.07253 [gr-qc]]. binary mergers,” Phys. Rev. D 100, no.10, 104004 (2019) [70] A. Bohe, S. Marsat and L. Blanchet, “Next-to-next-to- [arXiv:1907.06540 [gr-qc]]. leading order spin-orbit effects in the gravitational wave [80] N. Uchikata and T. Narikawa, “Prospects for estimat- flux and orbital phasing of compact binaries”, Class. ing parameters from gravitational waves of superspinar Quant. Grav. 30, 135009 (2013) [arXiv:1303.7412 [gr- binaries,” [arXiv:2104.12968 [gr-qc]]. qc]]. [81] J. Skilling, Bayesian Analysis 1, 833 (2006). [71] K. G. Arun, A. Buonanno, G. Faye and E. Ochsner, [82] J. Veitch and A. Vecchio, “Bayesian coherent analysis of “Higher-order spin effects in the amplitude and phase of in-spiral gravitational wave signals with a detector net- gravitational waveforms emitted by inspiraling compact work”, Phys. Rev. D 81, 062003 (2010) [arXiv:0911.3820 binaries: Ready-to-use gravitational waveforms”, Phys. [astro-ph.CO]]. Rev. D 79, 104023 (2009) Erratum: [Phys. Rev. D 84, [83] LIGO Scientific Collaboration, LIGO Algorithm Li- 049901 (2011)] [arXiv:0810.5336 [gr-qc]]. brary - LALSuite, Free Software (GPL), 2018; [72] B. Mikoczi, M. Vasuth and L. A. Gergely, “Self- https://doi.org/10.7935/GT1W-FZ16. interaction spin effects in inspiralling compact binaries”, [84] B. P. Abbott et al. [LIGO Scientific and Virgo], Phys. Rev. D 71, 124043 (2005) [astro-ph/0504538]. “GW170608: Observation of a 19-solar-mass Binary [73] Q. Henry, G. Faye and L. Blanchet, “Tidal effects in Black Hole Coalescence,” Astrophys. J. Lett. 851, L35 the gravitational-wave phase evolution of compact binary (2017) [arXiv:1711.05578 [astro-ph.HE]]. systems to next-to-next-to-leading post-Newtonian or- [85] LIGO Scientific and Virgo Collaborations, 2020, der,” Phys. Rev. D 102, 044033 (2020) [arXiv:2005.13367 https://doi.org/10.7935/99gf-ax93. [gr-qc]]. [86] P. Ajith, M. Hannam, S. Husa, Y. Chen, B. Br¨ugmann, [74] C. K. Mishra, A. Kela, K. G. Arun and G. Faye, Ready- N. Dorband, D. M¨uller,F. Ohme, D. Pollney, C. Reisswig to-use post-Newtonian gravitational waveforms for bi- et al., “Inspiral-merger-ringdown waveforms for black- nary black holes with nonprecessing spins: An update, hole binaries with non-precessing spins,” Phys. Rev. Lett. Phys. Rev. D 93, no. 8, 084054 (2016) [arXiv:1601.05588 106, 241101 (2011) [arXiv:0909.2867 [gr-qc]]. [gr-qc]]. [87] L. Santamaria, F. Ohme, P. Ajith, B. Bruegmann, [75] E. Thrane and C. Talbot, “An introduction to Bayesian N. Dorband, M. Hannam, S. Husa, P. Mosta, D. Poll- inference in gravitational-wave astronomy: parameter ney and C. Reisswig, et al. “Matching post-Newtonian estimation, model selection, and hierarchical models,” and numerical relativity waveforms: systematic errors Publ. Astron. Soc. Austral. 36, e010 (2019) [erra- and a new phenomenological model for non-precessing tum: Publ. Astron. Soc. Austral. 37, e036 (2020)] black hole binaries,” Phys. Rev. D 82, 064016 (2010) [arXiv:1809.02293 [astro-ph.IM]]. [arXiv:1005.3306 [gr-qc]]. [76] J. Veitch et al., “Parameter estimation for compact bina- [88] Planck Collaboration: P. A. R. Ade, N. Aghanim, ries with ground-based gravitational-wave observations M. I. R. Alves, C. Armitage-Caplan, M. Arnaud, M. Ash- using the LALInference software library”, Phys. Rev. D down, F. Atrio-Barandela, J. Aumont, H. Aussel, C. Bac- 91, 042003 (2015) [arXiv:1409.7215 [gr-qc]]. cigalupi et al., “Planck 2013 results. I. Overview of prod- [77] N. J. Cornish and T. B. Littenberg, “BayesWave: ucts and scientific results,” Astron. Astrophys. 571, A1 Bayesian Inference for Gravitational Wave Bursts and (2014) [arXiv:1303.5062 [astro-ph.CO]]. Instrument Glitches,” Class. Quant. Grav. 32, no.13, [89] H. Jeffreys, ”Theory of Probability, 3rd edn.” Oxford 135012 (2015) [arXiv:1410.3835 [gr-qc]]. University Press, 1961. [78] T. B. Littenberg, J. B. Kanner, N. J. Cornish and [90] R. Abbott et al. [LIGO Scientific and Virgo], M. Millhouse, “Enabling high confidence detections of “GW190814: Gravitational Waves from the Coalescence gravitational-wave bursts,” Phys. Rev. D 94, no.4, of a 23 Solar Mass Black Hole with a 2.6 Solar Mass Com- 044050 (2016) [arXiv:1511.08752 [gr-qc]]. pact Object,” Astrophys. J. Lett. 896, no.2, L44 (2020) [79] K. Chatziioannou, C. J. Haster, T. B. Littenberg, [arXiv:2006.12611 [astro-ph.HE]]. W. M. Farr, S. Ghonge, M. Millhouse, J. A. Clark and [91] A. Vijaykumar, N. K. Johnson-McDaniel, R. Kashyap, N. Cornish, “Noise spectral estimation methods and their A. Mukherjee, and P. Ajith, in preparation. impact on gravitational wave measurement of compact