Measuring the Low-Energy Weak Mixing Angle with Supernova Neutrinos Chun-Ming Yip, Xu-Run Huang, Ming-chung Chu, Qishan Liu Abstract The weak mixing angle θ W \theta_{W} is a fundamental parameter in the electroweak theory with a value running according to the energy scale, and its precision measurement in the low-energy regime is still ongoing. We propose a method to measure the low-energy sin ⁡ ( θ W 2 ) \sin{^2\theta_W} by taking advantage of Argo, a future ton-scale liquid argon dark matter detector, and the neutrino flux from a nearby core-collapse supernova (CCSN). We evaluate the expected precision of this measurement through the coherent elastic neutrino-nucleus scattering (CE ν \nu NS) channel. We show that Argo is potentially capable of achieving a few percent determination of sin ⁡ ( θ W 2 ) \sin{^2\theta_W} , at the momentum transfer of q ∼ 20 ​ MeV q\!\sim\!20{\,\text{MeV}} , in the observation of a CCSN within ∼ 3 ​ kpc \sim\!3{\,\text{kpc}} from the Earth. Such a measurement is valuable for both the precision test of the electroweak theory and searching for new physics beyond the standard model in the neutrino sector. 1 Introduction The weak mixing angle \theta_{W} is one of the most fundamental parameters in the Standard Model (SM) of particle physics, which characterizes the mixing of vector bosons from the SU\left(2\right)_{L}\times U\left(1\right)_{Y} symmetry Weinberg (1967) . Collider experiments in the last decades had achieved high precision measurements of \sin{^2\theta_W} near the so-called ’electroweak scale’ of \sim\!100{\,\text{GeV}} Workman et al. (2022) . The value of \sin{^2\theta_W} runs along with the energy scale where it is measured due to quantum effects, particularly at the next-to-leading-order. The SM prediction at zero momentum transfer q=0 is \sin{^2\theta_W}=0.23863\pm 0.00005 under the modified minimal subtraction ( \overline{MS} ) renormalization scheme Workman et al. (2022) . Experiments are also performed to check this prediction. Nevertheless, the precise determination of \sin{^2\theta_W} in the low-energy regime is still an ongoing issue Kumar et al. (2013) . Previous efforts can be categorized into the charged lepton and neutrino sectors. In the former, the Qweak Collaboration recently reported its result with a precision of \pm 0.5\% at q=0.158{\,\text{GeV}} via the measurement of parity-violation asymmetry in a polarized electron scattering experiment Androić et al. (2018) , which agrees well with the SM prediction. In the low-energy regime, the most precise results to date are provided at q\simeq 2.4{\,\text{MeV}} by the atomic parity violation (APV) experiments, but they turn out to be smaller than the SM prediction by \sim\!1.5\sigma Dzuba et al. (2012) ; Roberts et al. (2015) . In the neutrino sector, an experimental anomaly ( 3\sigma ) is reported by the NuTeV experiment, concerning the SM prediction for neutrino-nucleon scatterings at q\!\sim\!1{\,\text{GeV}} Zeller et al. (2002) . At the MeV scale, the coherent elastic neutrino-nucleus scattering (CE \nu NS) is a powerful probe, which was predicted theoretically in 1973 Freedman (1974) ; Freedman et al. (1977) but only observed experimentally for the first time in 2017 Akimov et al. (2017) . In this interaction, the recoil of the nucleus is the only observable signature, which is very difficult to detect since it shows up with an extremely small kinetic energy. Observations have been achieved with neutrinos from stopped pion decays by the COHERENT collaboration using a cesium-iodide (CsI) detector Akimov et al. (2017) and a liquid argon detector Akimov et al. (2021) . The COHERENT-Ge Adamski et al. (2024) , Dresden-II Colaresi et al. (2022) and CONUS+ Ackermann et al. (2025) experiments observed reactor antineutrinos by CE \nu NS on germanium. The recent detection of solar {}^{8}\text{B} neutrinos on dark matter (DM) detectors, reported by the PandaX-4T Bo et al. (2024) and XENONnT Aprile et al. (2024) collaborations, are also conducted via the CE \nu NS. Limited by the sparse data, an accurate extraction of \sin{^2\theta_W} still requires further efforts (for recent progress, see Refs. Pandey (2024) ; Cadeddu et al. (2023) ). Meanwhile, the nature of neutrinos remains elusive. New physics beyond the SM may exist in the low-energy neutrino-nucleon interactions and shift the extracted value of \sin{^2\theta_W} in these interactions Barranco et al. (2005) ; Kumar et al. (2013) ; Billard et al. (2018) ; Cadeddu et al. (2018) , e.g., neutrino non-standard interactions (NSIs) Wolfenstein (1978) ; Ohlsson (2013) ; Bhupal Dev et al. (2019) , neutrino charge properties Giunti and Studenikin (2015) , etc.. Any experimental constraints in this sector also render valuable information on searching for new physics beyond the SM. Such a scientific potential stimulates further experimental and theoretical attention to the CE \nu NS. Using low-energy neutrino fluxes from nuclear reactors or stopped-pion decays in accelerators, some terrestrial CE \nu NS experiments are running, under construction, or being planned, e.g.,CONNIE Aguilar-Arevalo et al. (2016) , MINER Agnolet et al. (2017) , NEON Choi et al. (2023) , NUCLEUS Angloher et al. (2019) , Ricochet Colas et al. (2023) , RED-100 Akimov et al. (2022) , and so on Baxter et al. (2020) . Meanwhile, based on some of the above reactor neutrino experiments, the prospect of measuring the low-energy \sin{^2\theta_W} has been discussed, e.g., a sub- 1\% extraction precision is found to be possible in ideal cases Cañas et al. (2018) . On the other hand, several determinations of the \sin{^2\theta_W} at the MeV energy scale have already been obtained by analyzing the recent CE \nu NS data (e.g., COHERENT, Dresden-II, CONUS+, PandaX-4T and XENONnT, etc.) Aristizabal Sierra et al. (2022) ; Atzori Corona et al. (2022) ; Majumdar et al. (2022) ; De Romeri et al. (2023) ; Atzori Corona et al. (2023) ; Maity and Boehm (2024) ; De Romeri et al. (2025) ; Alpízar-Venegas et al. (2025) ; Chattaraj et al. (2025) ; De Romeri et al. (2025) . The error bars are still large (see Fig. 5 ). Some efforts have also been made in the context of combining other low-energy probes and the measurement of the neutron radius of {}^{133}{\rm Cs} Huang et al. (2024) ; Atzori Corona et al. (2024) . CE \nu NS measurements can also be conducted using the low-energy neutrino flux from a galactic core-collapse supernova (CCSN). The first-ever supernova neutrino detection was done for SN 1987A, by the Kamiokande II, Irvine-Michigan-Brookhaven (IMB), and Baksan detectors Hirata et al. (1987) ; Haines et al. (1988) ; Alekseev et al. (1988) . Sophisticated CCSN simulations, in the meantime, have been performed by different groups to predict the neutrino signals of CCSNe Hudepohl (2013) ; Mirizzi et al. (2016) ; Nagakura et al. (2020) ; Burrows and Vartanyan (2021) . It is found that \sim\!10^{58} neutrinos, with mean energies of \sim\!10-20{\,\text{MeV}} , are released within a few seconds after a CCSN explosion occurs. Such a neutrino flux can be detected by large-scale DM detectors. For example, the sensitivity of future xenon and argon DM detectors to supernova neutrinos has been investigated, and \gtrsim\!10^{2} CE \nu NS events are expected to be observed for a typical CCSN at 10{\,\text{kpc}} from the Earth Lang et al. (2016) ; Agnes et al. (2021) . Meanwhile, other dedicated large-scale neutrino observatories, such as Super-Kamiokande Abe et al. (2016) (Hyper-Kamiokande Abe et al. (2021) ), DUNE Abi et al. (2021) , RES-NOVA Pattavina et al. (2020) ; Pattavina et al. (2021) , JUNO Abusleme et al. (2022) , and so on Scholberg (2012) ; Mirizzi et al. (2016) ; Horiuchi and Kneller (2018) , are capable of collecting a substantial number of supernova neutrino events, e.g., \mathcal{O}(10^{5}) for a source at 10{\,\text{kpc}} Mirizzi et al. (2016) ; Horiuchi and Kneller (2018) . So, a combined analysis of data from multiple detectors may potentially achieve a precise reconstruction of supernova neutrino energy spectrum for a nearby CCSN Nikrant et al. (2018) ; Gallo Rosso et al. (2018) ; Li et al. (2018) ; Li et al. (2019) ; Nagakura (2020) ; Huang et al. (2023) . Measuring the low-energy \sin{^2\theta_W} is feasible by taking advantage of the reconstructed supernova neutrino spectrum from large-scale neutrino observatories and the CE \nu NS events in large-scale DM detectors. In this work, we estimate the potential of measuring the low-energy \sin{^2\theta_W} with Argo, a proposed next-generation liquid argon DM detector Aalseth et al. (2018) ; Agnes et al. (2021) , in the observation of a nearby CCSN. The Argo experiment is suitable for this task. It has a large expected active mass of 362.7 tons Agnes et al. (2021) , considerably larger than other large-scale DM detectors (e.g., 43 tons for PandaX-xT Abdukerim et al. (2025) , 5.9 tons for XENONnT Aprile et al. (2024) and 40 tons for DARWIN Aalbers et al. (2016) ), which renders a high sensitivity to supernova neutrinos. Compared with heavier nuclei (e.g., Xe in other DM detectors), selecting the Ar nucleus can alleviate the uncertainty from the undetermined neutron distribution in the nucleus. Currently, the ab initio calculation of the neutron distribution is easier and more precise for light and medium-mass nuclei (see, e.g., Ref. Hagen et al. (2015) ). This paper is organized as follows. In Sect. 2 , we introduce the general properties of supernova neutrinos and 4 example fluxes obtained by numerical simulations. In Sect. 3 , we present the detection prospect in Argo. The expected precision and relevant uncertainty analysis of the low-energy \sin{^2\theta_W} measurement are shown in Sect. 4 . Finally, we conclude in Sect. 5 . 2 Supernova neutrino flux In a real measurement, the neutrino flux can be determined by observations, including the direct measurements by dedicated neutrino observatories and outcomes from CCSN models which have been calibrated by multi-messenger observations. In this paper, we use neutrino fluxes from CCSN simulations in the literature. Thanks to tremendous progress in modeling CCSN phenomena in the last decades, modern hydrodynamic codes are capable of showing robust explosions in simulations Müller (2016) ; Burrows and Vartanyan (2021) . The supernova neutrino energy spectrum of one species ( \nu_{\beta} in \nu_{e},\nu_{\mu},\nu_{\tau} , and anti-particles) is well approximated by a quasi-thermal distribution (the so-called Garching formula Keil et al. (2003) ; Tamborra et al. (2012) ) as a function of neutrino energy E_{\nu_{\beta}} : f_{\nu_{\beta}}\left(E_{\nu_{\beta}}\right)=C_{\beta}\left({\frac{E_{\nu_{\beta}}}{\langle E_{\nu_{\beta}}\rangle}}\right)^{\alpha_{\beta}}\exp{-\left( \alpha_\beta+1 \right) \frac{E_{\nu_\beta}}{\langle E_{\nu_\beta} \rangle}}, (1) where \langle E_{\nu_{\beta}}\rangle is the mean neutrino energy, \alpha_{\beta} is the shape parameter which controls the suppression of high energy tail, and the normalization constant C_{\beta} is given by C_{\beta}=\frac{\left(\alpha_{\beta}+1\right)^{\alpha_{\beta}+1}}{\langle E_{\nu_{\beta}}\rangle\Gamma\left(\alpha_{\beta}+1\right)} (2) for the species \nu_{\beta} , with \Gamma denoting the gamma function. Neutrino flavor transitions Scholberg (2018) have no impact on our results since the CE \nu NS is flavor-blind. The total neutrino flux on the Earth can be evaluated by summing over all neutrino species, i.e., \Phi\left(E_{\nu}\right)=\sum_{\beta}\frac{1}{4\pi d^{2}}\frac{E^{tot}_{\nu_{\beta}}}{\langle E_{\nu_{\beta}}\rangle}f_{\nu_{\beta}}\left(E_{\nu_{\beta}}\right) (3) as a function of neutrino energy E_{\nu} regardless of species, with d being the distance of the CCSN and E^{tot}_{\nu_{\beta}} the total energy emitted through one species. In practice, Eq. ( 1 ) is used to describe the energy spectrum of both the time-integrated flux as f_{\nu_{\beta}}\left(E_{\nu_{\beta}}\right) , and the time-dependent flux as f_{\nu_{\beta}}\left(E_{\nu_{\beta}},t\right) with \alpha_{\beta}(t) and \langle E_{\nu_{\beta}}(t)\rangle . Correspondingly, Eq. ( 3 ) can also describe the total differential flux \Phi\left(E_{\nu},t\right) with \langle E_{\nu_{\beta}}(t)\rangle and f_{\nu_{\beta}}\left(E_{\nu_{\beta}},t\right) , after replacing E^{tot}_{\nu_{\beta}} with the neutrino luminosity \mathcal{L}_{\nu_{\beta}}(t) . The time-integrated flux is defined as \Phi\left(E_{\nu}\right)\equiv\int\Phi\left(E_{\nu},t\right)\mathrm{d}t . Table 1: The CCSN models from the Garching group Hudepohl (2013) ; Mirizzi et al. (2016) . The progenitor mass and nuclear EoS of each model are listed Model Progenitor mass [M ⊙ ] EoS s11.2-S 11.2 Shen s11.2-L 11.2 LS220 s27.0-S 27.0 Shen s27.0-L 27.0 LS220 The neutrino fluxes from CCSN simulations depend not only on the properties of progenitor star (e.g., mass, compactness, etc. O’Connor and Ott (2013) ; Seadrow et al. (2018) ) but also on the nuclear matter equation of state (EoS) Pan et al. (2018) ; da Silva Schneider et al. (2020) ; Raduta et al. (2021) . We use realistic values of \alpha_{\beta}(t) , \langle E_{\nu_{\beta}}(t)\rangle and \mathcal{L}_{\nu_{\beta}}(t) from 4 CCSN explosion models provided by the Garching group, considering the above factors. These models are listed in Table 1 , and refer to Refs. Hudepohl (2013) ; Mirizzi et al. (2016) . The temporal evolution of the neutrino spectral parameters is shown in Fig. 1 for the models s11.2-S and s27.0-L, where the former (latter) model results in the smallest (largest) expected number of event counts among the 4 models (see Sect. 3 ). The explosions of these CCSN models were artificially triggered at \sim\!0.5\,\rm s after bounce and some kinks appear accordingly in the neutrino curves at \sim\!0.5\,\rm s , which do not lead to any significant perturbing consequences for the subsequent proto-neutron star cooling evolution Mirizzi et al. (2016) . The feature seen during this time will not appear in self-consistent explosion models. We find that the neutrinos emitted during 0.5\!\sim\!0.53\,\rm s only contribute to the total CE \nu NS events by <\!2\% . Note that the time-integrated flux \Phi\left(E_{\nu}\right) is already sufficient to compute the total number of CE \nu NS events, and that is also what we can directly construct from the detection by other neutrino observatories. Figure 1: Temporal evolution of supernova neutrino spectral parameters for the models s11.2-S (dashed-dotted lines) and s27.0-L (solid lines). \nu_{e} , \bar{\nu}_{e} , and \nu_{x} ( \nu_{x} denotes one of \nu_{\mu} , \bar{\nu}_{\mu} , \nu_{\tau} , and \bar{\nu}_{\tau} ) are colored blue, red, and purple, respectively. Upper panel: neutrino mean energies \langle E_{\nu_{\beta}}\rangle . Middle panel: neutrino luminosities \mathcal{L}_{\nu_{\beta}} . The total neutrino luminosities are also shown in black. Lower panel: shape parameter \alpha_{\beta} . Note that the feature appearing in all curves at \sim\!0.5\,\rm s is due to the artificial triggering of the explosion models (see more discussion in the text) Mirizzi et al. (2016) 3 Detection prospect in Argo The primary purpose of conventional direct DM detectors is to search for nuclear recoils induced by galactic DM particles (e.g., the weakly interacting massive particles). The CE \nu NS signals induced by solar neutrinos, reactor neutrinos, atmospheric neutrinos, the diffuse supernova neutrino background, and geoneutrinos will become an unavoidable background as the sensitivity of DM detectors approaches keV level Marrodán Undagoitia and Rauch (2016) ; O’Hare (2021) . Conversely, these detectors can act as excellent low-energy neutrino observatories. The opportunity of observing supernova neutrinos has been investigated for large-scale DM detectors, e.g., liquid xenon detectors like XENONnT, DARWEN, and LZ Lang et al. (2016) ; Khaitan (2018) as well as liquid argon detectors like DarkSide-20k and Argo Agnes et al. (2021) . In this section, we demonstrate how to obtain the expected CE \nu NS event counts of the Argo experiment. The Argo experiment is proposed as an extension of DarkSide-20k by the DarkSide Collaboration Aalseth et al. (2018) ; Agnes et al. (2021) . It is assumed to have a dual-phase Liquid Argon Time Projection Chamber (LAr TPC) with an active mass of 362.7 tons Agnes et al. (2021) . Such a dual-phase TPC is primarily filled with liquid argon and has a gaseous argon phase on the top. When a CE \nu NS event occurs in the instrumented volume, two measurable signals are produced, i.e., a prompt light pulse (S1) induced by scintillation in the liquid phase, and a delayed pulse (S2) associated with ionization electrons. The S2-only mode can achieve a lower energy threshold and a better detection efficiency than the S1-only or S1-S2 coincident modes for supernova neutrino detection Lang et al. (2016) ; Agnes et al. (2021) . The differential cross section of CE \nu NS in the SM, as a function of the neutrino energy E_{\nu} and the nuclear recoil energy T , is given by Drukier and Stodolsky (1984) ; Lindner et al. (2017) ; Huang and Chen (2022) ; Tomalak et al. (2021) \derivative{\sigma}{T}\left(E_{\nu},T\right)=\frac{G_{F}^{2}M}{4\pi}Q_{W}^{2}F_{W}^{2}\left(q\right)\left[1-\frac{T}{E_{\nu}}-\frac{MT}{2E_{\nu}^{2}}\right], (4) where G_{F} is the Fermi coupling constant, M the mass of the target nucleus with Z ( N ) protons (neutrons), Q_{W} the weak charge of the target nucleus, and F_{W}\left(q\right) the corresponding weak form factor. The momentum transfer q is given by q^{2}=2ME_{\nu}^{2}T/(E_{\nu}^{2}-E_{\nu}T)\simeq 2MT for E_{\nu}\gg T . Note that Eq. 4 describes the interaction between a neutrino and a nucleus with spin-0 Lindner et al. (2017) . Natural argon is 99.6\% {}^{40}\mathrm{Ar} with spin-0 Meija et al. (2016) , so we neglect the contributions from other isotopes. The mass of {}^{40}\mathrm{Ar} can be calculated via M=Z\times m_{p}+N\times m_{n}-\left(Z+N\right)E_{B} , with m_{p} ( m_{n} ) being the rest mass of proton (neutron) and E_{B}=8.595259{\,\text{MeV}} the binding energy per nucleon Wang et al. (2021) . The weak charge of a nucleus Q_{W} is given by Q_{W}=-2\left(Zg_{V}^{p}+Ng_{V}^{n}\right), (5) where the proton and neutron vector couplings are g_{V}^{p}=1/2-2\sin{^2\theta_W} and g_{V}^{n}=-1/2 , respectively, at the tree level. The weak form factor F_{W}\left(q\right) can be computed by F_{W}\left(q\right)=-2\frac{\left[Zg_{V}^{p}F_{p}\left(q\right)+Ng_{V}^{n}F_{n}\left(q\right)\right]}{Q_{W}}, (6) where F_{n(p)}\left(q\right) is the neutron (proton) form factor. We adopt the Helm parametrization Helm (1956) ; Piekarewicz et al. (2016) since it is very successful in analyzing electron scattering form factors Rosen et al. (1967) ; Raphael and Rosen (1970) ; Friedrich and Voegler (1982) and has also been used in previous CE \nu NS studies Cadeddu et al. (2018) ; Huang and Chen (2019) ; Cadeddu et al. (2020) . It has the form as F_{i}\left(q\right)=3\frac{j_{1}\left(qR_{0,i}\right)}{qR_{0,i}}e^{-q^{2}s^{2}/2}, (7) where i=p,n labels the form factor associated with proton and neutron, respectively, and j_{1}\left(x\right)=\sin{x}/x^{2}-\cos{x}/x is the spherical Bessel function of order one. The diffraction radius R_{0,i} is related to the root-mean-squared (rms) radius R_{i} by R_{i}^{2}=\frac{3}{5}R_{0,i}^{2}+3s^{2}, (8) with s quantifying the surface thickness. We adopt the value s=0.9\,\text{fm} , following Ref. Cadeddu et al. (2020) . R_{i} is associated with the point-nucleon distribution radius R_{i}^{\text{point}} and the radius of nucleon \langle r_{i}^{2}\rangle^{1/2} , i.e., R_{i}^{2}=(R_{i}^{\text{point}})^{2}+\langle r_{i}^{2}\rangle. (9) The radius of a proton equals its charge radius and has the value of \langle r_{p}^{2}\rangle^{1/2}=0.8414(19)\,\text{fm} Hammer and Meißner (2020) . For neutron, we assume \langle r_{n}^{2}\rangle^{1/2}\simeq\langle r_{p}^{2}\rangle^{1/2} , which is supported by the neutron magnetic radius measurement Tanabashi et al. (2018) . Electron scattering experiments have made precise measurements on the charge radius of a nucleus ( R_{c} ), which is given by Ong et al. (2010) ; Horowitz et al. (2012) R_{c}^{2}=(R_{p}^{\text{point}})^{2}+\langle r_{p}^{2}\rangle+\frac{N}{Z}\langle r_{n}^{2}\rangle_{c}, (10) with \langle r_{n}^{2}\rangle_{c}=-0.1161(22)\,\text{fm}^{2} Tanabashi et al. (2018) being the squared charge radius of a neutron. For {}^{40}\text{Ar} , it is measured as R_{c}=3.4274\pm 0.0026 fm Fricke et al. (1995) ; Angeli and Marinova (2013) . As a consequence, we obtain the value of R_{p} as R_{p}=3.448\pm 0.003\,\text{fm}. (11) On the other hand, the neutron rms radius is experimentally unknown Thiel et al. (2019) , and different nuclear models would give different predictions (see, e.g., Table I in Ref. Cadeddu et al. (2020) ). Theoretical predictions indicate that R_{n}^{\text{point}} is universally larger than R_{p}^{\text{point}} by 0.08-0.11\,\text{fm} for {}^{40}\text{Ar} Cadeddu et al. (2020) . We take such an assumption: R_{n}\simeq R_{p}+0.1\,\text{fm}. (12) It should be further pointed out that the neutron form factors at \mathcal{O}(10){\,\text{MeV}} are not sensitive to uncertainties of neutron distributions in nuclei since its value is close to 1 , especially for light nuclei like {}^{40}\text{Ar} (see, e.g., Fig. 1 in Ref. Kerman et al. (2016) ). We check it by varying R_{n}-R_{p} by \pm 20\% and the resulting CE \nu NS event counts only deviate by about \mp 0.1\% for the 4 aforementioned CCSN models. Also, the choice of the form factor parametrization could hardly affect the results here, e.g., the symmetrized Fermi and Helm form factors give the same results in the analysis of the COHERENT data Cadeddu et al. (2018) . We estimate the number of expected events \mathcal{N} in Argo by the expression, \mathcal{N}=N_{\text{Ar}}\int\mathrm{d}E_{\nu}\mathcal{E}\left(E_{\nu}\right)\int_{T_{\text{min}}}^{T_{\text{max}}}\mathrm{d}T\derivative{\sigma}{T}\left(E_{\nu},T\right)\Phi\left(E_{\nu}\right). (13) Here, N_{\text{Ar}}=N_{A}m_{\text{det}}/M_{\text{Ar}} denotes the number of Ar atoms contained in the active volume of the detector, with N_{A} being the Avogadro constant, m_{\text{det}} the active mass of the detector, and M_{\text{Ar}}=39.96\,\text{g/mol} the molar mass of argon. We consider the E_{\nu} -dependent detection efficiency \mathcal{E}\left(E_{\nu}\right) , which is evaluated in the range of [3, 100] ionization electrons (see Fig. 4 in Ref. Agnes et al. (2021) ). This detection efficiency is preferred in the study of supernova neutrino detection by DarkSide-20k and Argo, to suppress the single-electron background and the {}^{39}\text{Ar} events Agnes et al. (2021) . When integrating over T , an energy threshold of T_{\text{min}}=0.46 keV is considered, since it has been proven as feasible by DarkSide-50 Agnes et al. (2018a) and also applied for Argo in Ref. Agnes et al. (2021) . For the scattering between a Ar nucleus and an incoming neutrino with an energy E_{\nu} , T_{\text{max}}=2E_{\nu}^{2}/(M_{\text{Ar}}+2E_{\nu}) . The integration over E_{\nu} starts from 1{\,\text{MeV}} , where \mathcal{E}(E_{\nu}) already decreases to zero, and ends at 60{\,\text{MeV}} , which is high enough for the energy spectrum (Eq. 1 ) to almost fade out. The scattering rate \mathrm{d}\mathcal{N}/\mathrm{d}t can also be computed by Eq. ( 13 ) if we replace \Phi\left(E_{\nu}\right) with \Phi\left(E_{\nu},t\right) . Figure 2: Expected event counts of CE \nu NS in Argo for the 4 models, evaluated at d=10{\,\text{kpc}} . Upper panel: scattering rate \mathrm{d}\mathcal{N}/\mathrm{d}t . Lower panel: cumulative event counts \mathcal{N} over time Figure 2 shows the scattering rate ( \mathrm{d}\mathcal{N}/\mathrm{d}t ) and the corresponding cumulative event counts for the 4 CCSN models, with the \sin{^2\theta_W} in the SM at d=10{\,\text{kpc}} . The cumulative event counts \mathcal{N} are calculated to be 1012, 1284, 1820, and 2357 for the 4 CCSN models presented in Table 1 , respectively, at 7\,\text{s} after bounce and saturate afterward. More events are recorded for the models s27.0-S and s27.0-L with heavier progenitors. We also find that about half of the events are recorded within the first second after the bounce. 4 Low-energy weak mixing angle sensitivity 4.1 Statistical analysis To assess the precision of measuring \sin^{2}{\theta_{W}} , we perform a \chi^{2} analysis, assuming that the number of observed events by Argo is \mathcal{N}_{\text{obs}}=\mathcal{N}_{\text{true}}-\mathcal{N}_{\text{bias}} , where \mathcal{N}_{\text{true}} is the actual number of total CE \nu NS interactions and \mathcal{N}_{\text{bias}} is the number of unresolved interactions due to event pile-up (see later contents for more detailed discussions). Following Refs. Cañas et al. (2018) ; Huang and Chen (2022) , we consider the chi-squared function \chi^{2}=\frac{\left(\mathcal{N}_{\text{obs}}^{\text{SM}}-\mathcal{N}_{\text{obs}}^{\text{th}}\right)^{2}}{\sigma^{2}_{\text{stat}}+\sigma^{2}_{\text{syst}}}. (14) Here, \mathcal{N}_{\text{obs}}^{\text{SM}} denotes the number of observed events computed with the SM value of \sin^{2}{\theta_{W}} at zero momentum transfer, i.e., \sin^{2}{\theta_{W}}=0.23863 Workman et al. (2022) , while \mathcal{N}_{\text{obs}}^{\text{th}} varies according to the value of \sin{^2\theta_W} . We divide the uncertainty estimation into two parts, i.e., a pure statistical uncertainty \sigma_{\text{stat}}=\sqrt{\mathcal{N}_{\text{obs}}^{\text{SM}}} and a systematic error \sigma_{\text{syst}} . In this section, we attempt to draw an estimation based on the current limited knowledge of \sigma_{\text{syst}} . Firstly, due to the ultra-high event rate and limitation from the time resolution of LAr TPC, the event pile-up will become inevitable for a nearby target. This will reduce the total count collected in Argo and introduce a systematic uncertainty. Secondly, the energy spectra of supernova neutrinos are measured with an uncertainty, and it introduces an uncertainty on the expected number of events in Argo. This uncertainty will be more pronounced for a far target. The total systematic error of the low-energy \sin^{2}{\theta_{W}} measurement considering the above two types of uncertainties can be written as \sigma^{2}_{\text{syst}}=\sigma^{2}_{\text{bias}}+\sigma^{2}_{\text{flux}} . Here, \sigma_{\text{bias}} denotes the uncertainty in event pile-up estimation in Argo; \sigma_{\text{flux}} denotes the flux uncertainty of expected events in Argo, which originates from the uncertainty of the extracted supernova neutrino flux \sigma_{\nu} in other detectors. Nonetheless, there may also exist more uncertainties from other sources, e.g., the high-order corrections to the cross-section, the configuration of the Argo detector, and so on. In this section, we only consider the previous two types of uncertainties and leave a comprehensive uncertainty evaluation to a real measurement in the future. According to the study based on DarkSide-50 Agnes et al. (2018b) , the observed S2 signal in a dual-phase LAr TPC is a pulse with a width of \sim\!10\,\text{$\mu$s} (or a FWHM of \sim\!5\,\text{$\mu$s} ). When an intense neutrino flux from a nearby source comes into the LAr TPC, a large number of interactions occur within a short period of time, and some S2 pulses may overlap with each other. Such a phenomenon will reduce the total number of observed events by a bias number \mathcal{N}_{\text{bias}} . Meanwhile, \mathcal{N}_{\text{bias}} will get a value randomly in the detection according to a probability distribution. To evaluate \mathcal{N}_{\text{bias}} and its uncertainty, we use the Monte Carlo method to study the pile-up issue, which was also used in dual-phase xenon TPCs in Ref. Lang et al. (2016) . We adopt the assumption that two S2 signals are distinguishable only if the spacing from the start of one pulse to the start of the next is larger than 10\,\text{$\mu$s} . Otherwise, these two close S2 signals will be combined as one. The expected events are distributed randomly in time according to the scattering rate \mathrm{d}\mathcal{N}/\mathrm{d}t and uniformly throughout the TPC. Argo is assumed to have a height of 5 m and a drift velocity of 0.93\,\text{mm}/\text{$\mu$s} Agnes et al. (2021) . We also take into account the time delay due to the drifting of ionization electrons. With the above setup, a set of Monte Carlo simulations is performed, which consists of 3\times 10^{4} individual mock triggers for each d . After that, we count the number of unresolved events in each trigger and obtain a histogram of \mathcal{N}_{\text{bias}} for this dataset. Finally, the mean value \langle\mathcal{N}_{\text{bias}}\rangle and uncertainty \sigma_{\text{bias}} can be extracted from the histogram. The Monte Carlo simulation results with \sin^{2}{\theta_{W}}=0.23863 are displayed in Fig. 3 . As d decreases, the mean portion of pile-up events \langle\mathcal{N}_{\text{bias}}\rangle/\mathcal{N}_{\text{true}} increases, but the relative 1\sigma variation \sigma_{\text{bias}}/\mathcal{N}_{\text{true}} decreases. This is generic for all 4 CCSN models. The magnitude of \sigma_{\text{bias}} grows to \sim\!100 when d approaches 1{\,\text{kpc}} , which is \sim\!0.1\% of \mathcal{N}_{\text{true}} in all models. The prediction of \mathcal{N}_{\text{bias}} depends on the CCSN model (e.g., the progenitor mass) and d , while \sigma_{\text{bias}} is less affected by these factors according to Fig. 3 . In a real measurement, the multi-messenger observation of a nearby CCSN will be used to calibrate the CCSN model, and also to measure d (further discussions can be found in Sect. 4.2 ). Hence, we only consider the distance dependence now. We remark that, technically speaking, two S2 signals with a spacing smaller than 10\,\text{$\mu$s} can still be distinguished if the S2 pulse shapes get read out in a real measurement Agnes et al. (2018b) . We also test the case of shrinking the criterion of resolving two successive events from 10\,\text{$\mu$s} to 5\,\text{$\mu$s} in the Monte Carlo simulations. Using the new criterion, \langle\mathcal{N}_{\text{bias}}\rangle/\mathcal{N}_{\text{true}} is universally shifted down by a factor of \sim\!1/2 at d\gtrsim 3{\,\text{kpc}} in all models, and by a factor of \sim\!1/3 at d\sim 1{\,\text{kpc}} . Another way to avoid a large portion of pile-up events is to only consider the cooling phase where the neutrino luminosity is less intense (e.g., t\gtrsim 1\,\text{s} for the models in Fig. 1 ). Our simulations find that the mean portions of pile-up events are reduced by \sim\!50\% when only the cooling phases are considered. Figure 3: Simulated event pile-up in Argo with \sin^{2}{\theta_{W}}=0.23863 . For each set of Monte Carlo simulations at a certain value of d , 3\times 10^{4} individual mock triggers are generated. Upper panel: mean portion of pile-up events \langle\mathcal{N}_{\text{bias}}\rangle/\mathcal{N}_{\text{true}} (in percentage). Lower panel: 1\sigma variation of pile-up events \sigma_{\text{bias}}/\mathcal{N}_{\text{true}} Another issue is the uncertainty in the number of expected events in Argo. Although the time-dependent supernova neutrino fluxes are used to calculate the total number of expected counts in Eq. ( 14 ) in this work, the time-integrated neutrino fluxes should be used in a real measurement. They are expected to be measured by large-scale neutrino observatories in parallel to Argo. That is, by analyzing the data from one detector or a combination of multiple detectors, parameters of the time-integrated neutrino spectra can be reconstructed via various methods (see, e.g., Refs. Nikrant et al. (2018) ; Gallo Rosso et al. (2018) ; Li et al. (2018) ; Li et al. (2019) ; Nagakura (2020) ; Huang et al. (2023) ). We firstly estimate the uncertainty of the extracted supernova neutrino flux \sigma_{\nu} (in percentage) in other detectors according to the parameter extraction precision of all flavors in a Bayesian approach Huang et al. (2023) . The 2\sigma extraction precision for the observation of a 10{\,\text{kpc}} target is presented in Table 3 of Ref. Huang et al. (2023) , for both normal mass ordering (NMO) and inverted mass ordering (IMO). We first calculate the total neutrino spectra according to the 2\sigma parameter space and find the upper and lower boundaries. Then, we integrate the neutrino spectra to get the total numbers of neutrinos associated with the most probable spectrum as well as the two boundaries. The 2\sigma flux uncertainty can thus be estimated as the difference in total numbers of neutrinos (in percentage) between the most probable spectrum and the two boundaries. We find that \sigma_{\nu}\simeq 14.1\,(13.1)\% at d=10{\,\text{kpc}} assuming NMO (IMO). Furthermore, the authors of Ref. Huang et al. (2023) also tested the distance effect by assuming a closer source, e.g., d=5{\,\text{kpc}} , and they find that the precisions are improved by about 40-50\% among almost all parameters. Their results indicate that the reconstruction uncertainty is currently dominated by the statistics. Consequently, we make such an assumption: \sigma_{\nu}\propto 1/\sqrt{\mathcal{N}_{\nu}}\propto d since the total number of supernova neutrinos on the Earth is \mathcal{N}_{\nu}\propto 1/d^{2} . While \mathcal{N}_{\text{true}}^{\text{th}}\propto\mathcal{N}_{\nu} and \mathcal{N}_{\text{bias}}^{\text{th}} is also related to \mathcal{N}_{\nu} , we can estimate the flux uncertainty of expected events \sigma_{\text{flux}} in Argo according to the uncertainty of the extracted supernova neutrino flux \sigma_{\nu} in other detectors evaluated above. Thus, we adopt \sigma_{\text{flux}}\simeq\mathcal{N}_{\text{obs}}^{\text{th}}\times 14.1\%\times(d/10{\,\text{kpc}}) assuming NMO. To test such an estimate, we perform an additional set of Monte Carlo simulations for the model s27.0-L, following the aforementioned setting, and vary \mathcal{N}_{\text{true}} by \pm 10\% . The resulting variation in \mathcal{N}_{\text{obs}} is about \pm 10\,(7)\% at d=5\,(1){\,\text{kpc}} , smaller than the variation in \mathcal{N}_{\text{true}} itself at small d . Therefore, we remark that the above expression of \sigma_{\text{flux}} may slightly overestimate the flux extraction uncertainty for a very close source. The caveat is that the assumption of \sigma_{\nu}\propto d may not work for a very close source, since other systematic errors may overwhelm the statistical error. Finally, we obtain the total systematic error of the low-energy \sin{^2\theta_W} measurements by \sigma^{2}_{\text{syst}}=\sigma^{2}_{\text{bias}}+\sigma^{2}_{\text{flux}} . We evaluate the expected sensitivity of Argo to the variation in \sin{^2\theta_W} using Eq. ( 14 ). The sensitivity curves at a few benchmark values of d are shown in Fig. 4 for the model s27.0-L as an example. NMO is assumed when evaluating \sigma_{\text{flux}} , and a similar result is obtained when IMO is assumed. Currently, \sigma_{\text{flux}} is the most significant source of uncertainty for these benchmark values of d . The expected precision of the low-energy \sin{^2\theta_W} measurement is higher for a smaller d due to the proportionality of \sigma_{\text{flux}} to d . In particular, a theoretical precision of 2.8\,(1.3)\% at the 1\sigma confidence level can be achieved at d=3\,(1){\,\text{kpc}} , and it is 4.5\% at d=5{\,\text{kpc}} . Note that the corresponding curves for different CCSN models are nearly identical. We further remark that these results should generally be considered as upper limits of the precision due to incomplete uncertainty evaluation. Figure 4: Expected sensitivity of Argo to the variation in low-energy \sin{^2\theta_W} from the SM value of 0.23863 (indicated by the vertical black dashed-dotted line), at a few benchmark values of d . The 1\sigma and 2\sigma confidence levels are indicated by the horizontal grey dotted lines. The model s27.0-L and NMO are assumed 4.2 Discussions Figure 5 shows the comparison between the expected 1\sigma precision of low-energy \sin{^2\theta_W} measurement by Argo, as indicated in Fig. 4 , and the results of other experiments. It can be seen that the expected error bar is always smaller than those derived from recent CE \nu NS experiments, e.g., Dresden-II Atzori Corona et al. (2022) , COHERENT De Romeri et al. (2023) , and CONUS+ De Romeri et al. (2025) . Especially, we would like to compare our results with the expected sensitivities of the CONNIE, MINER, and RED-100 reactor neutrino experiments Cañas et al. (2018) , which also rely on the detection of CE \nu NS events. While the optimistic precision of these experiments is expected to be \sim\!1\% , a comparable precision ( \lesssim\!3\% ) may also be achievable for Argo if the supernova neutrino flux from a CCSN at d\lesssim\!3{\,\text{kpc}} is available. Figure 5: Comparison between the expected precision of low-energy \sin{^2\theta_W} measurement by Argo and other measurements Navas et al. (2024) , at the 1\sigma confidence level, at different momentum transfers q . Measurements from recent CE \nu NS experiments Atzori Corona et al. (2022) ; De Romeri et al. (2023) ; De Romeri et al. (2025) are also shown in colors The measurement of low-energy \sin{^2\theta_W} also sheds light on searching for new physics. For example, the interactions between low-energy neutrinos and nucleons could potentially involve new physics beyond the SM, which can be described by the neutrino NSIs Bresó-Pla et al. (2023) ; Cadeddu et al. (2023) . As a result, these NSIs could introduce an effective shift in the SM value of \sin{^2\theta_W} , which can be identified by Argo due to the modification in the differential cross section of CE \nu NS. By considering the neutrino NSIs with u and d quarks described by the effective Lagrangian Davidson et al. (2003) ; Barranco et al. (2005) ; Scholberg (2006) , the weak charge Q_{W} and the weak form factor of {}^{40}\text{Ar} are modified Coloma et al. (2017) . Based on the anticipated high-precision measurements of \sin{^2\theta_W} at low energies, as illustrated in Fig. 4 , the sensitivity to the NSI parameters, such as \varepsilon_{\alpha\alpha}^{uV} and \varepsilon_{\alpha\alpha}^{dV} , could reach the range of (-0.006, 0.006) at the 90\% confidence level ( \chi^{2}=2.71 ) for a CCSN event located at d\!\sim\!3{\,\text{kpc}} . Such an improvement in sensitivity enhances the precision of constraining the neutrino NSI parameters by approximately one order of magnitude compared to the first COHERENT data Papoulias and Kosmas (2018) ; Coloma et al. (2017) ; Khan and Rodejohann (2019) ; Papoulias (2020) ; Giunti (2020) . Note that the uncertainty of distance is not included yet. Currently, the distances of some candidate pre-supernova stars can be measured with a precision of a few percent ( <5\% ) in Gaia missions Vallenari et al. (2023) ; Mukhopadhyay et al. (2020) ; Brown et al. (2018) . Recent research finds that distance estimate with neutrinos can also achieve similar precision Healy et al. (2024) . It is reasonable to expect that a dedicated multi-messenger observation in the future can achieve higher precision. Here we make a simple estimate of the uncertainty originating from the uncertainty of distance. We take \mathcal{N}_{\text{obs}} at d=3{\,\text{kpc}} as the observed data, then we vary d by \pm 1\% when calculating the theoretical results and uncertainties. The combined 1\sigma precision of \sin{^2\theta_W} changes to 4.1\% from 2.8\% without considering the distance uncertainty. We remark that the supernova neutrino flux is important in our evaluation. When we evaluate the total number of CE \nu NS events, we can simply use the time-integrated neutrino flux reconstructed from other observations, and its uncertainty is considered in \sigma_{\text{flux}} . The time-dependent neutrino flux from a well-calibrated supernova simulation would be necessary for the estimation of pile-up events. So, the uncertainty in the CCSN model has less influence on the final results since it only affects the estimate of pile-up events in a real measurement. In this work, nevertheless, we adopt 4 CCSN models to estimate the event pile-up in Argo detector. The results (see Fig. 3 ) indicate that the number of unresolved events depends on the progenitor mass, nuclear EoS, and source distance. Unlike the progenitor mass and source distance that will be measured via multi-messenger observations for a nearby target Mukhopadhyay et al. (2020) ; Pledger and Shara (2023) ; Vallenari et al. (2023) ; Healy et al. (2024) ; Brown et al. (2018) , the nuclear EoS properties remain uncertain and require further determinations Fiorella Burgio and Fantina (2018) ; Burgio et al. (2021) ; Lattimer (2021) . For example, the value of nuclear matter incompressibility is reported to be 240\pm 20{\,\text{MeV}} from isoscalar giant monopole resonance experiment, and is adopted in various relevant studies Youngblood et al. (1999) ; Shlomo et al. (2006) ; Chen and Gu (2012) ; Colo et al. (2014) ; Garg and Colò (2018) ; Cai and Li (2021) ; Li et al. (2021) . The incompressibility of the LS220 and Shen EoS, employed by these CCSN models, is 220 and 281{\,\text{MeV}} , respectively Lattimer and Swesty (1991) ; Shen et al. (1998) . Such choices represent two extremes with respective to the aforementioned experimental constraint. The difference in pile-up portion \langle\mathcal{N}_{\text{bias}}\rangle/\mathcal{N}_{\text{true}} between the CCSN models using the LS220 and Shen EoSs increases from 0.11\,(0.22)\% at d=5{\,\text{kpc}} to 1.73(2.19)\% at d=1{\,\text{kpc}} for the models with progenitor mass being 11.2\,(27.0) M ⊙ (see Fig. 3 ). Hence, we believe that the influence of the nuclear EoS uncertainties on our results may be noticeable at a very small d , and further analysis is required to study its impacts quantitatively. Note that, however, different values and confidence ranges of incompressibility are obtained using other approaches (see, e.g., Refs. Leifels et al. (2015) ; Huth et al. (2022) ; Sorensen et al. (2024) ). It is worth paying attention to the latest progress on constraining nuclear EoS through astrophysical observations and nuclear experiments Raaijmakers et al. (2021) ; Zhang and Li (2021) ; Pang et al. (2021) ; Radice et al. (2018) ; Abbott et al. (2019) . We expect an even smaller uncertainty from the nuclear EoS once the supernova simulation is further calibrated by future observations and experiments. Another issue is the occurrence of the next galactic CCSN. A list of 31 CCSN candidates within 1{\,\text{kpc}} from the Earth is suggested in Ref. Mukhopadhyay et al. (2020) . More candidates can be expected within a range of 5{\,\text{kpc}} . Meanwhile, the supernova rate is estimated to be 1.63\pm 0.46 per century for the Milky Way Galaxy Rozwadowska et al. (2021) . Nonetheless, it has been 420 years since the last observed galactic supernova, i.e., the Kepler supernova in 1604 Bethe (1990) , and we can tell that no galactic supernova has occurred for at least 38 years, according to the neutrino detection, since the extra-galactic SN 1987A. Among the last 1,000 years, only 6 galactic supernovae were recorded: Lupus at 2.2{\,\text{kpc}} in 1006, Crab at 2.0{\,\text{kpc}} in 1054, 3C 58 at 2.6{\,\text{kpc}} in 1181, Tycho at 2.4{\,\text{kpc}} in 1572, Kepler at 4.2{\,\text{kpc}} in 1604, and Cas A at 2.92{\,\text{kpc}} in 1680, all happened within 5{\,\text{kpc}} from the Earth. 4 of these events out of 6 , namely Crab, 3C 58, Kepler, and Cas A, were considered to be CCSNe The et al. (2006) . The next galactic supernova seems to be imminent and possible to be a CCSN. It is meaningful to get prepared to catch such an once-in-a-life-time chance. 5 Conclusions We study the low-energy \sin{^2\theta_W} measurement by Argo in the observation of a nearby CCSN explosion, through the CE \nu NS interaction. A galactic CCSN event provides an intense neutrino flux with a mean energy of \sim 10-20{\,\text{MeV}} , and will be observed by both Argo and other dedicated neutrino observatories. Given the great progress in neutrino detection in the last decades, the next nearby CCSN will be well observed and also be used to calibrate the CCSN simulations. The Argo detector can collect the CE \nu NS signals when those MeV neutrinos scatter off the argon nuclei, providing a way to extract the value of low-energy \sin^{2}{\theta_{W}} . We further consider the systematic uncertainties due to the flux extraction of supernova neutrinos and event pile-up, and find that a precision of \lesssim\!3\% is potentially possible at a supernova distance d\!\lesssim\!3{\,\text{kpc}} . Such a precision is better than the results from recent CE \nu NS experiments, and comparable to the optimistic expected sensitivity of ongoing experiments (e.g., the CONNIE, MINER, and RED-100 reactor neutrino experiments). Our analysis of the systematic uncertainties can provide valuable information for the experimentalists to improve their strategy in the observation of the next nearby CCSN. However, the caveat is that this result should be taken as an upper limit of the precision due to the currently incomplete uncertainty estimation. We leave the detailed analysis of the systematic uncertainties in a real measurement to future work. The precise measurement of the low-energy \sin{^2\theta_W} is important for the precision test of the SM. Any deviation may provide a hint to new physics, especially in the neutrino sector where the existence of new physics has already been proven by neutrino oscillation experiments. Specifically, such a measurement can be used to constrain the neutrino NSI parameters. Recently, the PandaX Bo et al. (2024) and XENON Aprile et al. (2024) collaborations reported the first detection of solar {}^{8}\text{B} neutrinos by direct DM detectors, through the CE \nu NS channel. Since the solar {}^{8}\text{B} neutrinos have been well-studied, future multi-ton scale DM detectors may also provide excellent opportunities to determine the low-energy \sin{^2\theta_W} via such measurements.