GRB 230204B: GIT Discovery of a Fast Fading Afterglow Associated with an Energetic Gamma-Ray Burst from a Massive Star Progenitor

For the time-resolved analysis, we used the brightest NaI detector (NaI 8) and divided the light curve using Bayesian blocks with a false-alarm probability threshold of P = 0.01 (J. D. Scargle et al. 2013). To ensure sufficient signal-to-noise in the high-energy BGO 1 spectra, consecutive blocks were merged until a minimum significance of 5σ was achieved in each bin. Blocks were not combined across the four distinct emission episodes. The time intervals between episodes and the blocks with insufficient significance were excluded. This procedure resulted in a total of 16 time bins, as shown in the bottom panel of Figure 2.

We performed spectral fitting for each time bin using the PL+BB model and, for comparison, also fitted the Band function. Detailed fit parameters for both models are provided in the supplementary material (V. Swain et al. 2025).¹⁶

In our analysis, we find that the PL+BB model generally provided a better fit than the Band model across most time bins. However, in bin 8 of duration ∼16 s during Episode 3, the Band model showed a slightly better fit compared to the PL+BB model. This difference in their performance is indicated by the lower values of AIC and BIC, with differences of ΔAIC = 12 and ΔBIC = 12. The spectra for this bin revealed the highest peak energy of keV and a BB temperature of 66.56 ± 5.37 keV. This suggests that, during this interval, the prompt emission is increasingly dominated by nonthermal processes, while a thermal component may still be present but is less statistically favored.

In the Band model, the parameter α_p varies from a hard spectrum value of −0.002 to a soft spectrum value of −0.97, with an average of −0.62. The spectrum is hardest in bin 1, which indicates optically thick emission. The harder spectra, represented by higher values of α_p in the first three bins, suggest the presence of a BB component. As the episode progresses, the spectrum softens due to the expansion of the photosphere and the dominance of the nonthermal component (F. Ryde et al. 2010; B.-B. Zhang et al. 2011). The parameter E_p tracks the light curve intensity, ranging from 211 (± 23) to 970 (± 108) keV, with the highest value in bin 8 of Episode 3. The parameter β_p ranges from −3.39 to −2.15. The top three panels in Figure 4 illustrate the evolution of flux and spectral parameters derived from the Band model across the bins.

Zoom In Zoom Out Reset image size

For the PL+BB model, the PL index Γ_PL lies in the range from −1.4 to −1.7, while the BB temperature (kT) ranges from 39.6 (±2.5) to 121 (±7) keV. Both the peak energy and the BB temperature broadly track the intensity of the light curve and show a similar trend. We find that while the PL contributes more than half the flux in the full 10 keV–38 MeV range, the BB component dominates in the 10–1000 keV range (bottom four panels in Figure 4).

In contrast to the work of G26, which focused on Band and CPL models for their time-resolved analysis, our findings suggest that the PL+BB model is statistically preferred in most time bins. This model yields better constraints on the parameters, making it more suitable for exploring the photospheric emission and the physical properties of the jet.

The thermal emission originates near the photospheric radius (r_ph), where the optical depth drops to unity and photons decouple from baryon-associated electrons. In this section, we characterize the jet outflow at the photosphere. Such an analysis requires a segment of the light curve with significant photospheric emission and a freely expanding, nonmagnetized jet (F. Ryde 2004; F. Ryde & A. Pe’er 2009). At later times, the jet becomes optically thin, and the emission is typically dominated by nonthermal processes. Additionally, the curvature effect—arising from differences in photon arrival times and Lorentz boosting across a curved emitting surface—is more prominent in spectrally hard (thermal) bursts than in nonthermal ones (F. Ryde & A. Pe’er 2009). As a result, high-latitude emission can significantly contribute at late times. Thus, we exclude Episodes 3 and 4 from our photospheric analysis.

The first episode of GRB 230204Blasts for 6.5 s and shows significant photospheric emission; however, the signal-to-noise ratio is not high enough for a time-resolved analysis. In contrast, Episode 2 is both bright and prolonged, lasting approximately 34 s. This makes it an ideal candidate for detailed photospheric modeling using finer temporal binning.

We follow the method described in A. Pe’er et al. (2007), with assumptions supported by S. Iyyani et al. (2013) and F. Ryde et al. (2010). We further divide Episode 2 into 11 smaller time bins, such that the outflow can be approximated as quasi-static within each bin. The time-resolved analysis is performed using a PL+BB model, and the resulting data are available in the supplementary material at V. Swain et al. (2025).¹⁷ For the photospheric analysis, we model the outflow as adiabatic, following standard fireball dynamics under the assumption of quasi-thermal emission. In each time bin, we assume that a thermal component coexists with a nonthermal component. The evolution of the BB flux (F_BB), total flux (F), ratio F_BB/F, and the BB temperature (kT) are displayed in the upper panels of Figure 5.

Zoom In Zoom Out Reset image size

Using the observed BB flux and temperature, we can define the ratio () as follows:

where σ_SB is the Stefan–Boltzmann constant. For GRB 230204B, ranges from (0.76−1.86) × 10⁻¹⁹, remaining nearly constant at ∼10⁻¹⁹ during the early rising phase of the light curve and rising toward the end of the episode (Figure 5).

The photospheric radius (r_ph) and the bulk Lorentz factor (Γ_j) are two important parameters that characterize the outflow at the photosphere. In the fireball model, plasma expands from the initial radius (r₀), accelerating until it reaches the saturation radius (r_s). Beyond this point, the bulk Lorentz factor becomes constant, represented as , where L is the isotropic equivalent burst luminosity and is the mass ejection rate (A. Pe’er et al. 2007).

The location of the photosphere (r_ph) relative to r_s determines whether physical parameters like Γ_j and r_ph can be reliably estimated. If r_ph < r_s, the flow is still accelerating with Γ_j(r) ∝ r, and the relation r_ph/Γ_j = r₀ holds, but the individual values of Γ_j and r_ph remain unconstrained.

If the BB flux is subdominant to the total gamma-​​​​​​ray flux (F), i.e., F_BB ≪ F, then r_ph will be less than r_s (A. Pe’er et al. 2007; S. Iyyani et al. 2013). For GRB 230204B, the ratio F_BB/F ranges from 0.12 to 0.40 (shown in Figure 5), allowing reliable estimation of Γ_j and r_ph. In such a case, the bulk Lorentz factor is given by

where D_L is the luminosity distance, F is the total observed flux, σ_T is the Thomson cross section, m_p is the proton mass, and Y is the ratio of total fireball energy to observed gamma-ray energy, which ranges from 1 to 2. A lower value of Y indicates that the efficiency of converting the kinetic energy of the outflow into the nonthermal component is very high. Thus, for energetic GRBs, Y is expected to be 1. The photospheric radius (r_ph) can be estimated as

where is the isotropic equivalent burst luminosity.

As seen in Figure 5, for GRB 230204B the bulk Lorentz factor ranges from 400 to 626 and tracks the total flux intensity. On the other hand, the photospheric radius is almost constant around (0.5−1.0) × 10¹² cm.

The radius at the base of the outflow (r₀), which is independent of the cases (r_ph < r_s and r_ph > r_s), is given by

Finally, the saturation radius is defined as r_s = Γ_j × r₀, which depends on Γ_j. Thus, it can be defined for the case where r_ph > r_s.

For GRB 230204B, the base radius of the outflow, r₀, ranges from (0.4−1.3) × 10⁸ cm, and the saturation radius ranges from (2.0−6.7) × 10¹⁰ cm (indicating r_s < r_ph). We find that both r₀ and r_s strongly depend on the thermal contribution to the total flux.

The time-resolved photospheric parameters obtained for Episode 2 provide crucial insights into the physical state and evolution of the relativistic outflow. In the lower-left panel of Figure 6, we illustrate the relationship between the total flux and the bulk Lorentz factor, which follows the dependence . This relationship indicates that the total radiated flux is sensitive to changes in the Lorentz factor, consistent with expectations from emission due to internal shocks.

Zoom In Zoom Out Reset image size

Similarly, the lower-right panel of Figure 6 shows the correlation between the BB temperature and the Lorentz factor, following Γ_j ∝ T^(0.43±0.10). This suggests that the thermal and dynamical properties of the outflow evolve coherently during the prompt emission phase.

Conversely, the photospheric radius remains remarkably stable compared to the variations in temperature and flux. A similar characteristic was observed in the case of the thermally bright GRB 090902B (F. Ryde et al. 2010), supporting the interpretation that such bursts share a common photospheric emission mechanism.

We find that is negatively correlated with the BB temperature (kT) following for GRB 230204B, as shown in the upper right panel of Figure 6. As a consequence, despite the effective emitting area increasing at late times, the overall contribution of the BB flux declines due to the decreasing photospheric temperature. This trend is consistent with an expanding and cooling photosphere, which gradually becomes less significant as nonthermal emission processes become more prominent.

Overall, the photospheric properties of Episode 2 of GRB 230204B are consistent with a relativistic outflow that is initially weakly magnetized and dominated by quasi-thermal emission at early times. The gradual decline in thermal dominance, together with the stability of the photospheric radius and the correlated evolution of temperature and Lorentz factor, points to a transition from photospheric to nonthermal emission as the jet evolves and the photosphere cools. This behavior supports a scenario in which the jet is launched deep within a stellar progenitor and propagates adiabatically before becoming optically thin, giving rise to the observed interplay between thermal and nonthermal components in the prompt emission phase of the long GRBs.

In this section, we discuss three possible mechanisms responsible for producing nonthermal emission in this GRB and their connection to the observed photospheric component.

One possibility is internal shocks, where collisions between relativistic shells dissipate kinetic energy and produce nonthermal emission, typically via synchrotron radiation. Since both the thermal and nonthermal components are powered by the central engine, they are expected to be temporally correlated (F. Ryde et al. 2010), consistent with the behavior observed during the early episodes of GRB 230204B (Figure 4).

Alternatively, the correlation between the PL and BB components can be explained by inverse Compton processes, such as synchrotron self-Compton emission or the Compton upscattering of thermal photons (B.-B. Zhang et al. 2011). In this context, the low-energy PL component might originate from a different emission mechanism, such as synchrotron radiation. The positive slope of the PL component in νF_ν space suggests that a significant fraction of the emitted energy could be in the GeV range or higher. While the Fermi Large Area Telescope (LAT) has detected such emissions in thermally bright GRB (e.g., GRB 090902B; F. Ryde et al. 2010), LAT observations for GRB 230204B are not available due to its large boresight angle.

A third possibility is that the nonthermal PL component may arise from magnetic reconnection, where energy is transported via Poynting flux within the baryonic outflow. GRBs of this type, such as GRB 080916C, are expected to possess highly magnetized jets with purely nonthermal emission that is well described by the Band function (A. Pe’er & F. Ryde 2011). For GRB 230204B, the presence of a BB component in the observed spectrum indicates that the jet is either moderately or weakly magnetized (P. Veres et al. 2012), making this interpretation less likely.

GRBs exhibit several empirical correlations among prompt emission observables, which are often used to characterize the common behavior of the jet outflows and constrain the physical properties of their progenitors. Here, we compare the prompt emission properties of GRB 230204B with those of (energetic) LGRBs using rest-frame quantities derived from the time-integrated Band function parameters.

To enable a consistent comparison, we compute rest-frame quantities in the standard 1–10⁴ keV energy band using a k-correction. Detailed derivations are provided in Appendix B.1. For GRB 230204B, we obtain a k-correction factor of ∼0.97 for the observed 10–38,000 keV band. Using a fluence of S_γ = (1.9 ± 0.2) × 10⁻⁴ erg cm⁻² and a luminosity distance of D_L = 5.34 × 10²⁸ cm, we derive an isotropic equivalent energy of E_iso = (2.2 ± 0.4) × 10⁵⁴ erg. Similarly, from the observed peak energy E_p = 560 ± 55 keV, we calculate the rest-frame peak energy of E_p,z = E_p × (1 + z) = 1758 ± 172 keV.

We note that although GRB 230204B follows both the Amati and Yonetoku relations, its isotropic equivalent energy is significantly higher than that of typical LGRBs, ranking it among the most energetic events in this class. GRBs dominated by photospheric emission are expected to naturally follow the Amati relation (D. Lazzati et al. 2013), which aligns with our observations for GRB 230204B, where a prominent thermal component is detected during the prompt phase. These empirical correlations are also discussed in G26. In Appendix B.2, we provide a comparison of samples in the context of energetic long GRBs, along with the relevant plots.

We also estimate the initial bulk Lorentz factor (Γ_j,0) using the empirical Liang correlation (E.-W. Liang et al. 2010), which relates the isotropic equivalent gamma-ray energy to the initial Lorentz factor of the outflow:

Using the measured value of E_iso, we obtain Γ_j,0 ∼ 700 for GRB 230204B. This estimate is consistent with the bulk Lorentz factors inferred independently from the time-resolved photospheric analysis (Section 3.2), supporting a highly relativistic outflow.

D. A. Frail et al. (2001) found that in pre-Swift GRBs, the jet opening angle θ₀ seems to be anticorrelated with E_iso through , so that the jet-corrected gamma-ray energy is roughly constant, ≈10⁵¹ erg for LGRBs. Subsequent studies have shown that this characteristic energy can extend up to ∼10⁵² erg, particularly for energetic GRBs (S. B. Cenko et al. 2010). The implication is that wider jets tend to have a lower energy concentration, while narrow jets have a higher energy concentration. Using this relation, we estimate a jet opening angle of θ₀ ∼ (1.7° − 5.5°), suggesting a relatively collimated outflow.

R. A. Chevalier & Z.-Y. Li (2000) extensively discusses the afterglow of a GRB exploding in a wind-type medium. In this section, we examine the relevant closure relations for such a medium, summarized schematically in Figure 8. The dashed diagonal line denotes the cooling frequency increasing with time (ν_c ∝ t^1/2), while the dotted diagonal shows the characteristic synchrotron frequency decreasing with time (ν_m ∝ t^−3/2). The epoch at which these frequencies are equal (ν_c = ν_m = ν_cm) marks the transition from the fast- to slow-cooling regime and typically occurs at very early times (R. Sari et al. 1998). Since the optical afterglow of GRB 230204B exhibits a single PL decay without any detectable temporal break, we restrict our analysis to the slow-cooling regime (t > t_cm).

Zoom In Zoom Out Reset image size

In the X-ray band, the afterglow emission generally satisfies ν > ν_cm. In this case, the temporal decay is expected to follow α = (3p − 2)/4 for ν_m < ν_c < ν, with a mild steepening to α = (3p − 1)/4 (ν_m < ν < ν_c) once ν_c crosses the observing band.

In the optical band, two scenarios are possible: either ν > ν_cm, analogous to the X-ray regime, or ν < ν_cm. In the latter case, the early-time light curve is expected to exhibit a plateau phase (ν < ν_m < ν_c), followed by a steeper decay with α = (3p − 1)/4 (ν_m < ν < ν_c) once ν_m drops below the observing frequency. Notably, the late-time decay index is identical in both scenarios. This degeneracy can be resolved by observing the afterglow over a broad wavelength range, as the passage of spectral break frequencies (ν_m or ν_c) through different bands creates distinct temporal signatures that help constrain the frequency ordering.

The optical afterglow of GRB 230204B follows a single PL decay out to ∼30.8 hr, indicating that the observations likely probe the final stages of the decay phase (ν_m < ν < ν_c). Under this assumption, we infer an electron PL index of p = 2.76 ± 0.01, consistent with expectations from the synchrotron forward-shock model.

At radio frequencies, synchrotron self-absorption plays a crucial role in shaping the light curve. In a wind-type medium, the self-absorption frequency ν_a decreases with time and eventually crosses the characteristic frequency ν_m. However, this transition typically occurs at very late times (B. Zhang et al. 2022). Consequently, during the epochs probed by our radio observations, the afterglow is expected to satisfy ν < ν_cm. In this case, the radio light curve will exhibit a nearly constant radio flux, followed by a late-time steeper decay, a hallmark of a wind-type environment. This behavior is indeed observed in our data (Figure 7).

Based on the various considerations in Section 6.1, we now fit our afterglow data to calculate the physical parameters of the outflow. The additional parameter A_⋆ characterizes the density normalization of the wind-type CBM, following the definition from R. A. Chevalier & Z.-Y. Li (2000). We assume an adiabatic top-hat jet with a high initial Lorentz factor, observed within the jet opening angle. In our modeling, we fix χ = 1, corresponding to efficient particle acceleration. As no jet break is detected in the available data, the jet opening angle remains weakly constrained.

For a given set of physical parameters, the characteristic synchrotron frequencies ν_m, ν_c, and ν_a are computed, allowing the model flux to be evaluated as a function of time and observing frequency, following J. Granot & R. Sari (2002). We infer the model parameters using nested sampling via the MultiNest algorithm, implemented through the PyMultiNest package (J. Buchner et al. 2014), with 2000 live points. Uniform priors for the model parameters are listed in Table 2. The posterior distributions are obtained by maximizing the likelihood of the model given the observed data, and the resulting parameter estimates and credible intervals are reported in Table 2 (Posterior-1).

Table 2. Priors and Posteriors of the Model Parameters

Parameter	Unit	Prior Type	Parameter Bound	Posterior-1	Posterior-2	Posterior-3
					(Fixed e and b)
	erg	Uniform	[53, 58]
	⋯	Uniform	[-2, 0]
	⋯	Uniform	[-2.0, -0.5]		−1	−1
	⋯	Uniform	[-6, -1]		−4	−5
p	⋯	Uniform	[2, 3]	2.76  ±  0.01	2.76  ±  0.01	2.76  ±  0.01

Note. The best-fit values are reported with their 95% confidence intervals.

Download table as:  ASCIITypeset image

In our initial run, we found significant degeneracies among the model parameters. The resulting posterior distributions span a wide range of values: E_K,iso ∈ (10⁵⁴−10⁵⁷) erg, _e ∈ 0.03−​​​​​​0.23, _b ∈ 10⁻⁶−10⁻⁴, and A_⋆ ∈ 0.01​​​​​​−0.77, all within the 95% confidence interval. However, the posterior for the electron energy index, p, is well constrained, yielding p = 2.76 ± 0.01, and is consistent with the analytic estimate derived in Section 6.1.

To reduce parameter degeneracies, we adopt a typical value of _e ∼ 0.1 (A. Panaitescu & P. Kumar 2002; S. B. Cenko et al. 2010). For the parameter _b, B. Zhang et al. (2007) place an upper limit of _b ≲ 10⁻⁴ for X-ray afterglows in the regime ν_m < ν_X < ν_c, which is expected for a wind-like CBM. We therefore explore two representative values of _b, 10⁻⁴ and 10⁻⁵, in our subsequent analysis.

With _e and _b fixed, the posterior samples yield E_K,iso = (2.7 ± 0.5) × 10⁵⁴ erg for _b = 10⁻⁴ and E_K,iso = (2.7 ± 0.5) × 10⁵⁵ erg for _b = 10⁻⁵. In both cases, we obtain A_⋆ = 0.15 ± 0.02 and p = 2.76 ± 0.01. The results are summarized in the Posterior-1 and Posterior-2 columns of Table 2.

The inferred light curve for the best-fit model corresponding to _e = 0.1 and _b = 10⁻⁴, overlaid on the observed data, is shown in the upper panel of Figure 9. The dashed line represents the model light curve based on the median parameter values, while the shaded region indicates the 95% credible interval. The temporal evolution of the characteristic frequencies, ν_a, ν_m, and ν_c, computed from the median parameters, is shown in the lower panel. As discussed in Section 6.1, the characteristic frequency ν_m crosses the optical band at very early times, before our first observation, and passes through the 10 GHz band at ∼48 days after the explosion, resulting in a slight decrease in the radio flux. The cooling frequency ν_c remains above all observational bands throughout the period of interest.

Zoom In Zoom Out Reset image size

For a wind-like medium, the inferred kinetic energy depends sensitively on the assumed microphysical parameters, particularly _e, _b, and A_⋆, leading to substantial uncertainties in the radiative efficiency. For GRB 230204B, adopting _b ≲ 10⁻⁴ and E_K,iso ≳ 2.7 × 10⁵⁴ erg yields an upper limit on the radiative efficiency of ∼45%. Lower values of _b imply larger kinetic energies and correspondingly lower efficiencies; for example, adopting _b = 10⁻⁵ yields E_K,iso ∼ 2.7 × 10⁵⁵ erg and a radiative efficiency of ∼7.5%. Overall, our modeling suggests a moderately broad efficiency range that is consistent with values typically inferred for energetic GRBs.

The afterglow modeling by G26 inferred a lower magnetic field fraction (_b ∼ 10⁻⁶) and a higher kinetic energy (E_K,iso ∼ 4 × 10⁵⁵ erg), resulting in a lower radiative efficiency of ∼4%. The differences between these estimates highlight the sensitivity of radiative efficiency to assumptions about the circumburst environment and the adopted microphysical parameters.

Overall, the inferred parameters and the modeled light curves provide a consistent description of the steeply decaying optical afterglow and the late-time radio plateau of GRB 230204B, supporting a relativistic jet expanding into a wind-type CBM. Our best-fit model yields p = 2.76 ± 0.01, _e = 0.1, and A_⋆ = 0.15 ± 0.02, with an upper limit of _b ≲ 10⁻⁴ and a lower limit on the isotropic equivalent kinetic energy of E_K,iso ≳ 2.7 × 10⁵⁴ erg.

As discussed in Section 6.1, the optical afterglow of GRB 230204B exhibits a steep decay with α ≃ 1.82 in the ν_m < ν_opt < ν_c regime, significantly steeper than that observed in a typical LGRB, as shown in Figure 10. In both wind-type and uniform density environments, the transition from fast to slow cooling occurs at very early times. For typical parameters, the cooling frequency lies above the optical band during most of the afterglow evolution, implying that the observed optical emission is powered by electrons that do not cool efficiently. Consequently, the optical light curve of most GRBs, including GRB 230204B predominantly traces the adiabatic slow-cooling segment ν_m < ν_opt < ν_c.

Zoom In Zoom Out Reset image size

The temporal decay of the afterglow in this regime depends on the surrounding medium. For a uniform density medium, the expected decay index is α = (3p − 3)/4, whereas a wind-type medium predicts a steeper decay, α = (3p − 1)/4. For a given electron PL index p, the afterglow in a wind environment therefore decays faster by Δα = 0.5 compared to the uniform density case.

This rapid decay can be understood by examining the peak synchrotron flux , which represents the maximum emissivity near the frequency ν_m in the slow-cooling regime (using the definition from R. Sari et al. 1998). In a wind-type medium, this peak flux decreases with time (scaling roughly as t^−1/2), while in a uniform density medium, it remains constant. The physical interpretation is that, in a uniform medium, the jet interacts with a constant-density environment and continuously sweeps up new material at a steady rate. In contrast, in a wind-type environment, the density decreases as r⁻², so the jet encounters progressively less material as it expands, leading to less radiation dissipation and creating a steeper decay in the afterglow flux.

While the jet remains relativistic, the observed light curve can be adequately modeled using a spherical expansion framework. However, once the jet decelerates sufficiently, edge effects become prominent due to the lateral expansion of the jet (J. E. Rhoads 1999; R. Sari & T. Piran 1999). This transition results in a significantly steeper light-curve evolution during the edge expansion phase. The edge effect becomes observable when the Lorentz factor of the jet becomes comparable to the inverse of the jet’s angular half-opening angle (Γ_j ∼ 1/θ₀). Following R. A. Chevalier & Z.-Y. Li (2000), the characteristic time at which edge effects become visible is

where θ₀ is the jet half-opening angle in radians, E₅₂ is the isotropic equivalent kinetic energy (E_K,iso) in units of 10⁵² erg, and A_⋆ is the wind parameter.

The analysis of the GRB 230204B light curve indicates that, up to the final radio detection at ∼87 days, no significant steepening in the light curve was observed. This sets a lower limit on the jet opening angle of θ₀ > 48 for an isotropic kinetic energy of E_K,iso ∼ 2.7 × 10⁵⁴ erg. This limit is consistent with our independently derived estimate of θ₀ ∼ (17–55), obtained in Section 4.3.

S. B. Cenko et al. (2010) found that the highly energetic GRBs can release a true energy of E_γ ≳ 10⁵² erg. For GRB 230204B, considering the collimated jet adopting θ₀ ∼ 5°, E_K,iso ≳ 2.7 × 10⁵⁴ erg, and E_iso ∼ 2.2 × 10⁵⁴ erg, we estimate a total energy release of E_γ ≳ 1.9 × 10⁵² erg. These results place GRB 230204B among the class of hyper-energetic GRBs.

Considering the observed wind interaction in GRB 230204B, the most plausible scenario is that the progenitor of GRB 230204B was a W-R star. Galactic W-R stars typically exhibit wind velocities of ∼1000 km s⁻¹ and mass-loss rates of ∼10⁻⁵ M_⊙ yr⁻¹ (S. Woosley & J. Bloom 2006). However, our derived wind parameter A_⋆ ∼ 0.15 for GRB 230204B implies a lower mass-loss rate of (1.5 ± 0.2) × 10⁻⁶ M_⊙ yr⁻¹. Mass loss in the W-R phase plays a critical role in determining the final fate of the star, as excessive loss can strip the core, favoring neutron star formation and preventing the formation of the massive, rapidly rotating core needed for GRB production. Moderate mass-loss rates, on the other hand, preserve sufficient core mass and angular momentum, aligning with the collapsar model (S. Woosley & J. Bloom 2006). Metallicity further influences these outcomes, as low-metallicity W-R stars have reduced mass-loss rates and are more likely to retain the rapid rotation necessary to power a jet (A. I. MacFadyen et al. 2001; S. Woosley & J. Bloom 2006). Thus, the mass-loss rate inferred for GRB 230204B is consistent with this theoretical framework, supporting a low-metallicity W-R progenitor that underwent moderate mass loss while retaining sufficient core mass and angular momentum to produce a successful GRB.

In this section, we examine the correlation between the prompt and afterglow properties. The details of the samples collected for this analysis are presented in Appendix C.

Figure 10 presents a comparative analysis of GRB 230204B against a larger sample of LGRB. Notably, GRB 230204B exhibits a steeper decay during the adiabatic phase (ν_m < ν < ν_c) of the afterglow compared to most LGRBs. We find that α is correlated with isotropic equivalent energy and weakly correlated with the spectral peak energy. We also find that the r-band luminosity at 0.5 day after the burst is strongly correlated with the isotropic equivalent energy:

These trends are broadly consistent with findings reported by S. R. Oates et al. (2015). GRB 230204B satisfies the observed correlation between the afterglow temporal decay index (α) and the isotropic equivalent prompt energy (E_iso) within the 95% confidence region, as illustrated in Figure 10. Its relatively high α value is consistent with expectations for a wind-type CBM, which typically produces steeper afterglow decay. In the α–E_p,z plane, GRB 230204B lies near the boundary of the 2σ confidence region, indicating that the E_p,z is lower than expected. On the other hand, in the L_r–E_iso relation, GRB 230204B clearly satisfies the correlation, owing to both the bright optical afterglow and the energetic prompt emission of this GRB.

These correlations are not predicted by the standard fireball model, which treats prompt and afterglow phases as physically decoupled, with the former arising from internal dissipation within the jet and the latter from external shocks with the CBM (R. Sari et al. 1998; P. Kumar & B. Zhang 2015). However, if both phases originate from a common energy reservoir injected by the central engine, then more energetic GRBs could naturally produce both brighter prompt emission (a few percent of the total energy) and more luminous afterglows powered by the residual kinetic energy.

Furthermore, the brightness and temporal decay of GRB afterglows are governed by several key factors: (i) the density profile of the CBM, (ii) the energetics and internal structure of the jet, and (iii) geometric effects, such as the observer’s viewing angle relative to the jet axis. The first two factors are closely related to the properties of the progenitor and the formation of the central engine. As a result, this leads to the expectation of correlations between the prompt emission and the afterglow observables.

In the collapsar framework, the mass-loss rate and metallicity of the progenitor star are crucial factors in determining whether a black hole forms and whether enough angular momentum is retained to generate a relativistic jet through an accretion disk (S. Woosley & J. Bloom 2006). These parameters simultaneously shape the jet energetics and the CBM into which the jet propagates. In contrast, the viewing angle is an extrinsic parameter and primarily introduces scatter into these correlations.

Based on our analysis, GRB 230204B is most plausibly associated with a W-R progenitor that generated a wind-type CBM while retaining sufficient core mass to collapse into a black hole capable of launching an energetic relativistic jet. This scenario naturally explains the highly luminous prompt emission, followed by a bright optical afterglow with a steep decay slope. The combined prompt and afterglow properties of GRB 230204B are broadly consistent with the empirical correlations observed in our comparison sample of GRBs.

Fermi-GBM was triggered at time T₀ = 21: 47: 51 UT on 2023 Feb 04 (S. Poolakkil et al. 2023). Here, we report the results of an analysis of the Fermi-GBM prompt data. GBM has 12 NaI detectors (8 keV–1 MeV) and two BGO detectors (200 keV–40 MeV). We choose the detectors according to the brightness of the spectral counts and the angle of the detector’s boresight to the source location. The source was detected as bright in the NaI 8 detector, at 18° away from the source. For analysis, we used two NaI detectors, NaI 7 and NaI 8, and a single BGO detector, BGO 1. The angle from the Fermi-LAT boresight at the GBM trigger time is 106°; thus, LAT data are not included in the analysis (M. Ackermann et al. 2012). The background-subtracted light curves are shown in Figure 2. The duration of the GRB is T₉₀ = 216 s in the energy range of 50–300 keV. The strongest peak occurred at T₀ + 156 s in the energy range of 10–1000 keV with a photon flux of 7.3 ± 0.4 ph s⁻¹ cm⁻² (S. Poolakkil et al. 2023).

The source was also seen by AstroSat-CZTI (V. Bhalerao et al. 2017), where both the CZT modules (20–200 keV) and the Veto detectors (100–500 keV) detected multiple peaks of the emission (G. Waratkar et al. 2021). The strongest peak in the CZT detectors occurred at T₀ + 51.5 s, whereas in the Veto detectors the strongest peak occurred at T₀ + 50.8 s. Our detailed reanalysis of the data shows that the peak count rate in the CZT was 548 ± 50 counts s⁻¹ above background, with a total of counts and a T₉₀ duration of s, consistent with the Fermi-GBM T₉₀. In the Veto, the peak count rate was above background, with a total of counts and a T₉₀ of s. A total of 3240 Compton events were associated with the burst; however, the detector was at a boresight angle of 75°, which is too large to perform polarization analysis (T. Chattopadhyay et al. 2022).

We used GIT located at the Indian Astronomical Observatory (IAO), Hanle-Ladakh, to acquire data of GRB 230204B optical afterglow (H. Kumar et al. 2022). GIT is a 0.7 m, wide-field, fully robotic telescope specifically designed for the study of transient astrophysical events. The afterglow was observed in a Sloan filter starting at 1.57 hr after the GRB detection at 2025-02-04T21:47:51 UT by MAXI-GSC (M. Serino et al. 2023). Data were downloaded and processed in real time by the GIT data reduction pipeline. Basic reduction, astrometry, and point-spread function photometry followed the procedure described in H. Kumar et al. (2022). The final magnitudes were calibrated against Pan-STARRS.

Once the afterglow was identified, we used the Rainbow Camera on the Spectral Energy Distribution Machine (SEDM; N. Blagorodnova et al. 2018; M. Rigault et al. 2019) mounted on the P60 telescope to acquire r- and i-band imaging in 300 s exposures, starting 10.88 hr after the burst. The SEDM images were processed with a Python-based pipeline version of Fpipe (Fremling Automated Pipeline; C. Fremling et al. 2016), which includes photometric calibrations. Afterglow spectroscopy was attempted, but unfortunately, the weather conditions did not allow for good signal-to-noise observations.

We conducted multiple follow-up observations with the Giant Metrewave Radio Telescope (GMRT) and the Very Large Array (VLA) at different epochs following the initial trigger. The GMRT observations, carried out on 2023 February 22, obtained an upper limit in Band 5 (1000–1450 MHz). The VLA observations, beginning on 2023 February 15, and extending over several epochs, yielded multiple detections in the X band (8–12 GHz; central frequency 10 GHz) and upper limits in the S band (2–4 GHz). A summary of the measured fluxes and upper limits from all observing epochs is presented in the supplementary material at V. Swain et al. (2025).¹⁹

Observations with Swift-XRT began approximately 80.6 ks after the trigger time. For the light curve analysis, we utilized the results publicly available from the UK Swift Science Data Centre (UKSSDC).²⁰ In the afterglow analysis, we considered two data points of unabsorbed flux density at 10 keV, and ignored the very weak upper limit at the third epoch.

We reanalyzed the data of the first interval to obtain some spectral parameters. Since the optical light curve showed no evidence of flares or temporal breaks, we performed a time-integrated analysis over the interval from T₀ + 80.6 ks to 110.0 ks in the 0.3–10 keV energy band. The X-ray data were obtained from Swift observations as described by P. A. Evans et al. (2007). The total number of observed counts was 17, which is insufficient for spectral fitting due to the large number of free parameters. Therefore, we fixed the value of the intrinsic hydrogen column density to the mean value of N_H = 2.2 × 10²¹ cm⁻² as reported by P. A. Evans et al. (2007). With this assumption, we performed a likelihood-based spectral fit using ThreeML, yielding a photon index of 1.4 ± 0.4. It corresponds to a spectral index of β = 0.4 ± 0.4. The resulting flux is 2.3 × 10⁻¹³ erg s⁻¹ cm⁻² in the 0.3–10 keV band. The results are consistent with the automated analysis provided by the UKSSDC.

To compare individual events in the broader GRB population, it is essential to calculate physical properties in a common rest-frame energy band, typically 1–10⁴ keV. This requires transforming the observed quantities to the source rest frame using a k-correction factor, defined as

where N(E) is the spectral model in units of counts per unit of energy, and (e₁, e₂) brackets the energy band of the detector.

The isotropic equivalent energy in the rest frame, E_iso, is calculated as

where S_γ is the gamma-ray fluence, D_L is the luminosity distance, z is the redshift of the source, and k is the k-correction factor (J. S. Bloom et al. 2001). The “1 + z” term in the denominator corrects for the cosmological time dilation of the burst duration, while the k-correction accounts for the spectral shift, effectively converting the observed fluence into the rest-frame energy (typically 1–10,000 keV).

We used the catalog from L. Lan et al. (2023) to study this GRB in relation to the Amati relation (L. Amati et al. 2002), which describes the correlation between the isotropic equivalent bolometric energy of the GRB and its rest-frame peak energy. This correlation shows a positive dependence, as illustrated in Figure 11(a), and is typically expressed as follows:

where a lies between 0.4 and 0.6. This relationship suggests that energetic GRBs tend to release most of their energy in a harder spectrum. Notably, while GRB 230204B follows this relation, its isotropic energy (E_iso) is significantly higher than that of typical LGRBs, placing it in the category of highly energetic GRBs.

Zoom In Zoom Out Reset image size

We further explored the Yonetoku relation using the catalog given in D. Yonetoku et al. (2010). It is a tighter relation (Equation (B4)) between the peak energy in the rest frame and the peak luminosity of the GRB’s prompt emission. The relation can be defined as

Our analysis yields a best-fit slope of b = 1.55 ± 0.08 and an intercept of C = −3.52, as illustrated in Figure 11(b). The figure shows that the energetic GRBs exhibit higher values of both E_p,z and L_p. However, they are not distinctly separated from the rest of the LGRBs. For GRB 230204B, we obtain L_p = (8.7 ± 2.5) × 10⁵² erg s⁻¹ at T₀ + 156 s. This burst conforms well to the Yonetoku relation and aligns with the population of energetic GRBs.