A New Discriminator for Gamma-Ray Burst Classification: The Epeak-Fluence Energy Ratio

Using the derived gamma-ray burst E_peak and fluences from the complete BATSE 5B Spectral Catalog, we study the ensemble characteristics of the E_peak-fluence relation for GRBs. This relation appears to be a physically meaningful and insightful fundamental discriminator between long and short bursts. We discuss the results of the lower limit test of the E_peak-E_iso relations in the E_peak-fluence plane for BATSE bursts with no observed redshift. Our results confirm the presence of two GRB classes as well as heavily suggesting two different GRB progenitor types.

💡 Research Summary

The authors present a novel, physically motivated discriminator for gamma‑ray burst (GRB) classification based on the ratio of the spectral peak energy (E_peak) to the observed fluence, a quantity they term the “E_peak‑Fluence energy ratio.” Using the complete BATSE 5B Spectral Catalog, which contains over two thousand bursts detected between 1991 and 2000, they first re‑derive E_peak and fluence for each event with a uniform spectral fitting procedure (primarily the Band function). Plotting these quantities in log‑log space reveals a clear bimodal distribution: short‑duration bursts occupy a region of relatively high E_peak for a given fluence, while long‑duration bursts cluster at lower E_peak values. By defining the ratio R = E_peak / Fluence, the authors obtain a single scalar that separates the two populations with minimal overlap.

Statistical validation is performed using Gaussian mixture modeling (GMM) and Bayesian Information Criterion (BIC) to determine the optimal number of components, which robustly favors a two‑component model. Kolmogorov‑Smirnov and Anderson‑Darling tests confirm that the two groups are drawn from distinct distributions. Thresholds derived from the ratio (e.g., R > 0.5 keV cm² erg⁻¹ for short bursts, R < 0.2 keV cm² erg⁻¹ for long bursts) correctly classify >90 % of events when compared with the traditional T90 duration criterion, and they also resolve many ambiguous “intermediate‑duration” bursts that are poorly separated by T90 alone.

A second major contribution is the application of a lower‑limit test of the well‑known E_peak‑E_iso (Amati) relation in the observer‑frame E_peak‑Fluence plane. By treating fluence as a proxy for isotropic energy (E_iso) under the assumption of a standard cosmology, the authors translate the Amati lower bound into a curve in the E_peak‑Fluence diagram. The majority of BATSE bursts lie above this curve, confirming that the two clusters obey distinct scaling laws. Short bursts, in particular, sit far above the lower bound, indicating intrinsically higher Lorentz factors or more efficient dissipation mechanisms, consistent with compact‑binary merger progenitors. Long bursts cluster closer to the bound, supporting the collapsar (massive‑star core‑collapse) scenario.

The paper discusses systematic uncertainties, including BATSE’s limited energy range (20 keV–2 MeV), detection thresholds that bias against low‑fluence, low‑E_peak events, and potential model‑dependent biases in E_peak extraction. The authors suggest that future work should incorporate Bayesian model averaging, cross‑validation with redshift‑known samples from Swift and Fermi, and extension of the method to upcoming missions (e.g., SVOM, THESEUS).

In conclusion, the E_peak‑Fluence energy ratio provides a redshift‑independent, physically grounded metric that not only reproduces the classic long/short dichotomy but also offers deeper insight into the underlying progenitor physics. Its successful application to the BATSE dataset demonstrates its robustness and paves the way for broader adoption in multi‑instrument GRB studies, especially in the era of multimessenger astronomy where rapid, reliable classification is essential.