The energy budget of GRBs based on a large sample of prompt and afterglow observations
We compare the isotropic equivalent 15–2000 keV γ-ray energy, E {sub γ}, emitted by a sample of 91 swift Gamma-Ray Bursts with known redshifts, with the isotropic equivalent fireball energy, E {sub fb}, as estimated within the fireball model framework from X-ray afterglow observations of these bursts. The uncertainty in E {sub γ}, which spans the range of ∼10{sup 51} to ∼10{sup 53.5} erg, is ≈25% on average, due mainly to the extrapolation from the BAT detector band to the 15–2000 keV band. The uncertainty in E {sub fb} is approximately a factor of 2, due mainly to the X-ray measurements’ scatter. We find E {sub γ} and E {sub fb} to be tightly correlated. The average(std) of η{sub γ}{sup 11 hr}≡log{sub 10}(E{sub γ}/(3ε{sub e}E{sub fb}{sup 11 hr})) are −0.34(0.60), and the upper limit on the intrinsic spread of η {sub γ} is approximately 0.5 (ε{sub e} is the fraction of energy carried by electrons and E{sub fb}{sup x hr} is inferred from the X-ray flux at x hours). E{sub fb}{sup 3 hr} and E{sub fb}{sup 11 hr} are similar, with an average(std) of log{sub 10}(E{sub fb}{sup 3 hr}/E{sub fb}{sup 11 hr}) of 0.04(0.28). The small variance of η {sub γ} implies that burst-to-burst variations in ε{sub e} and in the efficiency of fireball energy conversion to γ-rays are small, and suggests that both are of order unity. The small variance of η {sub γ} and the similarity of E{sub fb}{sup 3 hr} and E{sub fb}{sup 11 hr} further imply that deviations from a simple fireball model description, if present, are small. This puts stringent constraints on models incorporating such modifications (due e.g., to radiative losses, energy injection, off-axis viewing).