Assessment of comet Shoemaker-Levy 9 fragment sizes using light curves measured by Galileo spacecraft instruments

Two- and three-dimensional numerical radiation-hydrodynamic simulations of the bolide stage of SL-9 fragments intruding into the Jovian atmosphere have been conducted. Radiation fluxes in each wave band of the Galileo mission instruments (SSI, PPR, NIMS and UVS) in the direction of Galileo have been calculated. The simulations are based on detailed tables of spectral opacities calculated assuming thermodynamic equilibrium and the initial composition of 0.89H2+0.11He+0.00195CH4. The small amount of methane and its products strongly change the spectral absorption coefficients in the IR and visible region, compared with a pure hydrogen helium mixture. Simulations begin at the moment when the bolide head is at an altitude of 300 km. The intensity of light rapidly increases until the moment when the bolide head disappears below the clouds, but the wake still shines above the cloud tops. This changes the slope of the light intensity versus time curve and creates a plateau on the light curve. A small difference in the observed maximum intensity for K, W and N fragments registered by SSI and for L, G, H and Q1 fragments registered by PPR instruments (no more than 2.5 times) implies that the sizes of all the above-mentioned fragments vary by no more than a factor of about 1.5. A comparison of the theoretical peak intensity determined neglecting ablation with that observed for all these fragments gives an estimate of the fragment radius in the range 0.5–0.7 km. NIMS data at λ = 4.38 μm for the G fragment give the same value of radius. The radiating volume is located mainly at altitudes of about 30–100 km and slowly changes its size and shape. The brightness temperature is much smaller than the temperature of gas in the bolide head and in the hot core of the wake. This effect of screening due to absorption in cold outer layers of the wake explains a rather low brightness temperature of PPR and NIMS. A short duration of the signal in the ultraviolet is also explained by the simulations. Simulations taking into account radiation-driven ablation show that the dense vapor cloud formed at high altitudes tends to increase the shock wave radius by a factor of two or three. Taking this correction factor into account we can expect the fragment radius to be 0.3–0.4 km for K and 0.15–0.20 km for W. The theoretical profile of the light curve before and some time after the peak intensity is similar to the observed one (for K fragment until 15–20 s). The discrepancy at later time is probably due to the heating and evaporation of the clouds by the shock wave and mechanically induced ablation, which have not yet been taken into account in the physical model.