The phase function and density of the dust observed at comet 67P/Churyumov-Gerasimenko

The OSIRIS camera onboard Rosetta measured the phase function of both the coma dust and the nucleus. The two functions have a very different slope versus the phase angle. Here, we show that the nucleus phase function should be adopted to convert the brightness to the size of dust particles larger than 2.5 mm only. This makes the dust bursts observed close to Rosetta by OSIRIS, occurring about every hour, consistent with the fragmentation on impact with Rosetta of parent particles, whose flux agrees with the dust flux observed by GIADA. OSIRIS also measured the antisunward acceleration of the fragments, thus providing the first direct measurement of the solar radiation force acting on the dust fragments and thus of their bulk density, excluding any measurable rocket effect by the ice sublimation from the dust. The obtained particle density distribution has a peak matching the bulk density of most COSIMA particles, and represents a subset of the density distribution measured by GIADA. This implies a bias in the elemental abundances measured by COSIMA, which thus are consistent with the 67P dust mass fractions inferred by GIADA, i.e. (38 +/- 8) per cent of hydrocarbons versus the (62 +/- 8) per cent of sulphides and silicates.