What drives the dust activity of comet 67P/Churyumov-Gerasimenko?

Context. The gas-driven dust activity of comets is still an unsolved question in cometary physics. Homogeneous dust layers composed of micrometer-sized grains possess tensile strengths of ~1 kPa, which is far higher than typical gas pressures caused by the sublimation of the ices beneath the covering dust layer. This implies that the dust grains cannot be detached from the surface by the gas pressure of the sublimating ices. One possibility to avoid this problem is that the nucleus formed through the gravitational collapse of an ensemble of millimeter- to centimeter-sized aggregates. In this case, an aggregate layer with a tensile strength on the order of ~1 Pa is formed on the surface of the nucleus, which allows for the release of the aggregates from the surface by the gas pressure build up at the ice-dust interface. Aims. We use the gravitational instability formation scenario of cometesimals to derive the aggregate size that can be released by the gas pressure from the nucleus of comet 67P/Churyumov-Gerasimenko for different heliocentric distances and different volatile ices. Methods. To derive the ejected aggregate sizes, we developed a model based on the assumption that the entire heat absorbed by the surface is consumed by the sublimation process of one volatile species. The calculations were performed for the three most prominent volatile materials in comets, namely, H2O ice, CO2 ice, and CO ice. Results. We find that the size range of the dust aggregates able to escape from the nucleus into space widens when the comet approaches the Sun and narrows with increasing heliocentric distance, because the tensile strength of the aggregates decreases with increasing aggregate size. The activity of CO ice in comparison to H2O ice is capable to detach aggregates smaller by approximately one order of magnitude from the surface. As a result of the higher sublimation rate of CO ice, larger aggregates are additionally able to escape from the gravity field of the nucleus. Conclusions. Our model can explain the large grains (ranging from 2 cm to 1 m in radius) in the inner coma of comet 67P/Churyumov-Gerasimenko that have been observed by the OSIRIS camera at heliocentric distances between 3.4 AU and 3.7 AU. Furthermore, the model predicts the release of decimeter-sized aggregates (trail particles) close to the heliocentric distance at which the gas-driven dust activity vanishes. However, the gas-driven dust activity cannot explain the presence of particles small