[Previous] | [Session 39] | [Next]
K. Green, K. Lumme, A. Penttilä (Observatory, University of Helsinki)
Cosmic dust is almost everywhere in the solar system: on the surfaces of atmosphereless bodies, in the interplanetary space, and in the cometary comae. Dust plays an important role in star-forming regions, and is present in protostars, young stellar objects, up to main-sequence stars. However, there are significant variations in the structure of dust, arising, e.g., from different material parameters such as the complex refractive index or the geometry by which some dust aggregate has been built of constituent particles. All these parameters on different size scales obviously affect the observed electromagnetic field, and revealing even small pieces of that complicated mapping could provide valuable insights, e.g., into explaining why the maxima in linear polarization lie at different phase angles for planetary regoliths and comets.
Due to enormous computational burden, there are no truly general-purpose methods even for calculating the direct problem, i.e., the scattered field from a given structure, in the optical region. The time requirements in the differential and integral equation methods scale rather unfavorably with the system size: typical dust aggregates in random orientation are at present beyond the reach of supercomputing. Also, there are some resolution problems if weak polarization phenomena in the backscattering direction are being dealt with. The T-matrix method is non-approximative and applicable to relatively large single particles and small clusters of basic shapes. However, with non-symmetric and even slightly concave particles we have encountered the inherent numerical instabilities, which we are currently attacking using different quadratures and other variants of the T-matrix.
Our research group has been modelling the shapes of dust particles by suitably parametrized irregular polyhedral solids and stochastically deformed shapes such as spheres and cylinders with Gaussian perturbations. Stochastic aggregates have been formed by packing constituent particles by sampling a clustering process which is searching for a global minimum for the potential energy of the particle system.