A database of single scattering properties for hydrometeors at microwave and sub-millimetre frequencies

Doctoral thesis, 2020

Remote sensing observations of hydrometeors (ice or liquid water particles) at microwave and sub-millimetre wavelengths provide important input to numerical weather forecasting through data assimilation and give insight to cloud processes that are relevant for climate prediction. The utilization of such measurements requires information on the single scattering properties (SSP), i.e., knowledge on how single hydrometeors scatter, absorb, and emit radiation. However, SSP are dependant on the particle orientation, shape, and size which in the case of ice hydrometeors are highly variable in nature. Furthermore, simulating the SSP of hydrometeors is challenging and computationally costly. These are the main challenges that this thesis aims to address.

In the first study of this thesis, a new publicly available SSP database for randomly oriented ice hydrometeors was developed. In terms of covered frequencies, temperatures, sizes, and particle models it is the most extensive to date. Particle models include aggregates that were generated using a semi-realistic, stochastic aggregation simulator. The next study utilised the simulator for a more detailed investigation on the dependence of SSP upon aggregate characteristics. For instance, the size and aspect ratio of the constituent crystals were found to have a significant impact on the extinction and back-scattering cross-sections of the aggregates. The third study analysed the ability of the SSP database to reproduce a combination of real passive and active satellite observations, by the GPM (Global Precipitation Measurements) Microwave Imager (GMI) and the CloudSat Cloud Profiling Radar, in radiative transfer (RT) simulations. While the tested particle models could accurately reproduce the real observations, it was difficult to find a particle model that performed better than the others. However, complementary simulations show promise with respect to the upcoming Ice Cloud Imager. In the fourth study, SSP of ice particles that have a preference towards horizontal orientation were calculated and applied to passive RT simulations at 166 GHz. The characteristic polarization signals present in GMI observations of clouds were successfully reproduced by RT simulations. The final study provides SSP of non-spheroidal rain drops, accounting for the effect of aerodynamic pressure upon the drop shape. It was found that this effect can have a small, but non-negligible, impact on passive and active microwave observations.