In English

Soft model approximation of microwave scattering properties of ice particles

Maryam Jamali
Göteborg : Chalmers tekniska högskola, 2015. 77 s.
[Examensarbete på avancerad nivå]

Clouds consisting of ice particles have a crucial role in our climate system and profoundly influ- ence the Earth’s radiation. However, these clouds are poorly constrained in climate models mainly due to uncertainties associated with estimation of their ice mass. It has been investigated by sev- eral studies that the combined microwave millimeter/submillimeter spectral region is suitable for observing the mass of ice clouds. One of the major uncertainty sources of the microwave ice mass estimation is that the shape of ice particles is poorly known. It has been shown by several studies that ice particles can take on extremely variable irregular shapes. It is therefore necessary to find shape models that can approximate and simplify the reality. Two common simple shape models frequently used are spheres and spheroids. They can be either solid (consisting of pure ice) or soft (consisting of a homogeneous mixture of ice and air). This thesis is concerned with examining the applicability of the simple models in estima- tion of ice mass across the microwave spectral region. The focus is put on soft models. Three databases consisting of optical properties of some randomly oriented non-spherical ice particles and aggregates are considered as reference. The practical objective is to determine microphysical characteristics (i.e., shape, volume or mass fraction of air, and refractive index) of a soft model that mimics average optical properties of the reference data across all microwave frequencies from 90 to 874 GHz and size parameters (x) up to 6. It is found that the volume air fraction of a soft model should vary with both size and frequency to give the best fit of the reference data. It is therefore impossible to define a soft model with a single air fraction working in a broad range of frequencies and particle sizes. It is also demon- strated that determining the volume air fraction based on size-density parameterisations results in underestimation of mean optical properties at larger size parameters (x > 0.5). Furthermore, it is concluded that applying the Maxwell Garnett mixing rule with ice as inclusion and air as matrix media (see section 3.4) underestimates the imaginary parts of refractive index of the soft models at lower size parameters (x < 1 ). Overall, it is concluded that soft models can be tuned to give proper results over a narrow range of frequencies, but such a model lose its accuracy over other range of frequencies. One alternative approach to soft models is applying a sector-like snowflake model to represent the mean optical properties of the reference data. This was suggested by Geer and Baordo [2014] and confirmed by Eriksson et al. [2014].

Nyckelord: ice particle scattering, microwaves, remote sensing



Publikationen registrerades 2015-03-05. Den ändrades senast 2015-03-05

CPL ID: 213412

Detta är en tjänst från Chalmers bibliotek