Physical and Microwave Modeling of Electromagnetic Fields
within a Granular Snow Layer
Paul Siqueira1 and Jianchen Shi2
A fundamental component of monitoring snow characteristics from the air or from space is to under-
stand and model the interaction of electromagnetic fields within the layer itself. For this purpose, a
combination of empirical, numerical, and theoretical models are implemented to achieve a degree of
mathematical simplicity in conjunction with flexibility for addressing a variety of real-world situa-
tions.
From a remote sensing point of view, the most important parameter for characterizing a generalized
snow layer is to estimate its complex permittivity. Knowledge of this basic parameter is directly
equivalent to having knowledge of the layer's emissivity, extinction coefficient, and propagation
constant--critical components for modeling remote sensing observations using both radiometry and
radar.
Depending on the observing frequency, models of the snow layer may take on one of three forms:
i) a uniform dielectric slab (low-frequency), ii) a dielectric half space (high-frequency), or iii) a
granular medium whose components strongly interact with the ambient field (mid-frequency). Be-
cause of the complex, dense-medium scattering interactions involved in the granular medium case, it
is particularly problematic for developing reliable and consistent models. Current state of the art
techniques involve coherently modeling the interaction of electromagnetic fields with the pseudo-
crystalline structure of the snow layer via a theoretical approach called the quasi-crystalline approx-
imation (QCA).
Because of the coherent field approach employed by this technique it is necessary to statistically
describe the position of scatterers within the medium so that electromagnetic interactions between
these particles may be accounted for. To achieve this end, a number of physical models for the
arrangement of particles within the snow layer have been proposed. This paper will present one of
those models (a deposition model) as well as demonstrate its application to a full three-dimensional
determination of effective permittivity for the medium using both a numerical approach and the
theoretical approach of QCA. Our long-term goals are to utilize the packing algorithm presented in
this paper in conjunction with snow section data to firstly refine the model for crystal positions
within the snow layer and secondly to develop a robust method for estimating effective permittivity
without the requirement of in-situ measurements.
1 Radar Science and Engineering, M/S 300-218, Jet Propulsion Laboratory, 4800 Oak Grove Drive, Pasadena,
California 91109-8099, USA
2
ICESS, University of California, Santa Barbara, California 93106, USA
57
Previous Page
