 |
| Titel |
Polarimetric radar observation of ice crystals and aggregates: Backscattering modeling of signatures from C to Ka band |
| VerfasserIn |
G. Botta, M. Montopoli, F. S. Marzano |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250031251
|
|
|
|
|
|
| Zusammenfassung |
| Hydrometeor classification using polarimetric Doppler weather radar is based on the
characteristic polarimetric signature for each hydrometeor class. This signature can be
obtained by either experimental campaigns or by proper electromagnetic modelling. Both
approaches have advantages and drawbacks: the experimental approach is not easy to
conduct as it requires co-located measurements of a weather radar with in-situ sampler
(usually installed aboard an aircraft); moreover, it is generally strictly related to
the measurement configuration (e.g., frequency, range) it is performed. Of course,
experimental campaigns are needed for definitive validation, but the modelling approach
exhibits a high flexibility in terms of system and meteorological parameters very well
suited for retrieval algorithm design. On the other hand, a model approach is heavily
dependent on the model capability to represent hydrometeor volumes in a realistic
way.
Within the electromagnetic scattering modelling of hydrometeor radar response, a well
known technique to simulate the radar backscattering from an ensemble of particle is based
on the T-matrix algorithm (Kim, 2006). The T-matrix model is based on the equivalence
principle and can ensure numerical convergence for a small set of canonical shapes such as
ellipsoids. These shapes are useful to represent raindrops and vertically-oriented
small crystals, but are largely unrealistic when dealing with ice aggregates and
crystals.
In this work we use a different approach to the scattering modelling that fits well for
classes like ice crystals and aggregates of different shapes and sizes: the discrete dipole
approximation (DDA). The DDA model lets us simulate almost any kind of particle under the
hypothesis it can be approximated as a three dimensional array of dipoles that generate the
scattering field (the wavelength should be large compared to the distance between
dipoles). The DDA code used is DDSCAT, developed by Draine and Flatau (2004),
which computes the scattering by a single randomly oriented particle. With this
approach a variety of hydrometeor shapes have been simulated: cylindrical ice
crystals, aggregates of ice cylinders, snow crystals, mixed-phase particles, etc. From
DDA it has been possible to obtain the polarimetric signature for ground-based
radars at C and X band for these hydrometeor classes after solving some heavy
computational issues. An equivalent spheroid model has been also developed for the ice
hydrometeor classes in order to use a T-matrix code, faster than DDA, to simulate ice
crystals-equivalent spheroids (Weinman and Kim, 2007). Numerical results will be
discussed analyzing the sensitivity of the DDA model to the particle shape, wavelength,
size distribution and orientation. The accuracy of T-matrix approximation of the
ensemble particle polarimetric signature will be also evaluated within the context of
hydrometeor classification schemes based on either fuzzy-logic or Bayesian techniques. |
|
|
|
|
|
|