Polarimetric observables#
Dual-polarization weather radars transmit and receive orthogonal
linear polarizations (conventionally horizontal h and vertical v).
A handful of combinations of the resulting amplitudes and phases
form the standard polarimetric observables implemented in
rustmatrix.radar. Definitions, sign conventions, and units below
follow [Bringi and Chandrasekar, 2001] and [Doviak and Zrnić, 1993].
The scattering matrix#
For a single particle at a fixed geometry, the far-field scattered electric field is related to the incident field by the \(2 \times 2\) amplitude scattering matrix \(\mathbf{S}\):
The \(S_{ij}\) are complex — they carry both amplitude and phase.
rustmatrix returns them from the T-matrix solver at any specified
incident / scattered geometry via
rustmatrix.scatter.amplitude_matrix.
Polarimetric observables come in two flavours:
Back-scatter quantities (\(Z_h\), \(Z_{dr}\), \(\rho_{hv}\), \(\delta_{hv}\), LDR) use \(\mathbf{S}\) at the 180° back-scatter geometry and integrate \(|S|^2\)-type quantities over the PSD.
Forward-scatter quantities (\(K_{dp}\), \(A_h\), \(A_v\)) use \(\mathbf{S}\) at the 0° forward-scatter geometry and integrate \(\Re\) / \(\Im\) linear in \(S\).
Switch geometries with s.set_geometry(geom_horiz_back) /
s.set_geometry(geom_horiz_forw). The T-matrix itself is cached on
the Scatterer, so the switch is cheap.
Back-scatter observables#
Reflectivity factor \(Z_h\)#
Units: mm⁶ m⁻³ (linear); dBZ after \(10 \log_{10}\). The dielectric
factor \(|K_w|^2\) for water at the radar band is tabulated in
rustmatrix.tmatrix_aux.K_w_sqr. Use radar.refl(s, h_pol=True).
Differential reflectivity \(Z_{dr}\)#
Positive \(Z_{dr}\) means oblate scatterers aligned with their long axis horizontal (the equilibrium-drop configuration). Rain at C-band gives \(Z_{dr} \approx 0.3\)–3 dB; pristine ice columns give slightly positive values; randomly-tumbling aggregates give ~0 dB.
Co-polar correlation \(\rho_{hv}\)#
Bounded by \([0, 1]\). Values near 1 mean a uniform population; drops below 0.97 indicate mixed-phase or irregular scatterers. A key discriminator between meteorological and non-meteorological echo.
Differential backscatter phase \(\delta_{hv}\)#
Non-zero only when scatterers are large enough that the Rayleigh approximation fails — so a resonance fingerprint at C-band for \(D \gtrsim 5\) mm. The HydroMix tutorial shows it for mixed rain + ice.
Linear depolarization ratio LDR#
Requires non-zero cross-polar response, i.e. non-trivial canting. LDR is small (< −25 dB) for rain and rises sharply in the melting layer and for oriented ice [Kumjian, 2013, Ryzhkov et al., 2005].
Forward-scatter observables#
Specific differential phase \(K_{dp}\)#
Units: ° km⁻¹. Positive for horizontally-oriented oblate particles (rain), near-zero for spheres and tumbling ice, slightly negative for vertically-oriented crystals. \(K_{dp}\) is immune to attenuation and calibration bias — the workhorse for rain-rate retrieval.
Specific attenuation \(A_h\), \(A_v\)#
Units: dB km⁻¹. Rises sharply at the higher radar bands (Ka, W) — see the radar-band sweep tutorial.
Sign and geometry conventions#
rustmatrix follows Bringi & Chandrasekar [Bringi and Chandrasekar, 2001]:
Horizontal polarization is
h, vertical isv.Equilibrium drop axis ratio is reported as \(h/v\) ≥ 1; the
Scatterer(axis_ratio=...)argument expects \(v/h\) (the value the Mishchenko code wants), so rain scripts useaxis_ratio = 1.0 / dsr_thurai_2007(D).Back-scatter geometry is
geom_horiz_backfromrustmatrix.tmatrix_aux(the BSA convention — incident antenna frame). Forward-scatter isgeom_horiz_forw.Doppler velocity is positive downward (toward the radar for a vertically-pointing profiler) throughout
rustmatrix.spectra— matching the convention in [Kollias et al., 2002, Zhu et al., 2023].
Relationship to the scattering cross-sections#
Back-scatter cross-section \(\sigma_b\) and total scattering
\(\sigma_{sca}\) are also available via
rustmatrix.scatter.radar_xsect and scatter.sca_xsect — useful for
Mie parity checks and for feeding spectra.SpectralIntegrator a
custom \(\sigma_b(D)\).
Further reading#
[Bringi and Chandrasekar, 2001], chapters 3–4 — the definitive reference.
[Kumjian, 2013] — a three-part practical tutorial aimed at operational meteorologists.
[Doviak and Zrnić, 1993] — for the radar-equation derivations and the sign conventions used in the Doppler / spectral modules.
V. N. Bringi and V. Chandrasekar. Polarimetric Doppler Weather Radar: Principles and Applications. Cambridge University Press, 2001.
R. J. Doviak and D. S. Zrnić. Doppler Radar and Weather Observations. Academic Press, 2nd edition, 1993.
P. Kollias, B. A. Albrecht, and F. D. Marks Jr. Why Mie? Accurate observations of vertical air velocities and raindrop size distributions using a cloud radar. Bulletin of the American Meteorological Society, 83(10):1471–1483, 2002. doi:10.1175/BAMS-83-10-1471.
M. R. Kumjian. Principles and applications of dual-polarization weather radar. Parts I–III. Journal of Operational Meteorology, 2013. doi:10.15191/nwajom.2013.0119.
A. V. Ryzhkov, D. S. Zrnić, and D. W. Burgess. Polarimetric tornado detection. Journal of Applied Meteorology, 44(5):557–570, 2005. doi:10.1175/JAM2235.1.
Z. Zhu, P. Kollias, and F. Yang. Particle inertia effects on radar Doppler spectra simulation. Atmospheric Measurement Techniques, 16:3727–3737, 2023. doi:10.5194/amt-16-3727-2023.