The Doppler velocity spectrum width (<em>&sigma;</em>_v) observed by dual-polarization weather radar is widely used as a proxy for atmospheric turbulence, including convectively induced turbulence (CIT) hazardous to aviation. However, <em>&sigma;</em>_v also responds to the aerodynamic behavior of frozen hydrometeors, so its interpretation within the dendritic growth zone (DGZ) remains ambiguous. This study combines a spheroidal aerodynamic response model with quasi-vertical profile (QVP) observations from eight precipitation cases recorded by the S-band dual-polarization radar at the Yongin testbed. Simulations show that particle geometry governs how efficiently a wind perturbation is converted into radial-velocity variance: the whole plane-crystal series, from plates to dendrites, responds far more strongly than columnar or rimed habits. A quantitative procedure identified a <em>&sigma;</em>_v peak zone in five cases, all stratiform. In four of them <em>&sigma;</em>_v showed no systematic relationship with vertical wind shear (<em>VWS</em>) but a temperature-dependent negative relationship with differential reflectivity (<em>Z</em>_DR); static stability and near-vertical radial-velocity observations argue against a thermally driven origin, while the ice-relative vapor density excess (<em>&Delta;&rho;</em>_v) maximum was positively related to, and vertically adjacent to, the <em>&sigma;</em>_v maximum. Beneath the peak, an aggregation index derived from <em>Z</em>_H, <em>Z</em>_DR, and <em>K</em>_DP increased with <em>&sigma;</em>_v magnitude, whereas a counterpart index for vapor-depositional growth did not. The cases separate into weak plate-dominated growth, strong dendritic growth, and one shear-driven peak unrelated to the DGZ. These results suggest that <em>&sigma;</em>_v within growth layers is partly determined by the hydrometeors observed, so estimating turbulence inside precipitating systems requires knowledge of the microphysical state and hydrometeor type.