An Improved Morphology‐Aware Model for Predicting the Settling Velocity of Oncomelania hupensis

Accurate prediction of drag coefficient ( C D ) and terminal settling velocity for irregular, bio‐origin particles is essential for dispersal assessment and eco‐hydraulic simulation; however, substantial uncertainty persists for living Oncomelania hupensis because its conical–spiral shell, surface roughness and heterogeneous mass distribution are not adequately represented by classical spherical or generic non‐spherical correlations. We quantified the settling of living O. hupensis collected from representative middle‐Yangtze habitats, obtaining 70 valid individuals and stratifying them into three size classes by major‐axis length. Tests were conducted in a transparent cylindrical column with a mid‐column observation window. Terminal settling segments were objectively delineated from the near‐zero‐acceleration intervals of vertical displacement–time records, enabling the back‐calculation of C D and particle Reynolds number ( R e ) via force–balance inversion. Measured terminal velocities spanned 7.08–15.324 cm/s, with corresponding C D ranging 0.52–2.97 over R e = 0.5–15. Building on the two‐term Haider–Levenspiel structure, a morphology‐aware drag correlation was developed in which model coefficients were parameterized as functions of shape descriptors (e.g., sphericity and Corey shape factor), supplemented by a smoothness‐stabilized correction and a nonlinear morphology–drag coupling term, and it was calibrated using constrained nonlinear least squares with multi‐start initialization. Relative to baseline spherical and commonly used non‐spherical formulas applied to the same dataset, the proposed correlation increased the coefficient of determination from 0.45 to 0.76 and reduced mean square error from 0.23 to 0.11. Size‐resolved results indicated the tightest C D – R e distributions for medium individuals, whereas small and large classes exhibited higher dispersion, consistent with posture/orientation variability and morphological heterogeneity. The resulting correlation and reproducible parameter‐inference workflow provide practically accurate inputs for particle‐tracking modules, improving predictions of settling distance, residence time and deposition hot spots of O. hupensis to support habitat management and schistosomiasis control strategies in river–floodplain systems.