Separating the horizontal advection component from field-measured suspended sediment concentration profiles

Area of Science:

• Coastal geomorphology and sediment dynamics
• Environmental fluid mechanics
• Data-driven modeling and analysis

Background:

• The classical one-dimensional vertical diffusion-settling (1DV) model is limited for fine-grained sediments due to its inability to account for horizontal advection.
• Existing methods for separating advection effects are often impractical, relying on expensive experimental techniques or complex physical models.
• There is a need for efficient and practical methods to analyze sediment dynamics in environments with significant horizontal transport.

Purpose of the Study:

• To introduce a novel data-driven approach using dynamic mode decomposition (DMD) to separate horizontal advection components in sediment transport analysis.
• To overcome the limitations of the 1DV model by enabling the analysis of fine-grained sediment dynamics on silty coasts.
• To provide a computationally efficient and practical method for understanding complex sediment transport processes.

Main Methods:

• Application of a hierarchical dynamic mode decomposition (DMD) method to measured suspended sediment concentration (SSC) profiles.
• Reconstruction of three distinct components: storm-induced vertical mixing/background concentration (Profile I), tidal resuspension (Profile II), and M2 tidal advection (Profile III).
• Analysis of the relative variance contribution rate (RVCR) of each component throughout the study period, including storm and post-storm phases.

Main Results:

• Profile I dominated during storm events (peak 98.7% RVCR), while Profile II became dominant after the storm dissipated (max 70.9% RVCR), controlling fine-grained SSC.
• Profile III, representing horizontal advection, showed a low contribution during storms (<2% RVCR) but increased post-storm (max 7.8% RVCR).
• The reconstructed total SSC profiles achieved high accuracy, with 83.7% of RMSE values below 0.3 g/L, validating the DMD method's effectiveness.

Conclusions:

• The data-driven DMD method successfully separates horizontal advection components from observed SSC profiles, offering a practical alternative to traditional methods.
• This approach provides crucial technical support for extending the applicability of the 1DV model to silty coasts and complex sediment environments.
• The study presents a new, effective methodology for analyzing sediment transport dynamics, particularly in distinguishing the contributions of vertical mixing, resuspension, and advection.