As an important element of water resources management, suspended sediment (SS) management plays a crucial role in water quality control of river systems. Effective management and planning require accurate predictions, reliable diagnostic tools and clear interpretations of model-based inference to provide guidance for designing proactive management actions. Understanding each stage of the SS production, storage, and transport processes of the fluvial system is crucial to understand sediment induced impairments and target management practices. While there are different approaches to characterize various components of complex fluvial processes, which are significantly influenced by streamflow (Q), this dissertation aims to address one major challenge to develop robust method for predicting SS yield and diagnosing changes in sediment transport processes. Part of the challenge relates to the difficulty in characterizing the changing SS-Q relationship across multiple timescales, as well as other nonlinearities in the underlying relationship, which highlights the importance of model formulation and parameter interpretation in capturing the multi-scale dynamics in the SS-Q relationship. This dissertation forwards a novel approach to incorporate fluvial SS dynamics into model development for management and planning through a series of advances and applications of Dynamic Linear Models (DLMs). First, an inter-model comparison suggests that a variant of DLMs with additional modeling on the residuals produces the most robust SS forecasts at 1-7 day lead times among seven rating curve models, and model selection can significantly influence the inference of optimal system operations during extreme events (Chapter 2). Then, the ability of DLMs to capture high-resolution changes in fluvial SS processes can also help evaluate SS reduction management practices by isolating the signals attributed to infrastructure and natural fluvial processes (Chapter 3). Lastly, DLMs can quantify the long-term dynamicity and capture seasonal patterns of SS transport processes, which can help develop management strategies that incorporate multi-scale dynamics in the SS-Q relationship based on local basin physiographic features (Chapter 4). The overarching goal of this work is to introduce a reliable and efficient modeling framework to quantify multi-scale SS-Q dynamics and provide strategies to address multi-scale dynamics in SS management and planning studies.