Uncertainty is a defining feature of environmental and climate systems, arising from natural variability, incomplete observations, and imperfect model representations. Accurately characterizing and propagating this uncertainty remains a central challenge for reliable prediction and decision-making. Bayesian approaches offer a principled probabilistic framework that enables the integration of prior knowledge with observational data, providing a coherent mechanism for uncertainty quantification across complex and multiscale systems. This review synthesizes recent advances in Bayesian methods for environmental and climate modeling, with a focus on their role in handling uncertainty. Key methodological developments, including parameter estimation, hierarchical and spatial modeling, data assimilation, and modern computational techniques, are examined alongside their applications in climate model calibration, hydrology, air quality assessment, and extreme event prediction. Particular attention is given to how Bayesian frameworks facilitate the propagation of uncertainty from data to model outputs, thereby supporting more robust and transparent inference. The review further discusses current challenges, including computational scalability, prior specification, data limitations, and the communication of probabilistic results to policymakers. By integrating methodological perspectives with applied insights, this work provides a coherent synthesis that bridges fragmented literature and highlights emerging directions such as hybrid physics-informed models and Bayesian machine learning. Overall, the paper underscores the critical role of Bayesian approaches in advancing uncertainty-aware environmental modeling and supporting evidence-based decision-making in the context of global change.