<p>Ice jams, which are prevalent in rivers of cold regions, can escalate into severe flooding disasters. Understanding variations in ice jam thickness is crucial, generating significant scholarly interest in developing accurate computational methods. Current models primarily rely on mechanical equilibrium equations to estimate ice jam thickness, representing a significant advancement in theoretical research. However, these models often overlook critical factors such as the cohesion of ice jams and the distribution of equilibrium stress across the river’s width, which can undermine their accuracy. This study introduces an enhanced model that incorporates these aspects, thereby improving the mathematical rigor. Validated against empirical data from natural rivers, the proposed model demonstrates strong agreement with observed values. This research not only refines the theoretical framework for calculating ice jam thickness but also improves the prediction and management of ice jam evolution and related disasters in cold regions.</p>