<p>Sedimentary depositions from the Mississippi River shape the geologic condition of southeast Louisiana. The region’s soft and highly compressible soil presents a significant challenge for civil infrastructure projects, including the need for ground improvement tasks. Models and observations are used to characterize the soil and predict settlement is a common practice. Accurately characterizing soil compressibility properties is particularly challenging due to the soil’s complex and heterogeneous behavior.</p><p>Most methods predict consolidation settlement using one-dimensional models whose parameters are estimated through expensive and time-consuming lab procedures. Models describing consolidation behavior are evaluated numerically, analytically, or using approximate methods. More recent studies have shown promising results by exploring the assimilation of field observations to improve accuracies. This approach is based on the premise that both modeling and observation can provide complementary information for prediction, thereby enhancing the overall accuracy of the prediction.</p><p>In this study, we analyze the field data to evaluate the Asaoka and hyperbolic methods for settlement prediction. Despite being accurate, these methods have limited application in regions and conditions beyond those factored in observational data. To broaden the application of empirical equations, we propose a model-informed empirical approach. Synthetic soil settlement curves are generated based on numerical simulation of soil settlement models by sampling soil geotechnical parameters. Plausible synthetic settlement data is then fitted to empirical models to establish the functional relationship between the empirical coefficient and the geotechnical parameters. Complementing model-generated synthetic data and limited observation provides a viable approach toward timely and reliable consolidation settlement estimation.</p>