Bank erosion in rivers is hard to predict because the processes behind it are complex. This makes it difficult to use computer models for real-world engineering projects. To address this, researchers developed a new one-dimensional model called LRBEM, which combines calculations for water flow, sediment movement, bank erosion, and groundwater levels.The model was tested on the Middle Yangtze River and compared with HEC-RAS, a commonly used software. The results showed:LRBEM performs just as well as HEC-RAS in predicting water flow, water levels, and sediment concentration.LRBEM is better at identifying bank erosion locations, with 61% accuracy compared to HEC-RAS’s 39%.The groundwater level part of the model is very accurate, with only a 0.86% average error compared to real measurements. The model is most sensitive to changes in riverbed roughness—especially when roughness decreases.This new model can help improve flood control and river management by providing more reliable predictions of bank erosion.