Comparative Analysis of Water-Level Forecasting Using SWMM and HEC-RAS Flood Simulation Models: A Case Study of Waryong, Jinju

Frequent flooding occurs in many rivers and urban areas, exacerbated by recent increase in localized heavy rainfall events induced by global climate change. Particularly, the area encompassing Songbaek to Waryong districts in Jinju is frequently affected by flooding due to typhoons and heavy rainfall-related outflow from the Namgang Dam, resulting in significant economic losses, though it is located in a flood plain. In this study, two flood analysis models, namely HEC-RAS and EPA-SWMM, were used to simulate the flood overflow in Waryong district. HEC-RAS is specialized in 2D analysis, and simulates topography-based inundation effects and spatial spread of water, thus making it advantageous for flood prediction in rivers or low-lying areas. In contrast, EPA-SWMM is specialized in 1D analysis, and focuses on sewer systems and conduit flows, thus rendering it suitable for urban drainage system analysis. These two models were used for simulations based on (i) a measured discharge of 792 ㎥/s from a recent flood event that caused overflow of the study area and (ii) a projected flood discharge of 3,540 ㎥/s for Waryong district. The simulations demonstrated that models a-1, a-2, b-1, and b-2 estimated flood depths of 15.25, 14.81, 19.20, and 17.11 m, respectively. A comparison of these results with the actual flood depth in Waryong District showed that HEC-RAS, which considers 2D flow and reflects detailed topographic data, yielded results closer to the actual flood depth. In contrast, EPA-SWMM, which focuses on 1D conduit flows, tended to underestimate the flood depth. In essence, this study suggested that considering the different analysis characteristics and limitations of the two models, their comparative analysis can provide a more accurate approach for flood prediction and overflow analysis. Future research should focus on combining 1D and 2D analyses or calibrating these models to enhance accuracy.