Investigating the role of model structure and surface roughness in generating flood inundation extents using one‐ and two‐dimensional hydraulic models

TLDR

Hydraulic models are essential for predicting flood inundation areas, and selecting an appropriate model and its calibration is crucial for accurate predictions across diverse 1D and 2D configurations. The study compares the performance of four commonly used 1D and 2D hydraulic models with respect to model structure and sensitivity to surface roughness characterisation. The comparison evaluates HEC‑RAS 1D, HEC‑RAS 2D, LISFLOOD‑FP diffusive, and LISFLOOD‑FP subgrid models across four study reaches with varying river geometry and roughness characterisation. Across four reaches, model geometry had little effect, 1D models performed comparably to 2D models (with 2D slightly better), 2D accuracy improved with higher channel roughness while 1D accuracy improved with higher floodplain roughness, and uniform floodplain roughness characterisation yielded the best results compared to distributed roughness.

Abstract

Hydraulic models play an important role in determining flood inundation areas. When considering a wide array of one‐ (1D) and two‐dimensional (2D) hydraulic models, selecting an appropriate model and its calibration are crucial in an accurate prediction of flood inundation. This study compares the performance of four commonly used 1D and 2D hydraulic models, including HEC‐RAS 1D, HEC‐RAS 2D, LISFLOOD‐FP diffusive, and LISFLOOD‐FP subgrid, with respect to their model structure and their sensitivity to surface roughness characterisation. Application of these models to four study reaches with different river geometry and roughness characterisation shows that for a given set of roughness condition, the geometry, including the sinuosity, reach length and floodplain width, does not affect the performance of a 1D or 2D model. Overall, the performance of a 1D model is comparable to the 2D models used in the study, with the 2D models showing slightly better results. The performance of 2D models is affected by low channel roughness, and it improves with increasing channel roughness that enables more water to enter into the floodplain. On the contrary, the performance of 1D model is positively affected with increasing floodplain roughness. When the models are evaluated for uniform versus distributed roughness characterisation in the floodplain, the uniform surface characterisation provides the best results compared to the distributed roughness characterisation.