Introduction

Floods rank among the most destructive natural disasters worldwide. Natural factors—such as hydrological, environmental, and topographic elements including intense and prolonged rainfall, landslides, climate change, and land use patterns, are among the most significant contributors to flood risk. However, the effects of human factors, including structural failures of dams, encroachment into riverine zones, and urbanization, also play a crucial role in exacerbating flood severity. Arid and semi-arid regions such as Iran often face flash floods triggered by heavy rainfall during intense downpours. This study is therefore necessary to delineate flood-prone zones along the Kengir River and to propose flood management strategies aimed at mitigating potential damages.



Methodology

The Kangir watershed, with an area of 122.71 km², is located southeast of Eyvan-e Gharb in Ilam Province. The Kangir River originates from the Sarab-e Eyvan spring, situated 9 km southeast of Eyvan city. The long-term average annual flow at the Siahgel hydrometric station is approximately 2.5 m³/s. To perform two-dimensional flood modeling of the Kangir River, primary data were first compiled. This dataset included a Digital Elevation Model (DEM), river channel geometry, land use/land cover data, Manning’s roughness coefficients, and peak instantaneous discharge values corresponding to various return periods.

To determine the peak discharge for different return periods, the maximum annual discharge series from the Siahgel hydrometric station (with a 25-year statistical record) was fitted to various statistical distributions using the Hyfran Plus software. Among the tested distributions, the Gamma distribution provided the best fit to the observational data. Following the completion of data preparation, the model was executed for each return period to analyze the river's hydraulic behavior under flood events of varying magnitudes. The model outputs—including flow depth maps, velocity maps, flood inundation extents, and spatial energy distribution patterns—were extracted and analyzed within a GIS environment.



Results and discussion

In this study, hydraulic modeling of the Kangir River was conducted over an 11-kilometer reach for short-term (2, 5, and 10-year), medium-term (25 and 50-year), and long-term (100-year) return periods to analyze flow behavior under flood conditions of varying intensities.

2, 5, and 10-year Return Periods: The results of the two-dimensional modeling indicate that the average flow depth in the Kangir River for discharges corresponding to the 2, 5, and 10-year return periods is 0.92 m, 1.1 m, and 1.25 m, respectively. The corresponding mean flow velocities are 0.33 m/s, 0.55 m/s, and 0.7 m/s, respectively. These values suggest the prevalence of unsteady flow conditions across a significant portion of the study reach. In this return period, the flow depth is not limited solely to the main channel but also extends into portions of the floodplain; however, a significant portion of the flow remains confined to the main river channel.

Conversely, in shallower sections, sediment transport capacity is reduced, making sediment accumulation the dominant process. Within the studied reach, the depth was greater in sections where the river width was narrower (specifically at the locations of Shoorabeh Mel and Shaleh Shouri villages).

Return periods of 25, 50, and 100 years: For flood events with return periods of 25, 50, and 100 years, the average flow depths were 1.45, 1.52, and 1.97 meters, respectively. This indicates that the river occupies a significant volume of water within its channel and floodplain.

The mean depth in the 100-year return period reflects the inundation of wider, low-gradient areas of the floodplain, illustrating the expansion of the inundated area under extreme flood conditions. The depth values suggest that during a 100-year flood event, a large portion of the floodplain will experience extensive inundation. The presence of localized points along the river with considerable depth (7 meters) indicates the existence of scour holes, areas of deep local erosion in the riverbed, or sites of hydraulic constriction. These locations are regarded as “critical points” from an engineering and management perspective.

In the 50-year event, broader sections of the floodplain become actively involved in the flood cycle. Images from the 50-year return period clearly show that the floodplain is activated along the entire river reach. Flatter river sections exhibit the greatest lateral expansion, and inundation around the villages of Shaleh-Shouri, Mir-Alamdār, and Shoorabeh-Mal becomes noticeably deeper and more pronounced. During a 100-year return period flood, the inundation reaches its maximum lateral extent, with depths of 1 to 3 meters across vast areas of the floodplain, indicative of severe flooding. The water flow is completely diverted from the first approximately 3 kilometers of the stretch, severely affecting agricultural lands, orchards, and scattered settlements.



Conclusion

According to the modeling data, the floodplain area for a discharge with a 100-year return period (1.64 km²) was twice that of the inundation area for a 2-year return period (0.79 km²). The increasing trend of the floodplain area and flow depth for discharges with return periods from 25 to 50 and 100 years was not uniform across different reaches and depended on the topography of the riverbanks and margins. The floodplain area values for the 50- and 100-year return periods were not significant in reaches where the riverbank had a high slope and was elevated, and most changes occurred in the river's depth and velocity. Furthermore, due to the increase in ground slope and the restriction of the floodplain by high terraces, the increase in flood inundation level for the 100-year discharge was not very high. The results showed that with increasing discharge for different return periods, the depth, flow velocity, and extent of the floodplain increased, but these values were not uniform across different reaches.

The results emphasize that flood management in this area cannot rely solely on structural methods; instead, adopting an integrated approach based on partial floodplain restoration and spatial land-use planning is inevitable.

In general, the results demonstrate that two-dimensional flood modeling is a powerful tool for accurately assessing flood behavior, identifying high-risk areas, and supporting management decision-making. The output of this research can serve as a basis for developing risk reduction strategies, improving watershed management, designing protective structures, and determining river boundaries under events with different occurrence probabilities.