Flood modeling using HEC-RAS 2D and IBER 2D: a comparative study

Today, our world is experiencing more natural disasters than ever (Vignesh et al. 2021; Hallil & Redjem 2022). This can lead to a wide range of damages that endanger human life and the environment (Sami et al. 2021; Afzal et al. 2022). Floods have been identified as one of the most common, dangerous, and destructive natural disasters (Tamiru & Wagari 2022; Debnath et al. 2023; KatipoğLu & Sarıgöl 2023), negatively affecting humans and ecosystems that occur around the world (Ali et al. 2019). Floods have adverse impacts in rural and urban areas, due to their frequency, intensity, and resulting damage (Bourenane et al. 2019; Awad et al. 2022; Benzougagh et al. 2022; Eslaminezhad et al. 2022; Leghouchi 2023). According to statistics published by the Center of Research on Epidemiology of Disasters (CRED), a total of 387 natural disasters occurred in 2022, where floods alone accounted for 176 events (CRED 2022; Ochani et al. 2022). Floods are natural disasters where the overflow of water from its normal confines, such as rivers, lakes, and oceans (Namara et al. 2022), resulting in the inundation of surrounding areas (Chouki & Makhloufi 2022; Korkmaz 2022). The reasons why the frequency of floods is increasing around the world are mainly related to climatic changes (precipitation and evapotranspiration) (Imani et al. 2023), vegetation issues, human interventions including dam failure (Güvel et al. 2022) as well as the change in land use as a result of urbanization (To et al. 2022), which leads to increasing surface runoff and associated flooding (Hussain et al. 2022; Mumtaz & Abustan 2022).

In recent decades, the encroachment of cities and towns across floodplains has significantly increased the potential for flood damage (Madi et al. 2023). Algeria has been exposed to devastating floods that caused considerable human and economic losses (Hafnaoui et al. 2022), making it the country's most dangerous natural hazard (Mokhtari et al. 2023). In general, floods in urban areas pose a significant danger due to several factors (Esmaiel et al. 2022). Located in the west of the country, El Bayadh City is one of the Algerian cities that has experienced multiple floods, the largest of which occurred on 1 October 2011. A storm with heavy rainfall (approximately 60 mm) caused floodwaters to spill from the main river traversing the city. This led to 12 deaths, affecting 400 families, and multiple injuries in addition to damage to hundreds of homes and two bridges that were destroyed (Hachemi et al. 2019; Hafnaoui et al. 2020).

To raise flood awareness in order to prepare for and protect vulnerable communities against floods, it is essential to establish and produce flood risk maps using the best available tools and software (Shustikova et al. 2019). Hydraulic modeling is widely used by practitioners to prepare flood risk maps (Diaconu et al. 2021), and it also plays a significant role in flood management (Shah et al. 2022). There are a range of numerical models available of varying complexity. These include one-dimensional (1D), quasi-two-dimensional, two-dimensional (2D), and three-dimensional (3D) models which each have their own benefits and limitations (Gharbi et al. 2016). Hydraulic modeling tools such as HEC-RAS, IBER, MIKE, SRH, FLO-2D, LISODFP, TUFLOW, MODCEL, ISIS, and TELEMAC have been used in different flood studies (Iroume et al. 2022). Due to differences in their capabilities accuracy and run-time, choosing a hydraulic model best suited to generating the required flood risk maps is significantly important (Diedhiou et al. 2020). 2D hydraulic modeling has been increasingly used in the last few years (Edirisooriya et al. 2022) due to the growing availability of LiDAR data to support the 2D representation of floodplains. In addition, 2D modeling is capable of representing complex overland flow paths in urban landscapes and is able to simulate unsteady flow dynamics across multiple flow paths, including backflow in floodplains (Merwade et al. 2008).

Several scientists have expressed their opinions regarding 2D hydraulic modeling (Yang et al. 2006) including the view that the application of 2D numerical models is more efficient for flood modeling. Pérez-Montiel et al. (2022) have mentioned that 2D models have made significant progress in flood modeling, and they are considered a more complete alternative to 1D models.

The functionality and capabilities of free hydraulic modeling software have increased significantly and can be comparable to proprietary software (Martins et al. 2020). Two examples include HEC-RAS and IBER. HEC-RAS is a software developed by the Hydrologic Engineering Center of the U.S. Army Corps and is a well-known tool for hydrodynamic modeling in the water resource research community (Loi et al. 2019).

The latest release of HEC-RAS (version 6.4.1) can be downloaded from the website (https://www.hec.usace.army.mil/software/hec-ras/download.aspx) and runs on 64-bit Windows operating systems.

Quirogaa et al. (2016) applied HEC-RAS 2D to simulate the floods that occurred in the Llanos de Moxos region in Bolivia in February 2014 while Bhandari et al. (2017) used HEC-RAS in an unsteady 2D flow simulation in the Nor Brazos catchments. The HEC-RAS software has also been applied to assess the risks of flooding in urban areas (Rangari et al. 2019; Ongdas et al. 2020; Dahal et al. 2021; Zeiger & Hubbart 2021; Cerneaga et al. 2022). Islam & Rahman (2020) used HEC-RAS 2D to prepare flood maps for a number of rivers in Bangladesh, including the Ganges River, Padma River, Jamuna River, and upper and lower reaches of the Meghna River. Studies such as these have shown the usefulness and effectiveness of this program when preparing flood risk maps and identifying sites potentially subject to flooding, the depth of floodwaters, and the expected duration of inundation, which is of great benefit when formulating strategies and plans to reduce the loss of life and property damage due to flooding.

IBER is software developed by the Water and Environmental Engineering Group, GEAMA (University of A Coruña), in collaboration with the Flumen Institute (Polytechnic University of Catalonia, UPC), the International Center for Numerical Methods in Engineering (CIMNE), the EPhysLab laboratory from the University of Vigo, and the Centre for Hydrographic Studies of CEDEX (Pérez-Montiel et al. 2022). It has been widely used for flood risk analysis and flood level assessment (Ilioski et al. 2022). The software package IBER provides a user-friendly graphical user interface for pre- and post-processing and can be freely downloaded from (https://iberaula.es/space/54/downloads).

The IBER program has also been used in several studies, as it was developed for consulting and applied research investigations. Due to its adaptability, usability, and availability to all users, it has emerged as a viable tool in the academic community for teaching topics related to free surface hydraulics and flood modeling (Cueva-Portal et al. 2021). For example, García et al. (2022) used IBER 2D to create maps of flood-prone areas in the Tesechoacán Basin. Naiji et al. (2021) used IBER 2D to produce flood risk maps for the urban area of Zaio in Morocco. Martins et al. (2018a) assessed the risk of the flood phenomenon in the urban center of Amarante while Cruz-Castro et al. (2022) used IBER as well as a collaborative map to prepare flood maps, which identified the most vulnerable areas of Peñón de los Baños in the borough of Venustiano Carranza in Mexico.

The main objective of this study was to assess the areas vulnerable to flooding in El Bayadh City using HEC-RAS 2D 6.4.1 and IBER 2D 2.5.2 software in order to assess the comparative performance of these two software. This comparison was based on the predicted flood depths and flow velocities, as well as the predicted extent of the areas at risk of flooding.

In this study, the models were based on a 7-m resolution digital model, created through a combination of a 30-m resolution digital surface model (DSM) and Google Earth data. The floods that were simulated were floods with return periods of 50, 100, and 1,000 years. The resulting flood risk maps can be used to inform decision-makers and concerned governments about the flood risks and to support the preparation of appropriate plans to mitigate the risk of flooding in existing developed areas as well as make appropriate plans for future development.

The topography of the study area was prepared through a combination of a 30-m resolution DSM and Google Earth data. A numerical model was created with 7-m accuracy to achieve the desired topography based on the global DSM database (ALOS World 3D – 30 m, AW3D30) with a resolution of 30 m, downloaded from the Japan Aerospace Exploration Agency (https://www.eorc.jaxa.jp/ALOS/en/aw3d30/). In addition, data were obtained from Google Earth Pro, which is represented by regularly distributed location points. ArcGIS system was used to prepare this model with the help of certain toolbox operations.

The geometry of the study area was represented using a rectangular mesh in the case of the HEC-RAS software and a triangular mesh in the case of the IBER software, the size of the mesh was 8 m. The Manning roughness coefficient adopted for this study was 0.030 because most of the study area is a natural stream and urban area (Chow 1959). For the upstream boundary condition, the flood hydrograph for each return period was used upstream for both software. The peak inflow at the upstream boundary was 435.59, 513.78, and 721.78 m³/s for the return periods of 50, 100, and 1,000 years, respectively, based on the study of Hafnaoui et al. (2020). For the downstream boundary condition, the normal depth was used for HEC-RAS software, and supercritical/critical flow was proposed for IBER software.

In this study, simulations of the El Bayadh city flood were carried out using the HEC-RAS and IBER software. To analyze the simulation results, we made a comparison between the extent of the flooded areas that were obtained using the two software and limits traced by the technical service of El Bayadh municipality. The estimated flood of October 2011 was equal to 426.45 m³ /s, corresponding to the 50-year return period (Hafnaoui et al. 2020).

In this study, several simulations were performed for different return periods, the simulated return periods are 50, 100, and 1,000 years. To analyze the simulation results, we made a comparison between the depths, the water velocities, and the extension of flooded areas for both software. The results of the simulations will be presented and discussed in this section.

The analysis of flow depths and velocities was carried out by determining the values at specific points along the river. The results of the numerical simulation of El Bayadh city for the return periods 50, 100, and 1,000 years are represented in the figures below.

The results showed that there is an increase in the water depth due to the higher inflows associated with the increase in the return periods for both software. This leads to the progressive expansion of the flooded area for each return period. There was an acceptable agreement between the two software, where the maximum depth values estimated by the HEC-RAS software were 6.28, 6.46, and 6.91 m, while the IBER software, estimated maximum depths of 6.11, 6.30, and 6.76 m for the return periods of 50, 100, and 1,000 years, respectively.

The HEC-RAS maximum water depths were 0.17, 0.16, and 0.15 m higher than the IBER maximum water depths for the return periods 50, 100, and 1,000 years. It was observed that the water depth difference is consistent and was attributed to the type of mesh used, computing time steps, as well as the nature of numerical schemes used for both software. It was also noted that there are many areas in which the depth values are higher than 4 m, this increase can be explained by three reasons. This increase can be explained by the topography of the wadi, the presence of depressions in the wadi bed, the decrease in slope, which leads to a reduction in the velocity, which is represented in Zone 1 (Figure 4(c)), and the change in the direction of the wadi channel, which makes it work as an obstacle to water, which is represented in Zone 2 (Figure 4(c)).

Figure 4(c) represents the water surface elevations simulated by two software for each return period of 50, 100, and 1,000 years, respectively. the results show a convergence between the water surface elevation profile obtained through the two software.

The difference in the maximum estimated velocities is 0.38, 0.5, and 1.98 m/s for the return periods 50, 100, and 1,000 years. The difference in the maximum velocity for the 1,000-year return period is rather significant. This difference can be explained by the topography of some zones in the wadi. The form and dimensions of the wadi play an important role in reducing or increasing flood risk (Hafnaoui & Debabeche 2021). The reduction in the cross-section of the wadi leads to a rapid increase in the flow velocity; this increase leads to a change in the flow regime, which requires specific equations to solve these phenomena. To reduce the velocity in these areas, it is suggested to widen and rehabilitate (section calibration) the wadi cross-sections to reduce erosion in the narrow areas of the wadi (Figure 6(c)).

The comparison showed that the limits of the flooded areas are almost identical for the three return periods. The flooded areas estimated by the HEC-RAS software were 0.4794, 0.4948, and 0.5304 km² while the flooded areas estimated by the IBER software were 0.4872, 0.5047, and 0.5452 km² for the 50, 100, and 1,000-year return periods, respectively. The differences between the estimated flooded areas for the three retour periods are 0.0078, 0.0099, and 0.0148 km², representing a relative deviation between 1 and 3%, which indicates a great agreement in the results of the two software.

The estimated also the percentage of developed areas in El Bayadh City using HEC-RAS are 26.30, 27.12, and 27.92%, and the percentage using IBER are 25.87, 27.38, and 29.35% for the return period of 50, 100, and 1,000 years, respectively.

In this study, the performance of two public domain 2D flood models, namely HEC-RAS 6.4.1 and IBER 2.5.2 to estimate flooding in the Wadi Deffa of El Bayadh City was compared. The flood depths, flood velocities, and flood extents were compared for floods with 50, 100, and 1,000-year return periods. In terms of the maximum flood depth, HEC-RAS software gave values 0.15–0.17 m higher than IBER software for the three return periods, while the maximum flood velocities estimated by IBER were higher than those estimated by HEC-RAS. The flood extents estimated by HEC-RAS and IBER were in good agreement across the three return periods with a relative difference of between 1 and 3%. Even though both hydraulic models provide different results in some zones, it was concluded that the results were quite close. The differences between the two softwares may be due to different numerical schemes.

A flood risk map was also prepared to identify areas most vulnerable to flooding. This map can be used to mitigate the risk of flooding and to inform the redevelopment of areas that are at risk from major floods in Wadi Deffa.

This work was supported by the Directorate General for Scientific Research and Technological Development (DGRSDT), Ministry of Higher Education and Scientific Research, Algeria.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.