Impacts of the road on flood inundation in the Tra Khuc–Ve River basin, Vietnam

Flooding is one of the most frequent and damaging natural disasters worldwide. According to recent estimates, flooding accounts for roughly 40% of all economic damage caused by disasters (Lyu et al. 2020). Urbanization, particularly the construction of infrastructure like roadways, significantly exacerbates both the frequency and severity of floods (Grove et al. 2001; Weng 2001; Li & Wang 2009). Boulomytis et al. (2016) highlighted the importance of urban planning and the vulnerability that arises from inadequate infrastructure planning. To effectively manage flood risks and develop long-term land-use strategies, it is essential to comprehend the intricate relationships between urbanization and flooding (Nguyen et al. 2021).

While the relationship between infrastructure development and flooding has been studied, the reverse impact – the influence of roads on flood patterns – remains insufficiently explored. Most prior studies have focused on how floods affect transportation infrastructure, such as road damage, with fewer studies examining how roads can alter hydrological patterns and create new flood risks (Beevers et al. 2012). Typically, roads are constructed at elevations higher than anticipated flood levels, which can obstruct surface water flow and disrupt natural flood distribution. This situation can lead to uneven flooding on either side of the road, resulting in increased flood risks in certain areas while reducing them in others. However, empirical studies addressing this phenomenon remain limited, particularly those utilizing real-world case studies or simulations based on actual flood data from affected regions.

In previous studies, Konrad (2003) emphasized the impact of urban development on flooding, highlighting how the construction of roads and other infrastructure in flood-prone areas increases the risks of both erosion and flooding. This study provides an overview of flooding in the United States but only presents information without detailed analysis of specific case studies. In contrast, Tyler (2016) specifically examined the effects of multiple roadways on flooding and concluded that roads significantly alter flood drainage patterns, affecting the extent of flooded areas and water depth. This study compared urban planning and development in peri-urban areas of three Vietnamese cities – Hue, Da Nang, and Can Tho – examining the causes of flood events and the role of urban planning in contributing to them. However, the study did not discuss the impact of flooding on the Quang Nam – Da Nang highway, nor did it clearly specify the extent of changes in flood patterns or present flood maps. Similarly, Douven et al. (2009) explored the effects of road design on flooding in the Mekong Delta of Vietnam and Cambodia, proposing solutions to mitigate road-induced flooding. Their research is detailed, particularly focusing on policy issues and comparing various proposed solutions, including their costs. However, it does not delve into the redistribution of floodwaters or the specific changes in flooded areas caused by roads. Recently, To et al. (2022) analyzed the interaction between urbanization and reservoir operations in the Vu Gia – Thu Bon River basin, although the study has yet to propose measures to mitigate downstream flooding.

Flood mitigation can be achieved through structural and nonstructural measures. Research on flood risk management has demonstrated the necessity of applying structural measures to mitigate the impact of floods. Kundzewicz et al. (2018) highlighted the significance of structures like dykes, dams, and reservoirs in controlling water flow and reducing flood risks. Similarly, Dasandara et al. (2022) provided an overview of structural measures in Sri Lanka, including dams, reservoirs, dykes, and drainage systems. However, these solutions address general flooding and may not be effective for road-induced flooding, which requires tailored solutions like culvert expansion and water diversion. In contrast, nonstructural approaches – such as legislation, public awareness campaigns, and flood prediction systems – also play a crucial role. For example, Pesaro et al. (2018) conducted a cost-benefit analysis, clarifying the relationship between investments in nonstructural solutions and the results achieved, demonstrating that these solutions can provide high effectiveness at a reasonable cost. Bardhan (2021) pointed out that nonstructural solutions have become increasingly popular in developing countries due to the financial burden compared to constructing structural measures. Among nonstructural measures, predicting flood inundations, a vital nonstructural measure, is crucial for effective flood control (Dhote et al. 2019). The location and depth of flooded areas, as well as their extent, can be detected with the help of an inundation map. Inundation maps can assist planners in identifying high-risk areas, assessing the efficacy of various adaptation strategies, and implementing risk mitigation measures (Boulomytis et al. 2024).

Thus, the balance between structural and nonstructural approaches, along with improved road planning, remains key to effective flood risk management (Chan et al. 2019). A pressing concern for road planners in floodplains is how to design roadways sustainably to minimize negative impacts while preserving the benefits of annual flooding. Elevated roads are commonly built to prevent floods in the immediate vicinity. However, when roads are built higher than the surrounding ground, they can cause barriers to the natural flow of water. Elevated roads can cause water to gather in neighboring low-lying areas, increasing the risk of flooding rather than spreading it throughout the terrain. When roads and drainage systems do not operate together properly, surface runoff increases and water is diverted into residential areas. Consequently, elevated roads, coupled with inadequately designed drainage systems, can create conditions that amplify the flooding impacts in surrounding areas.

The process of environmental impact assessment (EIA) is essential for determining, assessing, and controlling how projects affect the environment and public health. In Vietnam, the legal framework is primarily governed by the Law on Environmental Protection; however, implementation and monitoring remain limited. The quality of EIA reports is often inadequate due to a lack of expertise or intentional oversight, complicating the approval process. Furthermore, many projects do not adequately consider climate change factors during the EIA process, potentially leading to higher risks in the future. Road projects are also particularly affected by these challenges. While most roads undergo environmental impact assessments during construction, these assessments often prove ineffective for a variety of reasons, leading to significant alterations in water flow (Douven et al. 2012; Douven & Buurman 2013). Moreover, in Vietnam, there is still a lack of clear guidelines for constructing roads in flood-prone areas, resulting in inadequate technical solutions to mitigate the impacts of roads on flood regimes.

In this context, this study aims to fill the gap in understanding the impact of road infrastructure on flood distribution, particularly in areas with hydraulic systems like reservoirs and flood control mechanisms. By analyzing the interactions between road construction and existing flooding, the study offers practical solutions to mitigate the adverse effects of flooding. Additionally, the identification of shortcomings in Vietnam's road construction guidelines for flood-prone areas can lead to policy improvements, promoting safer infrastructure planning and reducing flood risks in the future. Specifically, it addresses the following research questions: (i) How does the Da Nang – Quang Ngai expressway influence flood distribution and characteristics in the Tra Khuc – Ve River Basin? (ii) What structural measures can be proposed to mitigate the adverse effects of road construction on flood patterns? (iii) What are the policy challenges and guidelines for designing roads through flood-prone areas in Vietnam? To achieve these objectives, this research employs hydrological and hydraulic simulation models based on real flood events to validate and assess the impact of road infrastructure on flooding. Additionally, structural solutions, such as the construction of more culverts, river dredging, and the development of new drainage canals, are modeled and analyzed to propose effective mitigation measures. The paper is organized as follows: the study area is introduced first, followed by a detailed description of the methods and hydrological data utilized in the simulations. The fourth section addresses the calibration and validation processes of the models, while the fifth section outlines the simulation scenarios used for analysis. In the sixth section, the impact of the road is examined, leading to a discussion in the seventh section. Finally, the paper concludes with a summary of the findings.

The Da Nang-Quang Ngai expressway began construction in 2013 and was completed in 2018. Spanning a total of 139 km, the road traverses several basins, including an 8-km section in Da Nang, 96 km in Quang Nam, and 35 km in Quang Ngai province. Designed for a maximum speed of 120 km/h, the expressway features a roadbed width of 26 m and includes 27 large bridges, 11 medium bridges, and 64 small bridges. This expressway is crucial for promoting regional economic development by facilitating the movement of goods and people, connecting key economic centers, and enhancing trade and tourism. Additionally, it strengthens Vietnam's international transport links, particularly through the East-West Economic Corridor, which connects Laos and Cambodia. However, the road runs in a north-south direction, perpendicular to the flood flow, and is believed to significantly alter hydrological processes, adversely impacting the ecosystem and environment. Therefore, understanding and assessing these changes is vital for hazard mitigation, recovery, and sustainable development. This study specifically focuses on the segment of the road that runs through the Tra Khuc basin in Quang Ngai province, covering approximately 35 km.

Flood impacts were simulated and analyzed under various frequency scenarios of 1, 3, 5, 10, and 20%. To conduct the simulations described above, input data included rainfall, evaporation, water levels, and flow data from gauging stations both upstream and downstream. There are four main hydrological stations in the Tra Khuc River basin. Son Giang and An Chi stations measure discharge, water level, and rainfall, while Tra Khuc and Song Ve stations measure only water level and rainfall. Figure 1 shows the locations of these stations. Additionally, eight smaller hydrological stations, primarily for rainfall measurement, are distributed throughout the basin. Rainfall data have been available from these stations since 1981. Our hydrological and hydraulic models were calibrated and validated using data from the major floods of 2009 and 2013. For these events, we collected comprehensive meteorological and hydrological data, including water levels and detailed flood traces for comparison and verification. These data were sourced from government meteorological reports and measurement data from various projects. For instance, river cross-section data measured in 2020 were obtained from our project titled Assessing the Impact of the Hydropower Reservoir System in the Tra Khuc River Basin (Hùng & Tô 2021). The two reservoirs considered in this study are Dakdrinh and Nuoc Trong. The authors obtained reservoir parameters and operational documents, such as dam height, width, installed capacity, water usage, and power plant discharge, from the reservoir management authority.

Based on these results, the model demonstrates satisfactory performance. The parameters utilized in the model are detailed in Table 1, while Table 2 presents the flood transmission parameters. These parameters will be employed to simulate inflow into the Tra Khuc sub-basins effectively.

For validation performance, there is very good agreement between the observed and simulated discharges at the Tra Khuc station (see Figure 7(b)), while a slight discrepancy is noted at the Song Ve station (see Figure 8(b)). Specifically, the simulated curve shows a lower peak compared to the observed peak, and the observed data exhibit a faster decline after reaching the peak than the simulated data. For both stations, the correlation coefficient is higher than 0.96, and the Nash coefficients range from 0.88 to 0.97. The percent error in peak flow is relatively small, ranging from 5.5 to 7.9%.

The flood traces from the flood events in 2009 (13 values) and 2013 (21 values) were also used to validate the MIKE 21 model. Figure 3 illustrates the locations of these flood traces, while Figure 9 displays the simulated and observed flood depths at these locations. Most flood traces align closely with the simulation results, indicating that the models are reliable for simulating flooding events.

The impacts of the Da Nang-Quang Ngai expressway on flooding are simulated using flood scenarios with frequencies of 1, 3, 5, 10, and 20%. The topography considered includes conditions before and after the road was constructed. As previously mentioned, the two large reservoirs in the study area play a significant role in flood regulation and greatly influence flood volume; thus, their operating modes are also taken into account. During flood events, these reservoirs operate according to the inter-reservoir procedure approved by the Prime Minister (Prime Minister 2018).

Table 3 shows statistics on the flooded area according to the water depth. The flooded area is approximately 370.98 km² for a severe flood (P = 1%) and gradually decreases to 212.83 km² for a small flood (P = 20%). The shallow flooding (0–1 m) accounts for the largest portion of the total flooded area and decreases gradually as flood frequency increases. At P = 1%, the shallow flooded area is about 158.88 km² (with road) and 158.23 km² (without road). This area decreases to around 130.56 km² (with road) and 130.14 km² (without road) at P = 20%. The area with the deepest flooding (>3 m) is the smallest across all depth levels. At P = 1%, this deeply flooded area is approximately 18.30 km² (with road) and reduces to only 2.07 km² at P = 20%. The road significantly impacts flooding patterns. When built, it causes the flooded area at depths of 0–1 m to increase across all scenarios compared to before construction. In every case, the area with a flood depth of 1–2 m decreases, while the area with a flood depth over 2 m expands. This shift toward deeper flooding suggests an increased level of danger, as more extensive areas at greater depths pose heightened risks to safety and infrastructure.

The road has significantly altered flood distribution across all scenarios (see Areas 1–3). It is important to note that the primary direction of flood flows is from west to east in the absence of the road. However, with the road present, zones of substantially increased flood depth are typically found in the west (left side), while areas with reduced water depths are mainly located in the east (right side). Due to the road's design, which aims to prevent flooding to ensure safe traffic, its elevation is consistently higher than the surrounding terrain. Consequently, if the road has an inadequate culvert system, it may obstruct the flow, leading to elevated water levels on one side of the road.

There is a significant change in flood depth in the vicinity of the Tra Khuc River (Area 1) for flood frequencies of P = 1–10%. This area was previously a drainage flood zone for the Tra Khuc River. However, the presence of the road obstructs the flow, resulting in severe flooding on one side and a drastic reduction on the other. Specifically, there is a 0.5-m rise in flood depth on the left side of the road and a 0.5-m decrease on the right side. The water level difference between the two sides of the road is approximately 1 m (in the case of P = 1%). This indicates that the drainage capacity of the culverts in this area is insufficient. In Section 5.3, we will explore a structural solution that involves adding more culverts to mitigate flooding.

In Area 2 (Nghia Dien and Hanh Thuan communes), the increase in flood depth varies between the two sides of the road for flood frequencies of P = 1–5%. Specifically, there is an increase of more than 0.5 m in the west and between 0.3 and 0.5 m in the east. For smaller floods (P = 10–20%), the area experiencing increased depth is primarily in the west, while the water level decreases in the east. Since Area 2 is densely populated, changes in hydraulic features and flood distribution will have serious consequences. According to the survey, the number of houses in areas with increased flood levels is quite substantial, as shown in Table 4.

For P = 1%, most of the increased flooding area is located in Nghia Dien commune (Area 2), resulting in the highest number of affected houses – 319 in total. This is followed by Nghia Ky commune, with 177 houses impacted. Although Hanh Thuan commune is also situated in Area 2 and the road traverses only a small portion of its land, it still experiences a significant impact, with 131 houses affected. Conversely, Nghia Trung commune reports the fewest affected houses, totaling just 21.

Furthermore, the data indicate that even with smaller floods (P = 20%), the road's influence on flooding remains apparent in Areas 2 and 3. The water level difference between the two sides of the road is still extremely high. This shows that even with minor floods, culverts in certain areas continue to be ineffective in draining water. As flood frequency increases, the negative impact of the road on flood distribution persists in the southern regions (Areas 2 and 3).

Another crucial element in a flood study is flooding duration, as it significantly affects the resilience of plants, animals, and humans. This study demonstrates that the road can prolong flooding, exacerbating its impacts. Since the road's construction, floodwater has increased rapidly in some areas but has decreased very slowly due to blockages. The distinct variation in flooding duration after the road was built is illustrated through water level simulation points on both sides of the road. The locations of these six flood monitoring points are shown in Figure 15, numbered T1–T6. The symbol ‘past’ indicates conditions before the road was built, while the symbol ‘present’ refers to the situation after its construction.

Conversely, at point T2, the flood flow appears later than it did before the road's construction (as seen in T2_present compared to T2_past). At point T3, the flood depth increased by about 0.9 m due to the road, creating a water level difference of roughly 0.9 m between the two sides of the road. Notably, there is no change in the timing of the flood's onset and recession (see T3_past and T3_present). At point T5, although the flood depth increased only slightly, the duration of flooding was significantly longer (see T5_past and T5_present).

In this study, three structural measures are proposed to mitigate flooding: (1) constructing new culverts, (2) dredging existing rivers, and (3) building a new flood drainage canal. The details of these measures are outlined in Table 5. These measures are aimed at residential areas experiencing significant flooding, as illustrated in Figure 15. The effectiveness of these flood mitigation measures is evaluated based on a flood frequency of 10%.

Simulation results indicate that the depth and duration of flooding can be minimized through the implementation of appropriate structural measures. Among these measures, constructing additional culverts is considered one of the simplest solutions for reducing flood impacts in localized areas. The effectiveness of culverts has been documented in various studies, including those by Conesa & García (2013); Aranda et al. (2021); de Jager & van Dijk (2024). Building new culverts enhances their capacity to manage stormwater, significantly mitigating flooding risks. However, potential disruptions to transportation and traffic during the construction of new culverts raise concerns. Therefore, it is essential to carefully evaluate the economic efficiency of this solution to ensure that it yields sufficient benefits without causing undue inconvenience.

The remaining two measures that do not affect traffic are dredging the river and constructing a new flood drainage canal. For the river dredging measure (Case 2), floods can be quickly drained, which reduces flooding duration and contributes to a healthier ecological environment. This traditional method offers several key advantages that are essential for effective flood management. Dredging removes sediment and debris, increasing the river's capacity to manage stormwater, thereby significantly lowering the risk of flooding. Additionally, it enhances water quality by eliminating pollutants, which supports healthy aquatic ecosystems. Dredging also stabilizes riverbanks by maintaining appropriate river dimensions, reducing erosion in the process. Furthermore, when conducted carefully, it can restore habitats for aquatic species, contributing to increased biodiversity (Alvin & Mardyanto 2018; Juarez et al. 2021; Lo et al. 2021). Meanwhile, the construction of a new canal (Case 3) may redirect flooding to other areas. It is essential to assess the land-use patterns and the capacity of the receiving water area to accommodate flooding, as well as the extent of potential damage. For instance, in this case, the area at the end of the canal experienced an increase in flooding of over 0.6 m. These cases illustrate typical structural measures designed to mitigate flooding. They are intended to serve as a technical foundation for research, development, and road rehabilitation efforts in other flood-prone regions.

The impact of roads on the extent, depth, and duration of flooding can be significant if not designed properly. This underscores the importance of prioritizing flood considerations during the design phase, particularly for roads traversing flood-prone areas. Many countries implement regulations and guidelines for road construction; however, these can vary widely in clarity and effectiveness. As a result, numerous completed road projects continue to exacerbate local flooding issues. The overall cost of constructing new roads is heavily influenced by the selection of drainage structures, such as culverts and bridges, along with the elevation of the road, which is closely tied to flood frequency. Opting for numerous bridges and culverts can increase costs while ensuring adequate drainage, whereas minimizing such structures may lead to increased or intensified flooding in adjacent regions. This situation can adversely affect both the environment and human lives. The objectives of effective flood management and cost efficiency often conflict, necessitating a careful compromise. Consequently, comprehensive and detailed guidelines are essential to navigate these challenges effectively.

Flooding can occur if the technical instructions are unclear or not mandatory during construction. For flood-vulnerable areas, a different approach to planning and engineering design is required for constructing roads compared to areas that are not regularly flooded. Roads passing through flooded areas must be built according to certain standards, and the scope of flood considerations should be broader at the basin scale. Currently, there are no specific national standards for designing roads that pass through flood-prone areas in Vietnam (Douven et al. 2009; Douven & Buurman 2013). Currently, design regulations only require road elevations to meet the design frequency for flood protection; however, there are no clear standards regarding the acceptable level of flooding that roads may cause in surrounding areas. This lack of guidance can result in unwanted risks in flood management. Furthermore, current regulations should incorporate the task of calculating simulation models to aid in the design of roads and culverts while also supporting environmental impact assessments to minimize adverse effects.

EIA is a crucial process in Vietnam, aimed at identifying, evaluating, and managing the impacts of projects on the environment and public health. While EIA is mandatory for major construction projects, the implementation and monitoring of these assessments remain limited. The quality of EIA reports is often inadequate, attributed to a lack of expertise or intentional oversight, complicating the approval process (Clausen et al. 2011; Pham et al. 2020). Furthermore, community participation is frequently insufficient, as local residents are often uninformed or excluded from the assessment process (Clausen et al. 2011; MONRE 2020). After approval, many projects fail to fully implement the mitigation measures committed to in their EIA reports, primarily due to a lack of rigorous oversight from authorities (Clausen et al. 2011). In addition, there is a lack of coordination among stakeholders involved in the EIA process. Sometimes EIA is only a formality and is conducted as an activity separate from the technical and economic aspects of project planning and design. There is inadequate collaboration between project owners and consulting units throughout the EIA implementation process (Nhi 2022). In many cases, the project owner has contracted and left it to the environmental consultant to carry out the EIA, while the legal responsibility for the content of the EIA report belongs to the project owner (Ngan 2022). This lack of close coordination can lead to inconsistencies within the EIA report itself, which may not align with the actual investment project details. Furthermore, environmental impact mitigation measures proposed in the EIA report are often not implemented (Pham et al. 2020). To enhance the effectiveness of EIAs in Vietnam, it is essential to increase awareness, strengthen the capacities of regulatory agencies, and promote greater community involvement in the assessment process (Kamijo 2022). Furthermore, it is important to recognize that environmental impact assessments are not the only tool available; effective flood response planning, natural resource management, and raising community awareness throughout planning and management processes are equally essential.

The presence of roads can significantly influence the magnitude and distribution of flooding. Inundation maps were created for various flood frequencies to investigate the extent, depth, and duration of flooding. The findings revealed that roads can obstruct water drainage in certain areas, resulting in increased water levels on one side and decreased levels on the other. In some instances, the difference in water levels can reach up to 2 m. Additionally, roads may prolong the duration of floods, contributing to more severe inundation.

Three typical structural measures have been proposed to mitigate flooding: constructing additional culverts, dredging rivers, and building new flood drainage canals. These measures have successfully reduced local flooding in some areas, with water levels decreasing by as much as 0.5 m, while also delaying the onset of floods and accelerating their receding. However, a limitation of this study is the lack of economic evaluations when comparing solutions.

Furthermore, there are notable constraints related to road construction in flood-prone areas, particularly the lack of coordination among stakeholders during the EIA process. To address these issues, Vietnam needs to continue researching and developing necessary regulations. Additionally, environmental impact assessments should be more stringent, and current standards must be refined to be more precise and comprehensive.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.