Floods and Structural Anthropogenic Barriers (Roads and Waterworks) Affecting the Natural Flow of Waters: Hydraulic Modelling and Proposals for the Final Section of the River Segura (Spain)

Abstract

Floods are the climate hazard that has the greatest socio-economic and territorial impact on the world. The root causes of these events are atmospheric and hydrological phenomena. However, human action usually aggravates their effects, as it alters the normal functioning of the river courses and water flows. The installation of road, rail and hydraulic infrastructures in a floodplain with no prior calculation or appropriate adaptation exacerbates the negative consequences of floods, increasing the extension of the flooded area and the height of the flood waters. This study addresses the problem of the barrier effect generated, on the one hand, by the layout of the N-332 road, as it is built at the same level as the ground, hindering the flow of overflowing water during episodes of flooding, and on the other hand, the channelling wall of the Segura River in the final stretch of its mouth, in the towns of San Fulgencio and Guardamar del Segura. These elements have aggravated flooding in this area. In order to analyse the consequences of the flood, IBER (v.3.3) software has been used to model a flood with similar effects to that of the episode of September 2019. The current situation has also been analysed with two openings in order to determine the effects that a future flood would have. After analysing the results, a proposal to correct the barrier effect of the N-332 road and the new channelling wall of the River Segura has been elaborated upon and then modelled. The results are positive and effective in reducing the negative effects of floods in the lower basin of the River Segura.

1. Introduction

Floods are the natural phenomenon that cause the greatest human and economic losses on a global level, as a result of meteorological and hydrological events combined with anthropogenic actions in a territory, particularly in the floodable areas of river courses [1,2,3,4,5,6,7,8]. Contrary to the case of developing countries, developed countries have been able to reduce the loss of human life over time, although the economic losses are greater year after year, derived from a greater exposure associated with the occupation of floodable areas [9,10,11], particularly in areas with a high economic value that coincides with areas of high risk, such as floodplains [12]. Since 1980, the accumulation of exposure due to economic growth and urbanisation has been the principal factor driving the increase in the associated damage [13].

Another type of anthropogenic action influencing the greater occupation of floodable areas is based on the regulation of river basins for their water usage and the fight against floods. This type of construction is carried out through control and evacuation engineering undertakings, such as reservoirs, dams, channelling, and derivation channels, among others [14]. These actions generate a “false sense of security” among the population of the river basin, leading them to believe that the flood problem is resolved and will not happen again. It is true that these types of actions are able to evacuate small and medium floods, depending on their design capacity, although they are usually overwhelmed or suffer damage in situations of extraordinary flooding, For this reason, the intense occupation of a river valley in an extraordinary flood situation will cause high levels of economic damage and the possibility of mortal victims.

In order to highlight the data that show the enormous effect of floods in terms of events, loss of human life, economic effects, and damage, the data available in EM-DAT (2024) have been consulted, based on an exclusive search for floods on a global level and different types of floods in Southern or Mediterranean Europe [15]. First, it should be indicated that the database does not contain all of the flood episodes recorded, so the data could be significantly larger.

On a global level, in the period 2000–2024, a total of 93,104 river flood events were recorded, causing 69,928 deaths and 1,266,543,017 casualties and an accumulated economic damage of USD 518,402,091, adjusted to 2024 [15]. With respect to the southern sector of Europe, including river floods, general floods, and flash floods, a total of 487 events were recorded in the same period. In these 487 events, a total of 2167 deaths were recorded, together with 8,625,773 casualties and economic damage amounting to USD 212,517,245 adjusted to 2024 [15].

On the other hand, in the current climate conditions, the effects of anthropogenic-caused climate change related to global warming should be taken into account, as in some territories, their effect is visible in the origin and recording of extreme weather events that cause floods [16]. In fact, some studies already indicate that the number of flood events in different world regions has already been displaying an increasing trend in recent decades [5,17].

In fact, there is a direct relationship between floods and climate change effects, and this has been analysed in detail in the JRC PESETA III and PESETA IV reports corresponding to Europe. The first report indicates that, since 1980, river floods in Europe have caused 4000 deaths and direct economic losses of more than EUR 150,000 million (values adjusted to 2013) and that the future risk is linked to climate and socio-economic factors [18]. Furthermore, it indicates that, in Europe, there are approximately 216,000 people exposed to river floods and flood damage each year worth EUR 5300 million [18]. In a climate scenario of warming of 2 °C by 2040, the impact of floods could be double, with 525,000 people exposed to floods annually and economic losses due to direct damage of EUR 13,000 million. In the period 2070–2100, it is estimated that a total of 700,000 people will be exposed to floods annually and that there will be direct losses of EUR 17,000 million for floods associated with return periods of 100 years [18]. As a consequence of the potential dramatic effects that could occur in relation to floods, it is necessary to take adaptation measures that will reduce the negative effects of floods on the world [19].

One of the European countries with the highest level of flood risk is Spain, holding fifth place in Europe in terms of the volume of population exposed to floods, with a total of 2.4 million inhabitants in 2018 concentrated in the coastal areas with a higher risk [20]. According to the data provided by EM-DAT for Spain related to river floods, general floods and flash floods for the period 2000–2024, a total of 23 events have been recorded, with 97 deaths, 19,798 casualties, and direct economic losses of EUR 5,240,503 [15]. Meanwhile, the Insurance Compensation Consortium in Spain (CCS) for extraordinary risks indicates that, for the series 1990–2023, the majority of the cases are related to floods (58.5%), and in 2019, a total of 1,370,860 claims had to be compensated, with an average cost of EUR 124,624 [21].

The three largest floods in terms of compensated damage in Spain were in August 1983 in the Basque Country, Cantabria, and Navarra, where the floods incurred a total of 25,664 claims with a compensation value of EUR 977,290,079 [21]. The second is that of September 2019, which affected the southeast of the peninsula, particularly the Segura basin (Murcia, the Region of Valencia, and Andalusia), being the one with the largest number of claims at 55,785 and a compensation value of EUR 478,921,488 [21]. The flood with the third largest compensation awarded by the CCS was also in the Segura basin (Murcia and the Region of Valencia) in November 1987, with a total of 18,800 claims and EUR 344,901,025 in compensation [21]. The first occurred in an urban environment, which explains the large economic losses and the greater compensation awarded by the CCS. However, in terms of the urban and agricultural areas affected, the Segura basin is ranked among the highest, which is evident in the size of the area affected by the floods in the mid-lower basin of the River Segura. This basin is located on the Spanish Mediterranean coast, and in 2019, the rainfall recorded was higher than that of November 1987 as a result of the effects of climate change, which have become evident in Spain. All of these data show the enormous exposure of the population in the Segura basin, particularly in its mid and lower sections.

An essential problem of flood events in river plains corresponds to the presence of architectural barriers, such as linear road–rail infrastructures and even hydraulic structures, which act as dams, retaining and accumulating water, influencing the duration and extent of floods, and in the worst case, generating regressive floods, altering the natural dynamics of floods in their flood plains, and which can affect population centres [22,23].

The international scientific literature has been working for years on the relationship between floods and the blockages caused by roads, railway lines [24], or bridges that are perpendicular to the flow of water and settled in the channels or their flood plains. There are a large number of studies and reports that deal with issues related to the impact (direct and indirect damage) [25]. However, the “dam effect” is only mentioned as one of the ways that the negative effects of floods are enhanced, although recently, there have not been many related works published in this line.

Few international examples and studies were found on strategies and measures for building infrastructure in river floodplains. Two general road development strategies, namely a strategy based on resistance and a strategy based on resilience, can be distinguished, both having their advantages and disadvantages. The resistance strategy, in principle, aims at preventing and regulating floods and, hence, has a strong impact on the natural floodplain dynamics, while the resilience strategy aims at minimising the consequences of floods but at the same time intends to maintain the natural floodplain dynamics as much as possible [26].

Analysing these aspects in greater detail, as regards the relationship between floods and roads, roads can be damaged by floods, but they can also increase the risk of flooding [23]. The flooding of a road produces two levels of consequences. On the one hand, injuries to people and the destruction of vehicles can occur. On the other hand, traffic disruption can have serious indirect consequences related to rescue and emergency services [22].

Roads can also act as obstacles during a flood, affecting the natural flow of water in several ways. (1) For obstruction of natural flow, if a road runs through a flood zone without sufficient drainage or water passages (culverts, bridges, etc.), it can disrupt the natural flow of water, creating reservoirs on one side of the road. This increases the depth of the water and the pressure on the road structure, which can lead to damage or overflow into adjacent areas. (2) For increased flooded areas, in some cases, roads can divert or redirect the flow of water into areas not intended for flooding, increasing the area affected. This is especially problematic in flood plains, where water should disperse naturally. (3) For erosion and scour, if a road significantly disrupts the flow of water, the velocity of the water can increase at certain points, increasing the risk of soil erosion and undermining the road’s foundations. This can cause parts of the road to collapse. (4) For the blockage of drains, roads, especially those with poor drainage infrastructure, can act as barriers by blocking natural drains or culverts, worsening the accumulation of water on the surface. (5) For the levee effect, roads can sometimes act as artificial levees, holding back water in certain areas and preventing it from dispersing into natural drainage areas, increasing the risk of localised flooding [23]. To mitigate these effects, appropriate drainage designs are often incorporated, such as bridges, culverts, or underpasses, which allow water to flow under the road or by elevating roads [23]. For these reasons, developing roads in a floodplain requires a different approach in terms of planning and technical design compared to the building of roads in areas that do not flood [27].

This problem is usually characteristic and common, for example, in the river plains of the large rivers of the world and even in the shortest river flows or intermittent flow rivers in the Mediterranean regions. In fact, in these regions, the problems are greater, given that for most of the year, no water flows through these river courses (dry riverbeds), which are reactivated in situations of intense and torrential rain. The absence of a constant flow in a dry riverbed, characteristic of the Mediterranean, leads to anthropogenic actions in the territory that can be installed perpendicularly to the natural water flows, acting as obstacles, barriers, or dams, such as road bridges. Furthermore, it should be noted that the form of the river valley is a determining factor for the behaviour of these road or rail barriers in the floodplains. In a valley with a “V” shape, roads can act as complementary barriers to the flooding of the river and can be covered and damaged.

On the other hand, river valleys with their courses above their floodplains are a rare occurrence and are generated by the geomorphological and climate conditions of the basin, together with anthropogenic actions through the construction of protection dykes, particularly in the lower sections of the river. These types of river valleys are the most problematic and difficult to manage in the case of a flood event. They cause the most problems and generate catastrophic floods, referred to in the scientific literature by the term out-floods. On a global level, there are certain lower sections of the Yellow River (China), the Mississippi (USA), and the Nile (Egypt). In Europe, the River Rhine (Germany and the Netherlands), the River Po (Italy), and the River Maas (the Netherlands) are good examples. And in Spain, we can cite the mid-lower section of the River Segura (Murcia and Alicante), the lower section of the Júcar (Valencia), the lower section of the River Guadalquivir (Seville), the River Ebro downstream of Zaragoza, and the river delta of the Ebro (Tarragona) [28].

The overflowed waters of the river in these types of river valleys expand until they get stuck in spaces where the roads are higher than the croplands of the flood plain (space where the water flow circulates), increasing the height and duration of the flood and aggravating the negative effects that the floods have on the population. If the roads are sufficiently resistant, the height of the water increases until it either covers the road or not and continues its course until the next road or barrier in the river valley. This cycle is repeated many times. In the case where the roads are not resistant enough, the waters make holes in them, opening up a path to flow through. However, the majority of the roads comply with international guidelines that are not adapted to floodplains, and there is a lack of integrated planning in the design of road and rail infrastructures [22].

After the roads have been covered or destroyed, the only way that the flood waters have to return to the course is via the lowest point of the river valley joining with the course. However, sometimes, the engineering works carried out in river courses, particularly in their final section, depending on how they are placed with respect to the natural flow of the water, can also act as barriers, preventing drainage and exacerbating floods in this section, in terms of their height, duration, and extension, among other factors, in the same way as the roads.

As previously indicated, the greatest exposure is related to urban developments that occupy floodable spaces and also the development of road infrastructures. In situations of flooding in river valleys, national roads and motorways and even provincial and local roads act as barriers or dykes, preventing the natural flow of the water. Even hydraulic infrastructures, such as channels or canals, whose sides are higher than the terrain, also act as barriers, preventing the water from returning to the river course when the flood subsides, as analysed by some authors studying the floods in Sri Lanka [29]. This same problem was experienced in the Mississippi canals after the catastrophic event of the Katrina Hurricane in 2005 and the subsequent floods [30]. A similar situation arose in the major flood occurring in July 1997 in Wroclaw (Breslavia) in Poland, part of the Czech Republic and Germany, generated by the River Oder, whose waters broke retention walls or lateral dykes, also preventing the return of the waters to the course once the flood had receded [31].

This leads to an increase in the depth of the flood water in urban and rural spaces, further aggravating the flood damage and increasing the time that the water remains or generating regressive floods.

In Spain, during the October 1982 flood caused by the Tous dam failure, the water could not be easily evacuated due to the presence of the A-7, which aggravated the effects of the flood. In September 2019, the presence of the N-332 national road also acted as a dam and generated a retrogressive flood that affected the town of San Fulgencio in the Segura plain, along with many other roads in the lower Segura basin. This also happened in the November 1987 flood. In addition, the wall that separates the old and new Segura riverbed after the channelling works carried out in the 1990s also acted as a barrier or dam, generating another retrogressive flood from the old Segura riverbed to the town of San Fulgencio. The same thing happened during the flood in November 1987. Furthermore, along the entire Mediterranean coast, the N-332 road is built perpendicular to the water flows of the rivers, which is why it presents the same problems in other river basins, such as the Júcar or the Ebro. This highlights the importance of taking into account both road and hydraulic infrastructures that act as walls or dikes.

Finally, we should note the importance of the passing of Directive 60/2007/EC, of 23 October on the assessment and management of flood risk, which assumes that floods are inevitable, and therefore, measures must be sought to diminish or reduce the negative impact of floods as much as possible [32]. This directive was transposed into the Spanish legal framework through the passing of Royal Decree Law 903/2010 of 9 July. The directive indicates a series of steps when elaborating a risk map, taking into consideration historical floods, the characteristics of the basin, and a description of the major floods occurring in the past in order to predict the adverse consequences of similar events in the future [33]. The regulation corresponding to the construction of roads in Spain is the Instrucción de Carreteras (Spanish Road Drainage Regulation) (Regulation 5.2-IC) and includes a specific section for calculating the conditions of surface drainage [33].

This study seeks to reconstruct the flood of September 2019 in the final section of the River Segura in the lower Segura basin, where the presence of the N-332 road and the channelling wall acted as retention walls or barriers perpendicular to the water flow, generating (a) an increase in the height of the water, (b) a regressive flood that affected the village of San Fulgencio, and (c) longer duration of waterlogged conditions and flooding, with water remaining for four weeks after the flood.

Therefore, the objectives of this study are the following:

• conduct a simulation of the September 2019 flood to identify the buildings that act as barriers;
• carry out a simulation of a flood similar to that of 2019 but taking into account the two current openings in the wall separating the old channel from the new one and checking whether they are really effective;
• carry out a simulation of a flood similar to that of 2019 but applying the proposals put forward in this article;
• verify the effectiveness and viability of the proposed actions.

The results obtained from fulfilling these objectives have been used to elaborate a comprehensive proposal to respond to the occurrences in this lower section of the Segura basin in order to reduce the negative effects of floods as much as possible.

2. Materials and Methods

2.1. Study Area and Contextualisation of the Problem in the Study Area

The area of study is located in the final section of the Segura basin, at the mouth of the River Segura and its nearby river valley, on the left bank of the River Segura, including the municipalities of San Fulgencio and Guardamar del Segura in the district of Vega Baja del Segura (Alicante) (Figure 1).

In order to determine how overflowed water functions in this space, it is necessary to highlight a series of characteristics pertaining to the area of study.

First, the area of study belonged to a former lacustrine area called the Sinus Ilicitanus (Figure 2), and the powerful activity of the floods of the River Segura and the Vinalopó, together with the torrential nature of the waters of certain gorges and watercourses, began to fill the lacustrine area, together with the formation of a shoal from Santa Pola to Guardamar del Segura, which favoured the sedimentation process [34].

Second, the altitude of the river valley, with respect to the sea level, is very low. In half of the valley, there are altitudes of 10 msnm and in the area of study between 2 and 0.40 m. This means that the average slope on the left bank of the River Segura is very slight (0.05%), so the flood waters circulate very slowly, hindering their drainage. This is why the whole of this sector was formerly a lacustrine area. Over the centuries, drainage and retention processes were implemented through the construction of irrigation channels called azarbes or drainage channels, which sought to move the excess irrigation water, collect the overflowed water, and move stagnated water in certain spaces (Figure 3).

They were channels that were designed to supply the lacustrine spaces and move stagnated water with a maximum capacity of approximately 10 m³/s. These types of channels are characterised by being at the same level as the ground so as to be able to collect surface water. The presence of the azarbes implies the existence of former lacustrine areas that were supplied by the drainage network designed between the end of the sixteenth century and the twentieth century. However, given the many swells and floods generated by the River Segura, together with the insufficient capacity of these channels, the population has built concrete channels (impermeable) and raised their walls, so the surface waters cannot join the channel and remain trapped, causing floods. Moreover, they cannot join the azarbes through infiltration as they did when the channels were made with soil, given that the current concrete channels are impermeable.

Third, as previously mentioned, the River Segura is higher than its river valley on the left bank, which implies that, in the case where it overflows or the channelling wall beaks, out-floods will occur and the waters cannot return to the river when the flood recedes.

Fourth, the overflowed waters circulating in the final section of the valley encounter several obstacles in the form of road infrastructures. In the event of extraordinary swelling and surface runoff, the overflowed waters encounter the presence of the national road N-332. The azarbes cross the N-332 through a very small space with a low hydraulic capacity, acting as a retention wall or dam and hindering the natural flow of the water. The only way the water is able to cross the N-332 is to accumulate, increase its volume, and pass over the road to continue its course. However, during this process, the overflowed waters continue to accumulate upstream, joining the trapped water and raising the height. The waters recede and begin to occupy and regressively flood the whole of the area upstream of the N-332. This includes the closest municipality, that of San Fulgencio.

Fifth and finally, when the waters cross the N-332, they continue to circulate until they join the old or original course of the River Segura. In this space, there is a breakwater that divides the old or original course of the river and the new course of the Segura with a greater capacity. However, this wall, similarly to the N-332, acts as a barrier, as it is placed perpendicularly to the natural flow of the waters that incorporate into the river at its lowest point. In this case, it would be the new course (Figure 4).

2.2. The Most Important Aspects of the Flood of September 2019

On 11 September 2019, the State Meteorology Agency activated a red alert due to a situation of torrential rain as a result of the formation of a pocket of cold air detached from the main jet stream (cold drop or DANA) in the Alboran Sea. The atmospheric conditions favoured the formation of a mesoscale convective system, generating two strong downpours of torrential rain for three hours on 12 and 13 September. In the first downpour, an average of 300 mm in three hours was recorded on 12 September. On the following day, another 200 mm in three hours were recorded, with values of a total of 525 mm being recorded in the town of Orihuela and an average of 300 mm in the rest of the lower basin of the Segura. Furthermore, intense rain fell in Murcia in the mid-basin. All of this reactivated many watercourses, generating floods in the mid-lower basin of the Segura. It also gave rise to the swelling of the River Segura, which overflowed at different points due to the many ruptures of the channelling walls constructed after the floods of November 1987. There were six ruptures in total. The largest and the one that drew the most media attention was in the town of Almoradí, where the waters began to fill the floodplain of the River Segura. This affected the towns of Almoradí, Daya Nueva, Daya Vieja, Dolores, and San Fulgencio and part of the crop fields of Guardamar del Segura. When this rupture was filled with soil deposits, there was a second breakage, which aggravated the flood situation in the floodplain. Moreover, there was a discharge of water from the Santomera reservoir, which was practically overflowing. This could have aggravated the flows of the River Segura to some extent and, therefore, also the flood.

It should be noted that, during the flood, the many roads running through the floodplain of the River Segura acted as architectural barriers, blocking the natural flow of the waters and aggravating the floods in certain spaces. We know that the roads that exacerbated the negative effects of the floods in the lower basin of the Segura were the CV-91, CV-911, CV-912, CV-859 CV-860, CV-855, CV-861, CV-930, the N-340, the N-332, the A-7, and the AP-7, among many other local or provincial roads of Alicante. All of these roads gave rise to enormous problems, particularly in the towns of Dolores, Daya Nueva, Daya Vieja, San Fulgencio, and Guardamar del Segura. In fact, the municipality of Daya Vieja has a rectangular shape and is surrounded by roads, which converted it into a lake.

In the flood of September 2019, the overflowed waters were blocked by the N-332. This led to an increase in the height of the flood water, and the land upstream was flooded through a process of regressive flooding. When the height of the water exceeded that of the N-332, the water cut off the traffic in several sections and continued to circulate towards the old course of the River Segura (Figure 5).

Of all of these roads that acted as a barrier, the N-332 is of particular interest for this study, together with the breakwater that separates the old and new courses of the River Segura.

The presence of the wall or breakwater that divides the two courses also acts as a wall in situations of a high water level, leading to regressive floods between the N-332 and the old river course. This also influences the waters accumulated upstream of the N-332, aggravating even more the volume and extension of the flood (Figure 6).

The image of the event in the area of study in September 2019 shows the complete flooding of all of the adjacent land (Figure 7).

2.3. Methodology for Preparing and Calculating the Hydraulic Modelling Using IBER (v.3.3) Software

This study has used the GIS called QGIS. First, the information provided by the National Geographic Institute based on the most recent orthophotos (aerial photographs) and digital models of the terrain generated with airborne LIDAR points with a resolution of 0.5 m of precision was downloaded. Second, the orthophoto was added to the QGIS program, and a vectorial layer was created in a square shape in order to cut the area of study. After making the cut and obtaining a new orthophoto of the area of study, another shapefile layer was created in which the geometries of interest of the terrain were mapped for modelling in the IBER program. With respect to the digital models of the terrain (DTM or DEM), the layers provided by the airborne LIDAR technology with RGB-coloured points (red, green, and Blue) were downloaded. They were incorporated into the QGIS program, and using the tools provided by LAStools, the point layer was converted into a digital model of the terrain. This was conducted layer by layer for the area of study, and a digital model of the terrain was created based on the points with values of 2 and 6 (where the value of 2 refers to the ground and the value of 6 to the buildings). The result obtained is a DTM that includes the altitude of the ground and the altitude of the constructions. This is the most valuable DTM for conducting hydraulic modelling. When converting each layer of the LIDAR flight into individual DTMs, it is necessary to use the process of combining raster images to obtain an all-in-one DTM. Using the same layer cut in shape format as that used to cut the original orthophoto, a cut in the all-in-one DTM was made, obtaining a detailed DTM of the area of study (Figure 8).

To finalise the process of preparing the layers in the QGIS, the geometry needs to be mapped. This can be done by constructing polygons in shape format or duplicating the layer cut and making cuts in the resulting shape layer. The polygons should be drawn based on the results that are sought. In the case of this study, geometries have been elaborated on the new course of the River Segura, the old course, the azarbes, the urban and industrial areas, and the floodplain area, among others (Figure 9).

After the necessary layers had been obtained to conduct the modelling, they were included in the IBER program. This program is a bidimensional hydraulic model (space and time) for simulating the free surface flow of rivers and estuaries [35].

The reconstruction of the 2019 flood was based on including an inflow of 150 m³/s for several hours to simulate the flooding of the Segura River. Since the overflow of the Segura River occurred upstream, the other inflows of water were made at each of the random areas with a flow of 10 m³/s each.

In the initial conditions, the value of the added draught is zero.

With regard to roughness, the corresponding land use has been introduced (river, meadow, residential, industrial, and agroforestry areas, among others), in which the values of Manning’s coefficient or roughness index have already been assigned (river: 0.025, meadow: 0.05; residential: 0.15, industrial: 0.1; and agroforestry areas: 0.1).

Regarding the creation of the mesh, an unstructured mesh with different sizes of triangles was used. For greater precision in the channels of the azarbes, and in the river, the mesh size assigned was 10 m.

A DTM from the LiDAR flight information with a precision of 0.5 m has been introduced into this mesh. The mesh adopts the form of the DTM to carry out the hydraulic modelling. This generates a flood wave that is very similar to what happened in the September 2019 episode. It is also corroborated by the flow rates reached in these areas.

After obtaining these results, a similar flood episode was also recreated, taking into account the two current openings made after the flood of September 2019.

After analysing the results obtained, it is clear that the N-332 road and the wall of the Segura River channel act as walls that retain water, preventing its natural outflow and worsening the effects of flooding. For this reason, the removal of the wall of the Segura River channel that separates the old channel from the new channel is proposed, creating a single channel.

To carry this out, the height of the wall separating the old and new courses of the River Segura was reduced at the current points where it is located but maintained at a height above the new channelling of the River Segura. After obtaining the results, the same flood was modelled with five openings, adding these outflows at the points where the azarbes discharge their waters to the old course of the River Segura, and others were also added at certain points downstream. Given that the results continued to be fairly negative, although each modelling exercise improved the previous situation, the existing wall separating the old course from the new course of the River Segura was eliminated completely, and the two courses were joined, forming a much wider course. Subsequently, the breakwaters existing at the river’s mouth were also eliminated, as they generated a blockage of the outflowing water and were poorly designed inversely [36] (Figure 10).

After improving and resolving the problem existing in the rive course, the N-332 road was raised. To carry this out, the hypothesis that the N-332 road should be raised uniformly with a bridge with pillars resting on the floodplain was used. The assigned altitude in the whole of the section of the N-332 is seven metres above sea level, as this height enables the entry of 5m tall machinery for the periodic maintenance of the road and the area under the bridge. In order to simulate the presence of pilasters or bridge walls, the “drainage” option of the IBER program was used in order to simulate the space existing between the pilasters. Due to the lack of information about the placement of the pilasters, spaces were made between them of 15 m in length and 5 m in width, creating a large capacity for the evacuation of water (Figure 11).

Moreover, the terrain had to be modified to the same height in the cropland sector between the final azarbes to enable an outlet due to the obstacle of several rural lanes and constructions (Figure 12).

3. Results

3.1. Modelling of the Flood of September 2019 and a Possible Similar Flood in the Current State

The first modelling conducted in the area of study sought to simulate the flood occurring in September 2019. The assigned data that best adjust to the real flooded area are a flow of 150 m³/s in the River Segura in one hour with the same level maintained for the 12 h of the simulation and, in the azarbes, a flow of 10 m³/s in one hour, with this level maintained for the 12 h of the simulation. The results obtained show each 30 min in a simulation of 12 h.

The simulation after 3.5 h is noteworthy, as this is when the flood waters make an impact on the N-332 road. It may be observed that this is when an increase in the height of the water begins upstream of the road, which produces a barrier effect. The waters crossing the N-332 do so through the spaces in the azarbes or over the road itself (Figure 13).

The waters that cross the N-332 continue their course until they reach the old course of the Segura, where they meet the wall that separates the old course from the new course, generating a regressive flood of the old course, which subsequently, prevents the discharge of water from other azarbes and overflowing water from the river. This generates the accumulation of water in the old course of the Segura, hindering the drainage of the azarbes. This gives rise to their overflowing in the adjacent areas, flooding the whole sector between the N-332 and the old course. Furthermore, the prevention of the discharge of waters from the azarbes generates a greater accumulation upstream of the N-332, reaching heights of two metres or higher in this sector (Figure 14). This leads to out-floods reaching the village of San Fulgencio.

This simulation perfectly reflects what happened in September 2019 and the problem of the barrier effect that roads and hydraulic infrastructures generate if they are placed perpendicularly to the natural flow of the waters.

As previously mentioned, in order to drain all of the water accumulated in this sector, the basin authority made two openings in the wall separating the old and new courses. These two openings were able to discharge a fair amount of water, alleviating the flood in the lower Segura basin. In the following months after the flood episode, the basin authority constructed two floodgates to discharge water in case of a future flood episode.

Analyses should be conducted to verify whether this solution is viable in a future event. To carry this out, modelling has been conducted that incorporates these two openings (Figure 15).

This moment has been chosen because the waters have occupied the whole of the old course of the River Segura and have begun to increase in height, despite the two existing openings. This shows that the two openings are insufficient for discharging water during a flood and that, after the flood, it is probable that it will take many days or even weeks to drain the flood water.

Therefore, the effects continue to be negative in terms of material damage and even the loss of human life due to the height of the water and the duration of the flood. Modelling was also conducted with six openings at different points, and although more water can be discharged than with two openings, the same problem exists due to the presence of the complete wall that separates the two courses of the River Segura. For these reasons, it was decided to conduct a series of modelling exercises that enabled these problems to be resolved by eliminating all existing barriers.

3.2. Proposal of a Solution to the Existing Problem of the Barrier Effect of the N-332 and the Hydraulic Infrastructures in the Final Section of the River Segura

Both the effects occurring in September 2019 and the modelling carried out highlight two important problems, namely the barrier effect of the N-332 and the barrier effect of the hydraulic infrastructures of the wall or breakwater of the new course of the River Segura that separates the old and new courses.

Arriving at a definitive solution that best resolves the problem of this geographic space in terms of floods has required many simulations. As shown in the previous section, the two openings that currently exist are insufficient. Making six openings causes a greater evacuation of water, although major problems of accumulated water and regressive floods continue to exist. Therefore, it was decided to eliminate the breakwater that separates the old and new courses, joining the two at the same height as the new course of the River Segura.

The result of this simulation revealed a high level of evacuation that resolved the blockage problem of the existing wall. However, at the river mouth, the inverse breakwater, which can be observed in the orthophotos, constitutes an obstacle to the natural outflow of the waters. In order to resolve this problem, the current inverse breakwater was removed, simulating the old natural mouth of the River Segura. The result of the modelling enables the natural alleviation without the problems of the River Segura, the drainage of the overflowed waters, and the azarbes to the only constructed course. It shows that the waters of the river discharged into the sea at the river mouth recovered their outlet towards the south, depositing sediments dragged by the river that, over the centuries, could regenerate the beaches of Guardamar del Segura as an adaptation measure to climate change.

The barrier effect of the hydraulic infrastructure is resolved. However, the barrier effect of the N-332 road has yet to be resolved.

The nodes in the grid elaborated in the IBER program were manually modified at a height of 7 m to make space for possible machinery for maintenance tasks that are between 3 and 5 m tall. The starting hypothesis is that there is a pilaster every 15 m that supports the bridge raised to 7 m. Therefore, different rectangular culverts are added with a base or length of 15 m and a height of 5 m. This technique is used for a total of 11 water outlets in the azarbes and 34 spaces or water outlets between the bridge pilasters.

Several interesting points should be highlighted with respect to the results obtained in the modelling.

First, the single course of the River Segura and the elimination of the inverse breakwater at the river mouth alleviates the swelling of the course of the River Segura with no problems. Furthermore, it may be observed that, when the water discharges into the Mediterranean Sea, the waves of the water move towards the south and southeast, as they would do naturally.

Second, the moment of the flood corresponds to six hours of modelling receiving constant flows. The surface overflow of the water and that of the azarbes seems to accumulate in the sector upstream of the N-332 (red circle), due to the depressed morphology of the terrain and the contributions of water constantly received. However, the water passes without difficulty through the simulated spaces between the pilasters of the bridge of the N-332.

Third, downstream of the N-332, the terrain had to be modified, as it was a depressed area where the water accumulated. The modification of the terrain is conducted in a way that creates a natural slope towards the single course of the River Segura (green circle). From this moment, the overflowing waters of the river before the N-332 and the water from the adjacent azarbes begin to accumulate at this point, temporarily increasing its height (Figure 16).

After 12 h of simulation, the flow of the entry into the azarbes stops, simulating that the flood waters of the river at another point upstream of the lower basin have reduced their level or have clogged the rupture in the wall. The maximum flood point is represented at 14 h of simulation, where we can observe that the contributions of overflowed water from the river and from the azarbes raised the height at which the River Segura flowed. This shows that the water is being discharged without any difficulty (Figure 17).

From this moment, in the following hours, a major reduction in the height of the overflowed surface water, the flood water from the azarbes, and the waters upstream of the N-332 can be observed twenty hours after the beginning of the simulation.

If we compare Figure 16 with Figure 17, we can observe how the height of 1 m has been reduced to approximately 0.70 cm in just six hours. We can also observe that the depth of the flows of the River Segura has also reduced.

Twenty-four hours after the beginning of the simulation, the height of the water has diminished considerably in the section between the N-332 and the course of the River Segura, dropping to a height of around approximately 0.4–0.6 m. The same occurs upstream of the N-332, where the heights fall from almost 2 metres (simulation at 14 h) to 0.70 metres (simulation at 20 h) and, ultimately to 0.5–0.6 metres (simulation at 24 h) (Figure 18).

The large decrease experienced in this problematic area shows that the proposal is feasible, efficient, and functional. There are no modelling results beyond 24 h, but the decreasing trend of the final 12 h of simulation shows that in barely 48–72 h at the most, a large part of the flood waters would have disappeared, with just a few waterlogged parts remaining in certain deeper areas.

Finally, a sketch has been drawn manually of the proposal to elevate the N-332 road. In order to obtain a digitalised drawing of the manual sketch of the possible future proposal, the artificial intelligence tool ChatGPT has been used. Below is a photograph of the current state of the road, in which we can clearly observe its layout at ground level acting as a barrier, and a sketch of the proposal to resolve this problem (Figure 19).

The proposal to elevate the N-332 and eliminate the wall that separates the old and new courses of the River Segura is shown to be efficient and feasible in terms of the hydraulic solution to the problem existing due to the barrier effect of the road, rail, and hydraulic infrastructures. Reducing the time that the flood waters remain in the space enables a reduction of the material, human and animal damage in the case of a flood episode, facilitating their recovery and the recommencement of land cultivation.

4. Discussion

This article shows the existing problem of flooding in the Segura basin, where the presence of various linear barriers (roads, railways, or hydraulic) can aggravate the effects of flooding. The approach proposed in this article can be extrapolated to other areas with problems related to flooding and the presence of roads or hydraulic infrastructures that act as retaining walls or dams to the natural flow of overflowing water. Hydraulic modelling software (IBER, HEC-RAS, SWMM, Infoworks ICM, FLUMEN) is a very useful tool to reconstruct past floods and develop modelling proposals to determine their viability in terms of flood risk in a flood zone.

This approach has been applied, as an example, to a study area corresponding to the final stretch of the Segura River, where the presence of the N-332 road, which acts as a dam, prevents the natural flow of overflowing water and where the channelling wall of the Segura river occurs. This causes the floods to increase in depth, generating regressive flooding and worsening their effects. Hydraulic infrastructures, such as channelling or the channelling of rivers, can also act as barriers or obstacles that prevent water from returning to rivers, worsening their effects and negatively affecting the population and economic activities of a territory.

Once the situation has been analysed, a scenario is recreated in which the proposed proposals are applied, and their effectiveness and viability are analysed. In this case, a significant reduction in droughts and residence time is observed, which would allow a large part of the study area to be drained in two or three days, instead of floods lasting between two weeks and a month in this area.

The ideas and proposals put forward in this article may lead to taking similar measures in other areas with similar problems and even following the same approach for other particular problems, where this type of risk analysis can work to carry out better flood risk management. It also provides the basis for the risk analysis posed by the European Floods Directive (Directive 60/2007/EC on flood risk assessment and management).

The methodology used could be subject to debate with respect to the data used more than the modelling process in the IBER program. This is because it is possible to modify the initial and boundary conditions and roughness index, obtaining similar results or a greater aggravation or reduction in the heights of the floods in the area of study. The flow of the River Segura introduced is 150 m³/s, as there are gauge readings of the Confederación Hidrográfica del Segura in Rojales of 125 m³/s after the rupture of the retaining walls in Almoradí. In this case, a larger flow was added to aggravate the swelling of the river, so as to analyse whether the drainage of the azarbes would be blocked by a greater flood or not. In this case, the modelling exercises show that they drain without difficulty in the case of a swelling of 150 m³/s.

The flows assigned to the azarbes (10 m³/s) correspond to their maximum capacity, although some could have a greater or lesser hydraulic capacity. Therefore, the data and results could vary, further aggravating or reducing the extension and height of the flood. However, establishing 10 m³/s for all of the azarbes gives uniformity to the swelling of the river, which perfectly simulates a wave of overflowed water circulating on the surface, without giving greater or less importance to a particular azarbe. By maintaining this uniformity, a progressive and uniform flood is obtained with which to gain an understanding of the behaviour of the water in this problematic space.

The elevation of the N-332 road could vary in height. However, the simulation considers 7 m, with a lower board at 5 m and with a free space of between 0.5–1 m and 5 m, in order to allow for the access of maintenance and cleaning machinery, which can measure between three and five metres tall. In order to resolve the problem of the N-332, it is sufficient to elevate the road enough to leave a free space where the waters can flow without being hindered. One debatable point is that the distance between the pilasters when using real pillars could differ from the simulated free spaces between the pillars in this study, which could partially modify the results. This study establishes a uniform distance of 15 m. However, it is possible that there are regulations that establish a shorter or greater distance, which could affect the free space where the flood waters can circulate. Nevertheless, the result obtained is fairly positive, as it allows the water to drain easily when this free space between the pillars is simulated.

The assigned land uses could also be modified, adjusting the roughness index and runoff coefficient established by Manning, generating a greater or lesser runoff.

The same is the case when assigning the grid size. If the size of the grid were larger, the results obtained would not be so detailed.

On the other hand, the advance and acceleration of the construction of roads in river plains seriously aggravate the effect of floods, generating structural damage, such as erosion in the road surface or soil displacement, the interruption of traffic due to impassable roads, or damage to bridges [5].

The results obtained in this study show the problem existing during the episode of September 2019 in the final section of the lower basin of the Segura as a result of the barrier effect of the roads and hydraulic infrastructures in the floodplains. This is particularly the case where the courses are located above the river valley. As previously indicated, the fact that the courses are above the river valleys is due to geomorphological causes and anthropogenic actions related to the heightening of the retaining walls. This has given rise to a situation of high flood risk for the population and economic activities established in the floodplains. Therefore, it is considered essential to undertake the proposed actions in order to reduce or minimise the negative effects of floods.

The socio-economic impact of these actions would probably be high and costly. It involves raising the N-332 road and removing the wall and the breakwaters from the Segura riverbed, with the environmental impact that this entails, in addition to the enormous investments. However, the population of the Vega Baja del Segura would support these actions because they are aware of the barrier effect of both infrastructures, especially, the raising of the N-332, which is a historically popular demand. In November 1987, a similar flooding episode occurred, and the mound of the old Segura riverbed hindered the outflow of water from the study area. Therefore, they had to dynamite several points of the river mound so that the water could evacuate towards the sea and reduce the flooding in the Segura floodplain.

This problem occurred in September 2019, where, to release the water, they had to open two floodgates to evacuate the water. The models show that, if a similar flood were to occur with both floodgates built, the situation would improve but would be equally serious because the problem is the entire wall.

These proposed actions can only be carried out by the Spanish Government, which has the authority to carry out these improvements and the financial capacity to do so, although it can count on the collaboration of other public administrations. For its part, the N-332 road is a national road, so it depends on the management, maintenance, and improvement of the Spanish State. On the other hand, the river basin authority (Confederación Hidrográfica del Segura) is responsible for the Public Hydraulic Domain, which depends upon the Spanish State. These are, therefore, bold bets and large investments to reduce the effects of flooding in the study area and the fluvial plain of the Segura River.

There are several examples on a global level where a road has acted as a barrier to floods, and after the elevation and modification of the section of the road in question, the problem has been successfully resolved. One example is that of the I-10 in Louisiana (USA), particularly in the areas close to New Orleans, which are vulnerable to floods due to their proximity to the sea and river systems. In order to resolve this problem, sections of the road were elevated, and more efficient drainage systems were installed to enable a faster water flow away from the road. This action has generated a reduction in the structural damage caused by flooding [37]. Another example is the A-8 motorway in Germany, one of the principal routes in Bavaria, which was prone to flooding as it passed through several hydrographic basins. In a key section, the road was elevated by several metres, and large drainage tunnels were added under the motorway to channel the water effectively during intense rains. Buffer zones were also created where the water could accumulate without affecting the road. Since the implementation of these improvements, the A-8 has experienced fewer interruptions due to floods, maintaining the traffic flow during storms [38]. Another example is National Route 1 in Thailand, where monsoon rains generated major floods. Raised bridges and secondary roads were constructed, and the drainage network was improved. This led to the road being closed to traffic much less frequently during intense rains [39]. Finally, on the banks of the Júcar (Spain), after the flood of October 1982, a consequence of the failure of the Tous dam, and after another flood episode in 1987, works were carried out to raise the A-7 and create water passages between pillars, which have worked successfully. No further flooding has occurred on the banks of the Júcar after this action. This proposal for action was contemplated in the Plan de Defensa Contra Inundaciones en la Ribera del Júcar (1987), which could be considered as a current flood risk management plan (FRMP) but, in the 1990s, was much more focused on engineering works of a structural nature, although it provided some green and holistic solutions [40]. Subsequently, in the successive Management Plans for the Júcar River Basin, these actions have been complemented with others. In the last cycle, proposals have been made to open up the basin for extraordinary floods with a return period of 250 years [41].

As we can observe from the successful practical examples indicated, the construction of bridges and the elevation of roads is necessary to enable the waters to pass and to create roads adapted to the needs of the territory, taking floods into account [5]. In fact, some kind of international guideline or regulation should be created for the construction of roads that are above floodplains, understanding this space to be completely anthropogenic, with urban nuclei and, particularly, crop fields that are at a lower height than the roads.

Some reports that address these issues highlight the need to elaborate specific actions according to the individual characteristics of the road and the floods, such as for example, converting roads into sponges or floating bodies as an alternative to the traditionally constructed roads with impermeable characteristics and also offering additional benefits in terms of adapting to climate change [6].

It is indicated that floating roads are usually more flexible than bridges, although they can be elevated to facilitate the flow of water or be equipped with gates that enable the passing of waters, such as the example of Afsbritdjk (dyke and 30 km long road surface that crosses the bay in the northwest of the Netherlands) [6].

Meanwhile, sponge roads are those made of permeable construction materials and enable part of the flood water to be absorbed. The construction of many roads with this type of material would enable the water to infiltrate into them as if they were a sponge, reducing the flows circulating [6]

Other authors recommend carrying out resistance and resilience actions on roads [22]. Resistance actions consist of maintaining the road’s impermeability, reinforcing or protecting them so that the floods do not damage them [22]. These types of actions would be appropriate, for example, for protecting certain specific areas where the damage (human and economic) is greatest. In these circumstances, it is of interest that roads are completely impermeable and that they act as a wall to protect a space [22].

Meanwhile, resilience actions are those that seek to reduce the impact of floods as much as possible, using permeable materials on the roads, elevating the roads or creating sufficient space for the waters to cross them without difficulty [22]

It is probably necessary to carry out mixed resistance and resilience actions depending on the problem and context to reduce the negative impact of floods. In those spaces where it is preferable to retain the waters so as to avoid greater damage downstream during flooding, it is necessary to use impermeable materials that act as dams or reservoirs. In those areas where retaining water generates a greater problem, it would be necessary to use resilience techniques so that the roads facilitate the natural flow of the water. It is also necessary to make a previous and rational plan when constructing roads in floodplains. First, it is necessary to understand the functioning of the natural dynamics and route of the flood water of a river in its floodplains in order to avoid constructing traditional roads perpendicularly to the flow of the water. In the case of not being able to ensure this, measures that enable the water to pass would be adopted with wide culverts to prevent obstruction.

With respect to the hydraulic infrastructures, urban development and frequent floods have led to the channelling of the majority of the river courses. Even some irrigation channels and azarbes are also channelled. When the swell of a river surpasses the sides of a channel, its walls are usually raised so that the next swell does not cause it to overflow. This action is negative on two levels. On the one hand, the channel or course is raised above the adjacent territories that are flooded when the river overflows. On the other hand, if the river overflows, the water cannot rejoin the course in the same way as when the courses are above their river valley. The flood waters cannot return to the course and have to reach the lowest point in order to join the course at another place. However, during this trajectory, populations and economic activities are affected and roads may be destroyed.

There are examples on a global level of flood situations when the channels themselves act as a barrier, preventing the water from returning to its course. Some examples are the River Mississippi in New Orleans (USA); the River Danube as it passes through Hungary and Austria, where the channels often block the return of the waters to the course when it overflows [42]; in sections of the River Ebro downstream of Zaragoza (Spain) [43]; and the River Segura in the mid-lower Segura basin (Spain) [44].

It is important to begin to contemplate all of these issues in order to adapt the territory to extreme meteorological and hydrological events and the effects of climate change. The actions carried out today will reduce future damage and costly investments that would be necessary in a few decades if action is not taken in the present [19].

The results obtained in this study show the efficiency of eliminating the barriers in order to minimise the effects of flooding, enabling us to (1) reduce the height of the flood (2) reduce the time that the flood water is trapped, (3) reduce the material damage and human casualties, (4) avoid closing roads for the emergency services, (5) help alleviate the water in other affected spaces, (6) help a quicker recovery and return to normality, and (7) enable the sediments of the River Segura and the overflowed river and azarbe waters to be deposited on the beach of Guardamar del Segura, among other positive effects

5. Conclusions

Anthropogenic actions, such as roads, railways, or hydraulic infrastructures built perpendicular to the natural flow of water, are a major obstacle that aggravates flooding in situations of river overflow. There are also other examples in the world of pedestrian bridges, roads, or railway lines that are installed in riverbeds that, on some specific occasions during flooding, have ended up being destroyed by the force and erosion of the waters.

All of this shows that floods can reach a catastrophic level due to the presence of this type of linear barrier against the natural flow of water.

This research has presented a particular case of this problem in the final stretch of the Segura River, especially in its flood plain, simulating the flood of September 2019, and seeing the effects caused by the national highway N-332 and the wall that separates the new and old channel of the Segura River.

This research has presented a particular case of this problem in the final stretch of the Segura River, especially in its flood plain, simulating the flood of September 2019, and seeing the effects caused by national highway N-332 and the wall that separates the new and old riverbed of the Segura River. The current situation has also been modelled with 2 openings and with 5 openings. It is observed in each of them that the result was improved, but there is still a significant flooding problem in the floodplain. To do this, it is proposed to completely eliminate the wall that separates both channels and raise the N-332 road using pillars that allow the passage of water in the event of a flood.

The results validate the effectiveness of these proposals by successfully evacuating the waters and reducing the time that the flood remains in the study area.

The results have been obtained using hydraulic modelling software (IBER), which allows us to understand the behaviour of the flood in space and time. It also allows the MDT to be modified, allowing proposals such as those raised in this article to be drawn up and their effectiveness to be verified.

The results obtained from the proposed measures (raising the N-332 and removing the wall that separates the new and old riverbed of the Segura River and eliminating the final breakwaters) have been successful in reducing the depths and the time that the water remains.

The findings of this research have broader implications for flood management practices, which can be extrapolated to other areas with similar problems. The use of hydraulic modelling programmes is key and essential to understanding the behaviour of water in the territory, knowing the affected areas, identifying conflict points, and verifying the effectiveness of proposals to improve the situation in the face of floods. In this way, greater detail and knowledge of flood behaviour is achieved, allowing for effective flood management.

This research supports the proposals mentioned above, but there may be the possibility of carrying out other types of action that also manage to reduce the negative effects of flooding.

In Spain, most rivers have problems associated with the presence of some type of barrier that can aggravate flooding, for example, in the entire river valley of the Segura River, the Vinalopó basin, the Júcar basin, and the Ebro basin, among others. Those areas where the river is located above its river valley are also prone to present this problem, so they are areas where future research could be carried out.

Even rivers with a “V” type river valley also present this type of problem, so different proposals could be analysed to solve this problem through hydraulic modelling anywhere in the world.

As mentioned above, there are areas where these types of actions have been carried out, and they have managed to reduce the effect of flooding successfully. Knowing these success stories and carrying out a follow-up or analysis of the before and after actions can be interesting to know the degree of protection obtained against flooding.