Flood Exposure Assessment of Railway Infrastructure: A Case Study for Iowa

Abstract

Floods pose a substantial risk to human well-being. These risks encompass economic losses, infrastructural damage, disruption of daily life, and potential loss of life. This study presents a state-wide and county-level spatial exposure assessment of the Iowa railway network, emphasizing the resilience and reliability of essential services during such disasters. In the United States, the railway network is vital for the distribution of goods and services. This research specifically targets the railway network in Iowa, a state where the impact of flooding on railways has not been extensively studied. We employ comprehensive GIS analysis to assess the vulnerability of the railway network, bridges, rail crossings, and facilities under 100- and 500-year flood scenarios at the state level. Additionally, we conducted a detailed investigation into the most flood-affected counties, focusing on the susceptibility of railway bridges. Our state-wide analysis reveals that, in a 100-year flood scenario, up to 9% of railroads, 8% of rail crossings, 58% of bridges, and 6% of facilities are impacted. In a 500-year flood scenario, these figures increase to 16%, 14%, 61%, and 13%, respectively. Furthermore, our secondary analysis using flood depth maps indicates that approximately half of the railway bridges in the flood zones of the studied counties could become non-functional in both flood scenarios. These findings are crucial for developing effective disaster risk management plans and strategies, ensuring adequate preparedness for the impacts of flooding on railway infrastructure.

1. Introduction

Natural hazards, such as floods, significantly threaten various aspects of society. These events frequently lead to loss of life, substantial economic losses, infrastructure damage, and disruptions to daily life [1]. The impacts of these disasters are expected to intensify due to climate change and land-use changes [2]. Rentschler et al. [3] demonstrate that population growth, economic growth, and land subsidence are exacerbating global flood losses, with the average annual cost of flood damage expected to rise from USD 6 billion in 2005 to more than USD 60 billion by 2050. Understanding the potential ramifications of these natural hazards is crucial, and identifying regions at risk is imperative for implementing measures to mitigate public safety risks, as highlighted by Zhao & Liu [4].

Global policy frameworks have underscored this urgency. The United Nations’ Sustainable Development Goals (SDGs) call for the development of resilient infrastructure (SDG 9.1), while the Sendai Framework for Disaster Risk Reduction advocates strengthening the resilience of transportation systems to ensure their functionality during and after disasters [5,6]. In the United States, the Federal Railroad Administration [7] highlights flooding as a major climate-related hazard to rail safety and continuity. Given that pre-disaster investment in infrastructure resilience is more cost-effective than recovery, there is strong motivation to assess and mitigate flood risks before catastrophic events occur [8].

Decision makers regularly need to assess the history of the vulnerable sites to determine the level of risk and potential losses. They also conduct risk impact analyses to assess potential mitigation options. Furthermore, structural or non-structural mitigation options can be evaluated to secure welfare and minimize losses [9]. This proactive approach allows for the implementation of targeted protective measures, strategic location choices, and robust emergency response plans, ultimately safeguarding communities and minimizing the cascading disruptions that can result from infrastructure failures during flood events.

In order to ensure the resilience and dependability of critical services during disasters, it is crucial to understand the flood risk to critical infrastructure. Flood risk is composed of hazards, which constitute the probability and intensity of flood events; exposure, which is the presence of people or infrastructure in flood-prone areas; and vulnerability, which is the extent to which these components are susceptible to damage in the event of flooding [10]. The railroad network, a crucial component of delivering goods and services in the United States, is one of those infrastructures. The primary freight and passenger rail network in the US comprises about 225,308 km in 49 states and serves various sectors of the economy such as agriculture, industrial, wholesale, retail, and manufacturing [11,12] The role of the Iowa rail system, a major hub for railway traffic, is to connect Iowa shippers and buyers to markets in the North, in and out of the country [13]. Although freight transport comprises the majority, Amtrak, which carries passengers, also provides service in the state. Therefore, it is vital to assess the railroad network’s vulnerability during flooding to maintain services.

One of the pillars of a thriving economy is reliable transportation infrastructure, which provides access to social services, employment opportunities, and marketplaces [14,15]. As an example, railway infrastructure is essential for providing both freight and passenger transportation, which enhances social and economic welfare [16]. Moreover, railroads play a crucial role in guaranteeing that emergency responders are equipped with the essential resources and information to manage hazardous material incidents efficiently [17]. They collaborate by sharing resources and information and providing staff and equipment to assist in containing incidents, safeguarding public health and the environment, addressing any negative effects, and ultimately restoring safe operations as part of their comprehensive planning, risk reduction, and response endeavors [17]. However, there is a risk that the reliability of railroad networks can be considerably hampered by riverine flooding, which can also seriously disrupt the environment and cause structural damage [18]. Therefore, future urban risk management must address the rail transportation systems’ greater sensitivity to flood risk resulting from climate change, urbanization, land-use changes, and uncontrolled urban expansion [19].

Flooding events can occur for several reasons, such as heavy rainfall, rising sea levels, and snow or ice melt, and they can damage the structures of railway networks, including buildings, bridges, rails, and overhead lines. These events are often accompanied by river-related hazards such as erosion and scour, which further compromise structural integrity. Each of these potential damages entails a discussion of evaluation, maintenance, and repair costs. Thus, any significant interruption in the rail service always attracts attention from the public and puts further pressure on the political and administrative establishment [20]. Furthermore, since railroads are less spatially flexible than other modes of land transportation, they are more susceptible to flooding [21]. For these reasons, rail systems rank among the most crucial infrastructures to protect against major flooding [22].

Iowa is one of the most vulnerable states and has experienced several major flood events in the United States [23,24]. The state is also in the top ten federal grant assistance beneficiaries, which is primarily due to flood events [25]. Historically, floods in the Midwest during 1922, 1991 [26], 2008 [21,27], and 2019 [28] severely damaged railroads and disrupted train movement, and Iowa’s rail network had also been affected by these floods. Figure 1 shows a flooded railroad track near Plainfield, Iowa, during the June 2008 flood event, providing an example of flood impacts on rail infrastructure in the region [29]. According to Changnon [21], the 2008 flood caused the railroads in the Midwest substantial damage and costs as well. On the Iowa side, six rail bridges and three train wrecks were destroyed or badly damaged; additionally, 24 rail lines were closed in total.

Although the consequences of floods cause serious economic losses, there are a limited number of rail flood risk assessments in the literature using the methods and methodologies suggested in damage assessment studies during and after floods at spatial scales. Nonetheless, using historical data and a Geographic Information System (GIS) analysis, Hong et al. [30] suggested a thorough technique to statistically evaluate the railway system’s vulnerability to floods for the Chinese railway system. In [16], Kellermann et al. proposed the Railway Infrastructure Loss (RAIL) model for the railway transportation sector; they developed a flood damage model for the estimation of both structural damage to railway infrastructure and incurred direct economic losses. Kellermann’s methodology is based on data collected during and after past flood events [16,18]. This methodology was applied to the Austrian Northern Railway and Mur River catchment areas, respectively. Bubeck et al. [22] utilized the same method for the European railway network.

In addition, recent studies in railway flood risk assessment have introduced innovative methodologies to evaluate infrastructure exposure and vulnerability. Sresakoolchai et al. [31] developed an automated machine learning model to diagnose the flood resilience of railway switches and crossings, achieving over 98% accuracy in identifying deterioration due to flooding conditions. Abdelkarim et al. [32] and Varra et al. [33] used GIS-based multicriteria methods, such as the Analytic Hierarchy Process (AHP), to look at how likely it is for floods to happen along railway corridors by combining factors such as the terrain, water flow, and infrastructure. Liu et al. [34] and Samela et al. [35] utilized geomorphological and remote sensing methods to delineate flood-prone rail segments and classify them into hazard levels. Pant et al. [36] presented a vulnerability assessment framework for interdependent critical infrastructures, applying it to Great Britain’s rail network to analyze how flooding and other hazards can propagate through interconnected systems, affecting railway operations. Adams and Heidarzadeh [37] examined historical records of storm-related disruptions along the Dawlish coastal railway in the UK to provide a long-term perspective. They showed how repeated floods have had a cumulative effect. Varra et al. [38] used hydraulic modeling to determine the extent of damage floods caused to railway infrastructure. Despite previous research efforts, the comprehensive analysis of railway assessments during flooding remains limited and needs more attention, especially in light of the challenges posed by climate change [39].

Specifically, even though railroads in Iowa play an essential role in the state’s economy and the region’s ability to compete on the global stage [40], the state is lacking a thorough examination of the rail network’s exposure to flooding. Many existing studies on small- and large-scale flood disaster risk models have predominantly focused on damaged buildings, affected populations, or road networks [41]. In another study, the shortest path analysis was applied to Iowa to evaluate the disruption of transportation networks in terms of accessibility to critical facilities such as hospitals and schools in different flood scenarios of the road network. To the best of our knowledge, this will be the first Iowa-wide assessment of potential flood exposure to railway infrastructure under standardized scenarios.

This research, unlike other Iowa studies, aims to assess the impact of flood hazard on the railroad network in Iowa by performing state-level analyses of railroads, railroad bridges, intersections of railroads, and public roads, as well as more detailed county-level analyses of railway bridges in the three most flood-prone counties, using floods of 100- and 500-year return periods. These flood scenarios were selected because they represent standard benchmarks in regulatory assessments and infrastructure planning [42]. The present study focuses on identifying infrastructure that coincides with mapped flood extents and provides insights into spatial vulnerability hotspots. Our study complements this research by providing a statewide geospatial exposure baseline, which can guide future work toward integrated risk and resilience modeling for railway infrastructure. The study outcomes in terms of assessing flood risk for rail infrastructure can help develop policies and plans for effective disaster risk management. At the same time, the results obtained from the study can be used to identify which locations are risky and need prioritized rail transportation investments and resources to reduce the risk of flooding.

2. Materials and Methods

This research delved into examining how railroads are impacted by flooding at both the state and county levels. By considering the extent of the flood, we were able to assess a specific geographical area and pinpoint areas where rail transportation is vulnerable to flooding. Figure 2 illustrates the various elements of the analysis conducted to assess the railways’ vulnerability to flooding.

2.1. Data Collection

This study utilizes various datasets by acquiring publicly available data sources and requesting Iowa Department of Transportation (Iowa DOT). Information about the origin, content, and format of each data set used in this research is explained below.

Flood Inundation Map: After the devastating 2008 flood in the state of Iowa, the Iowa Flood Center (IFC) was established to develop advanced tools and provide resources for better flood management, including creating 2- and 3-D flood models for various return period scenarios (i.e., 2-, 5-, 10-, 25-, 50-, 100-, 200-, and 500-year periods) [43]. These floodplain maps were generated by analyzing the hydrologic and hydraulic characteristics of basins and streams, employing high-resolution input data such as a 1 m digital elevation model, and utilizing MIKE FLOOD and HEC-RAS software. The modeling process is described in detail by Gilles et al. [43], and the IFC has previously validated these maps using historical flood data, including the well-documented 2008 flood in Iowa. The development of these inundation maps was made through a collaborative partnership between the IFC and the Iowa Department of Natural Resources (Iowa DNR). This initiative produced interactive flood mapping products for all 99 counties in Iowa, with assistance from the Federal Emergency Management Agency, the U.S. Army Corps of Engineers, and the Iowa Natural Heritage Foundation. To ensure thorough coverage of potential flood extents, IFC researchers used high-resolution LiDAR data from the Iowa DNR to identify all streams draining catchments larger than one square mile. While this study did not perform a new calibration against field data, the pre-validated nature of the IFC maps ensures reliability at the regional scale. The flood inundation products cover the entire state at a 1 m spatial resolution, with flood extents produced for each return period scenario. The modeling incorporated hydrologic boundary conditions derived from stream gauge data and rainfall records, and all scenarios were benchmarked against observed flood extents. The maps remain the primary reference used by state and local agencies and continue to serve as a critical resource to help citizens, emergency managers, community leaders, and decision makers identify and communicate Iowa’s flood hazards and support informed floodplain management decisions. Access to the generated flood inundation maps is available through the Iowa Flood Information System [44] at https://ifis.iowafloodcenter.org. While comprehensive flood map models are critical for this study, data-driven approaches for flood map generation [45] can be used for resource constraint communities.

Railroad Network: The Iowa DOT data is the main source for the rail networks and relevant information. Data cover the last 10 years of railroad data includes current active rail lines, abandoned rail lines, and historic rail lines for the State of Iowa. However, at the analysis stage, only actively used railways were taken as a basis in order to be up-to-date and consistent. The total active rail length in Iowa is 11,927 km. According to the Iowa Rail Toolkit [40], a total of six railway class groups derived from the data can be briefly explained as follows:

Main: The primary railway route where uninterrupted train travel occurs between terminals and rail yards within the rail network.

Turnout: A configuration of tracks that allows a train to change its path from one track to another.

Crossover: A relatively brief section of track that redirects train traffic from one parallel line to another.

Siding: A brief additional track connected to the main track at both ends using turnouts, allowing trains to meet or pass. It runs parallel to the main route.

Spur (or stub in): A shorter, often dead-end portion of a track constructed to provide specific facilities like loading and unloading ramps access to the main or secondary line. It can also serve as temporary storage.

Yard: A network of supplementary tracks is used for tasks such as organizing railroad cars based on their cargo or destination, assembling trains, storing cars, or conducting equipment repairs.

Railroad Crossings: It presents the location of the intersection of railway and public roads, and its data type is a point geometry. The feature class contains rail crossing information, and it was generated using the Federal Railroad Administration’s (FRA) US DOT Crossing Inventory Form 6180.71 [46]. Nonetheless, the data include all railway crossings that can be at road level, under the road, or over the road. If the rail crossing is over the road, it will not be affected by the flood. For that reason, we eliminated this crossing class from the open attribute table based on FRA’s data entry field description report. Eventually, the number of crossings in Iowa that are road-level and under the road was found to be 5503.

Rail Bridge: Railroad bridges were collected from Homeland Infrastructure Foundation-Level Data [47]. These bridges are represented as point geometry and include the intersections of railroads with roads, waterways, and other railroads. There are 2652 rail bridges located statewide.

Rail Facilities: The Iowa DOT provided information on several rail facility types, namely warehouses, grain facilities, and transload facilities. A warehouse is a type of commercial building used for storing commodities, which can be anything from raw materials to packing supplies to spare parts to finished products used in manufacturing and production. The purpose of grain storage facilities is to store grain like warehouses, and they must have either a state or federal license. Lastly, a transload facility is where bulk freight shipments are transferred from a truck or container in one mode to another at a terminal interchange point. In total, Iowa has 14 warehouses, 324 state- and federally licensed grain facilities, and 46 transload facilities.

Digital Elevation Model (DEM): This dataset contains a digital elevation model (DEM) of the geographical surface of Iowa [48]. The DEM has a 3 m horizontal resolution and was created by combining 1 m resolution elevation data acquired by Iowa’s LiDAR program between 2007 and 2010. Despite its age, this dataset remains the most detailed and comprehensive statewide elevation model available, with a 3 m resolution and bare-earth filtering. It continues to serve as the foundational dataset for flood modeling efforts by the Iowa Flood Center and is widely regarded as reliable for topographic and hydrologic analysis across the state. Under predetermined standards, the Iowa Light Detection and Ranging (LiDAR) Project [49] gathers position and elevation (X, Y, and Z) data for the entire state of Iowa. The data are cleaned to remove any man-made structures and tree cover to display bare earth. In Iowa, the highest elevation point, Hawkeye Point, is about 510 m, whereas the lowest elevation point, the Mississippi River, is 146 m [50].

2.2. Case Study

The study was carried out on the rail network in Iowa, which consists of a total of 99 counties located in the Midwest region of the United States. It is the only state that is bordered by two major streams: the Missouri River to the west and the Mississippi River to the east. Furthermore, Iowa has a history of recurrent flood events that have caused significant damage to infrastructure, crops, and human life during the past two decades [41]. An overview of the railway network and mapped flood hazard extents is presented in Figure 3.

According to rail infrastructure information [46], there are 18 freight railroad companies that run 6155 km of track in Iowa. These railroads are divided into 4 classes: Class 1, regional, short line, and tourist railroads. Eighty-three percent of Iowa’s total route miles, including most of its grain collection network, are covered by Class 1. It also has a passenger rail line called Amtrak that runs east and west through the southern part of the state (see Figure 3). This line includes two different Amtrak routes and has a total of six stations. However, these stations do not pass through big cities such as Des Moines, Iowa City, and Cedar Rapids. In addition to them, from the rail line data provided by the Iowa DOT, we can access the building information of networks, which shows that the rail networks were built between 1855 and 2003.

2.3. Flood Impact Assessment on Iowa Railroads

This section details the GIS-methodology used to assess the impact of the 100- and 500-year floods on the railway system using the data we described above. The phrase “100-year flood” describes an occurrence that has a 1% annual chance of exceeding a given flood level, which means that there is a 1 in 100 chance that the flood level will be equal to or higher in any given year. Likewise, the probability of exceeding a “500-year flood” is 0.2% per year. Analyses were carried out at two spatial scales: a statewide overview and a focused county-level study using geospatial analytical software (Quantum Geographic Information System (QGIS 3.34) and ArcGIS Pro 3.1) for all spatial processing. To ensure an accurate overlay of datasets, all map layers (flood extents, rail infrastructure, elevation grids) were projected to a common coordinate system (e.g., NAD83/UTM) and aligned. We also pre-processed the input data to include only relevant infrastructure: for example, the rail network layer was filtered to active lines only (excluding abandoned tracks), and the rail crossing dataset was refined to remove grade-separated crossings where the railroad passes over a road (since an elevated track would not be affected by surface flooding at the road level [13]). The data have three different types of geometric structures and raster images. Among the geometric data, flood inundation maps, which illustrate the extent of the flooding, have polygon geometry, whereas the rail network data have a linear structure. Rail crossings, nodes, rail bridges, and rail facilities are presented as point geometry. Moreover, flood depth, LIDAR, and DEM data are stored as raster data, and our workflow integrates these data to identify intersections between floodwaters and rail infrastructure.

Graph theory has been beneficial in the analysis of transportation networks. In this theory, a graph (G) is represented as G = {N, E} and it consists of vertices (N), referred to as nodes or points, connected by edges (E), also described as links or lines. In this research, we looked at the path length exposure for flood probabilities of 1% and 0.2%. For all railways, we identify segments to be affected by overlaying flood maps on railway networks using the intersection tool in QGIS. It is assumed to be closed if the rail section partially or completely intersects the floodwater. In addition, considering the theory that bridges (vertices) connect the railway networks segments (edges) to each other, additional data at the county level are used to calculate the flood depths on a particular railway bridge and examine whether that bridge is inundated. In our analysis, a railway bridge is considered inaccessible and submerged if it is connected to fully flooded roads. The following subsections describe the methodology for statewide and county-based analyses in detail.

2.3.1. State-Wide Analysis

For the statewide exposure assessment, we overlaid the 100-year (1% annual chance) and 500-year (0.2% annual chance) flood inundation maps from the IFC with each category of rail infrastructure across the state. Using GIS spatial intersection tools, we identified where the flood polygons intersect with (a) the rail lines and (b) the rail infrastructure points, including crossings, bridges, and facilities. For point features, a point-in-polygon query was performed: any rail crossing, bridge, or facility point located within the flood extent polygon was flagged as inundated under that flood scenario. This operation was executed in ArcGIS Pro using the intersection function, ensuring that only points falling inside the modeled flood boundaries were counted. Points representing rail-over-road crossings had been excluded beforehand, so the remaining crossing points represent at-grade or underpass crossings that are susceptible to flooding. Each identified flooded point was attributed to the flood scenario (100 years or 500 years) and its location information for subsequent aggregation.

For linear features, a polygon–line intersection was carried out between the flood inundation areas and the statewide rail network layer (using QGIS and ArcGIS Pro geoprocessing tools). This produced new line segments corresponding to the portions of track that lie within the flood zones. We calculated the length of each inundated rail segment (in kilometers) using the GIS field calculator and then summed these lengths to obtain the total affected railroad length under each scenario. By comparing this flooded length to the total rail length in the state (≈11,927 km of active track), we determined the overall percentage of the Iowa rail network at risk of 100-year and 500-year floods. In addition to these statewide totals, we aggregated the results on a county-by-county basis to reveal the geographic distribution of flood impacts. This was accomplished by spatially joining the flooded segments and points to Iowa’s 99 county boundaries and computing summary statistics for each county. In particular, for each county, we derived (i) the total rail line length inundated (and what percentage that represents out of that county’s rail mileage) and (ii) the number of rail crossings, bridges, and facilities inundated. The point-in-polygon counts were generated using ArcGIS Pro’s counting tool (counting points within each county polygon), and the line inundation lengths per county were calculated by intersecting rail segments with county areas and summing their lengths. This systematic statewide analysis yields absolute exposures (e.g., kilometers of track flooded, number of assets affected) and relative exposures (e.g., percentage of a county’s rail infrastructure inundated), providing a comprehensive picture of flood risk to Iowa’s railroads under the two flood scenarios.

2.3.2. County-Based Analysis

In the second part of the assessment, we conducted a more granular analysis for the counties with the highest rail flood exposure in the statewide results (Pottawattamie, Harrison, and Linn counties). Here, we incorporated flood depth information and high-resolution elevation data to evaluate where flooding occurs and how severe it is in vertical terms. First, we examined the flood depth affecting rail lines in these counties. We extracted depth values along the inundated railway segments using the flood depth raster grids produced by the IFC for the 100- and 500-year floods. In practice, this meant overlaying the rail line geometry with the depth raster and sampling the water depth at points along each flooded segment. This raster-vector integration was completed using GIS tools such as the “Extract Values to Points” function, which assigns the underlying raster depth value to specified locations on the rail line. By doing so for all affected rail segments, we obtained an estimate of how deep the water would be over the tracks in each location. This additional information on flood severity (shallow vs. deep inundation) complements the binary exposure mapping and is important for understanding potential damage; for instance, even a few centimeters of water over the rails can halt train operations, while deeper inundations could cause structural damage to the track bed. The flood depth profiles along the rail lines were noted for each of the three counties and for both flood scenarios.

The second component of the county-based analysis focused on railway bridges, using a three-dimensional approach adapted from [51]. Our goal was to determine how many rail bridges in the selected counties would likely be overtopped or made impassable due to flooding. We started with the complete statewide set of 2652 railroad bridge points (from the HIFLD dataset) and filtered it to isolate bridges that span waterways, since those are directly exposed to riverine flood flows. This was achieved by using the bridge metadata and location: the HIFLD data classify bridge points according to what they intersect (road, water, or another railroad), so we selected only those points corresponding to waterway crossings (i.e., where a rail line crosses a river or stream). Next, we addressed a data alignment issue to improve accuracy. We found that some bridge point locations did not exactly coincide with the rail lines on our map—a minor offset likely due to differences in data sources or resolution. To correct this, we utilized high-resolution satellite imagery (Google Satellite view within ArcGIS Pro) as a basemap and manually adjusted the position of these bridge points, snapping them to the centerline of the rail track at the correct crossing location. This realignment step ensures that each bridge’s coordinates truly reflect the physical bridge’s position on the rail network, which is crucial when extracting elevation data for the bridge.

After preparing the bridge locations, we combined the elevation and flood depth datasets to evaluate bridge inundation. For each bridge point, we retrieved the ground elevation from the 3 m Digital Elevation Model (DEM) (recall that this DEM is derived from LiDAR and represents bare-earth terrain). We then obtained the flood depth value at that location from the flood depth raster for the given scenario. We determined the water surface elevation at the bridge site by adding the flood depth to the ground elevation. To compare against this, we needed an estimate of the bridge deck elevation. The LiDAR data provided returns from the bridge structure, allowing us to approximate the bridge deck height. We then applied the inundation criterion: a railway bridge is classified as flooded if the water surface elevation at that location meets or exceeds the elevation of the bridge deck. In other words, if the flood depth is high enough to submerge the bridge, we consider it inundated and non-functional. Moreover, even if a bridge deck sits slightly above the floodwaters, it would still be considered effectively impassable in our analysis if the connecting rail segments on both sides are fully flooded—under such circumstances, the bridge becomes an isolated island with no usable track leading to it. Using this rule, we evaluated each waterway-spanning bridge in the three focus counties for the 100-year and 500-year floods, tallying how many bridges would be overtopped or isolated by floodwaters. The combination of statewide and county-specific analyses strengthens the credibility of the findings, as we capture both the broad scope of exposure and the fine-grained details of how flooding can affect critical rail components such as bridges.

3. Results and Discussion

This section presents the results of the flood impact assessment derived from a comprehensive case study and discusses these results. In Section 3.1, state-wide analysis results utilizing flood inundation data are given and discussed, while Section 3.2 covers the result of the analysis of the most impacted three counties primarily using flood depth data.

3.1. State-Wide Analysis Results

Table 1. Summary of rail elements exposed to flooding in Iowa State.

Flood Scenario	Railroad Length (km)	# of Railroad Crossings	# of Rail Bridges	# of Rail Facilities
Baseline	11,927	5503	2652	384
100-year flood	1040	437	1551	23
500-year flood	1952	793	1622	51

summarizes the rail components that were flooded in the state of Iowa during the 100- and 500-year flood scenarios. The results indicate that, out of the 11,927 km of railroads across the state, approximately 1040 km (8.7%) intersect the 100-year floodplain, increasing to 1952 km (16.4%) under the 500-year flood scenario. Similarly, 437 railroad crossings (7.9%) are affected in the 100-year scenario, compared to 793 (14.4%) in the 500-year case. Freight and passenger facilities also show increased exposure, rising from 23 (6.0%) in the 100-year to 51 (13.3%) in the 500-year flood scenario. Among all components, rail bridges exhibit the highest share of exposure. Of the 2652 bridges statewide, 1551 (58.5%) intersect with the 100-year floodplain, and 1622 (61.2%) with the 500-year floodplain. These percentages reflect spatial overlap with flood extents and do not necessarily imply operational failure. Nevertheless, the results indicate an elevated risk of service disruption, especially in areas with limited redundancies or key junctions. According to the county-based flood depth analysis discussed later, the number of fully inundated bridges is lower. Still, even partial submergence may hinder accessibility and functionality, especially in areas where alternate routes or redundancies are limited. These results provide a thorough picture of the exposure of statewide infrastructure and can be used to determine priority areas for more in-depth, site-specific flood impact assessments and mitigation strategies.

Figure 4 shows the percentage of impacted railway length per county during the 100- and 500-year floods. The railroad length was evaluated instead of the number of railroad segments in order to analyze the direct effect of flooding on the railroad. Considering the length of the railway that will be affected is an essential factor in estimating reconstruction or repair costs. Briefly, we calculated the inundated railroad length within each county and extracted the county damage percentage (Figure 4). Most counties have damage levels of up to 5%. Moreover, under the 100-year flood event, Fremont County, located in the lower-left corner of the Iowa boundary, would be experiencing major inundated railways exceeding 50% of its railway length, while, under the 500-year flood, four more counties (Pottawatomie, Harrison, Linn, and Allamakee) are impacted by that percentage. In general, we concluded that some counties, such as Pottawattamie, Mills, Plymouth, and Harrison, which are located along the Missouri River, as well as every other county along the Mississippi River except Louisa, will suffer significant rail damage (5–50%). On the other hand, there are a few counties that have zero percent damage. This means that either the railway network does not pass through those counties, or the railways do not overlap with the flood extent in both scenarios.

Although Fremont County has a high flood risk in both flood scenarios, looking at only percentages, Pottawattamie County ranks first when we consider the length of railroads that will be affected by flooding.

Table 2. Top 15 counties based on impacted railroad length (km) sorted by 500-year floods.

County	Crossover		Main		Siding		Spur		Turnout		Yard		Total
	100	500	100	500	100	500	100	500	100	500	100	500	100	500	All
Pottawattamie	0.6	0.6	68.7	108.0	9.3	10.4	28.2	33.7	1.4	1.5	101.4	144.5	209.6	298.7	445.0
Linn	-	0.8	24.3	56.3	1.1	2.3	12.1	23.4	0.6	1.4	34.1	83.3	72.5	167.5	335.6
Harrison	-	-	63.4	98.3	0.9	1.9	-	-	1.9	5.2	5.9	10.8	72.4	116.6	231.7
Scott	-	-	23.1	42.6	2.4	8.2	1.6	2.2	-	-	25.5	30.2	52.6	83.1	177.3
Lee	-	-	16.9	35.6	-	-	0.5	1.2	-	-	29.8	39.6	47.6	76.9	196.7
Polk	-	-	28.8	40.6	1.7	2.8	10.3	12.8	0.7	1.1	13.8	13.8	55.6	71.4	309.3
Plymouth	-	-	39.4	64.5	2.7	3.7	-	-	-	-	-	-	42.2	68.8	166.8
Allamakee	-	-	10.5	62.1	-	2.8	-	-	-	-	-	-	10.5	65.3	71.8
Clinton	-	-	20.2	41.5	1.9	4.0	-	3.3	-	-	2.0	14.0	24.3	62.8	238.0
Dubuque	-	-	9.1	34.2	0.6	1.3	-	7.4	-	-	-	19.0	10.0	62.0	145.2
Woodbury	-	-	1.5	16.7	-	0.7	-	13.2	-	1.0	-	28.8	1.8	60.8	193.0
Mills	-	-	40.2	45.1	3.0	3.1	-	-	4.1	4.1	3.3	3.3	50.9	55.9	122.4
Fremont	-	-	37.8	38.1	8.2	8.2	3.6	3.7	1.2	1.2	-	-	51.0	51.3	57.1
Clayton	-	-	7.9	35.8	1.3	3.0	0.8	0.8	0.8	0.8	0.5	5.0	11.3	45.4	108.2
Muscatine	-	-	9.8	21.8	3.1	4.9	2.6	8.0	-	-	-	4.4	15.5	39.1	156.1

summarizes the information for the top 15 counties based on the length of the railway that will be affected by the flood with different railroad-type classifications. At the same time, this table gives the total length of the railroad network for each of these impacted counties in the last column. The range of total impacted length during the 100- and 500-year flood events for the studied counties fluctuated from 10 to 209 km and 39 to 299 km, respectively. During a 500-year flood event, the total impacted length increased dramatically, often doubling or even exceeding the impacted length observed during 100-year flood events, as seen in areas such as Linn, Clinton, and Dubuque. Among the top 15 counties affected, the main class is the railway type that every county will be inundated with, with a range of 8–108 km. In contrast, the crossover class comes out as less vulnerable to flooding.

Additionally, we estimated the total length affected and the percentage of affected segments across the state by rail type (

Table 3. Impacted railroad length (km) for different railroad types sorted by the total length.

Railroad Types	100-Year	Percentage	500-Year	Percentage	Total Length
Main	658.9	4.97%	1289.0	9.73%	13,247.5
Yard	245.2	19.35%	431.2	34.04%	1266.9
Siding	50.4	6.57%	82.4	10.75%	766.7
Spur	72.8	10.90%	127.7	19.12%	667.8
Turnout	11.7	14.64%	19.3	24.14%	79.9
Crossover	1.3	9.90%	2.4	18.24%	13.2

). Overall, the state of Iowa is threatened with losses of up to approximately 8% and 16%, respectively, in 100- and 500-year flood events, and the class most affected during these two floods is the “main” class in terms of road length, while the “yard” class has the highest share when we consider their percentages.

Based on Table 2 and Table 3, as well as the description of railway types given above, it can be determined that two railway tracks, which are essential in terms of being main and complementary tracks, will be interrupted in the event of a flood. In particular, the class of yard that has the function of repair and maintenance should normally be a railway line that would be least strongly affected in an emergency situation. Therefore, when we examine these results, it can be concluded that there is a deficiency in this regard.

In Figure 5, we show the total inundated length over the entire state while considering the year of construction. Relying on the rail-built-year information provided by the Iowa DOT, railroads that will be inundated during 100- and 500-year flood scenarios are relatively older-built railroads. In Iowa, the first railroad was built in 1855, and the railroad lengths that are expected to be most affected in both 100 and 500-year flood scenarios were built between 1865 and 1874, with estimated lengths of 285 km and 557 km, respectively. This could potentially be attributed to the oversight of not factoring in flood scenarios during the construction of these older railways. We also noticed that the length of the railway impacted by a 500-year flood is almost twice the length impacted by a 100-year flood for each construction period. On the other hand, the railroad lengths that are expected to be the least affected in both scenarios are the recent railways made in 1900 and 2000. The findings indicate that a higher proportion of recently constructed railway segments fall outside the mapped flood extents.

The railway system is linked to land transportation. Disruption in one place on the railway will also affect the public-use road connected to that road. Figure 6 demonstrates the number of impacted crossings between the railroad and public road per county. According to our results, while Polk and Pottawatomie counties have the highest crossings inundated due to the 100-year flood, Linn and Woodbury counties will share the same range between 50 and 128 crossings with others in the 500-year flood scenario. This suggests that a substantial share of railway crossings in many counties is not located within the 100- and 500-year flood hazard zones.

A railway bridge is a specially constructed structure to carry the rail traffic flow across an obstacle, including rivers [52]. However, flooding can cut off access to bridges and create challenges to freight and passenger transport. Figure 7 gives information about the number of railway bridges within the 100- and 500-year flood zones. A total of 87 counties in Iowa have railroad bridges within the two flood extents. It has been found that the highest numbers of 66 (100-year) and 71 (500-year) railway bridges within the floodwater were in Plymouth County, and some western and eastern counties have a major share. Linn County, where there is the large city of Cedar Rapids, also has a high number of railway bridges in flood zones.

However, it should be noted that these results were created using only 100- and 500-year flood maps and actual flood events may differ in extent, duration, or severity. Therefore, the results indicate potential inundation rather than guaranteed impacts. Nonetheless, our analysis can provide a comprehensive examination of rail bridges, enabling the identification of counties that may be susceptible to the inundation of rail bridges. Additional data, such as flood depth, are needed for bridge closure information, and the analysis of this subject is explained in more detail in the next part.

When we compare the Iowa DOT’s railway facility locations with the 100- and 500-year flood maps, it appears that some of the storage facilities would be affected by flood scenarios, and

Table 4. Rail facilities within floodplain.

	Warehouse		Federal Grain		State Grain		Transload Facility
	100 yr	500 yr	100 yr	500 yr	100 yr	500 yr	100 yr	500 yr
Inundated	0 (-%)	4 (29%)	13 (6%)	24 (12%)	5 (4%)	11 (9%)	5 (11%)	12 (26%)
Total	14		206		118		46

summarizes that analysis. When we evaluate the percentage of facilities that will be inundated in general, it is shown that there will not be a major loss. Almost every flood probability scenario is below 30%; even in the 100-year scenario, it has been revealed that no warehouse will be affected. Yet, when we evaluate the capacity and location information of the facilities that will be affected by the flood, the losses are shown to be high and this may seriously affect other cities, perhaps even other states. For instance, among the grain facilities, the city that will see the most damage to storage capacity is Hamburg (Fremont County), with 10,238 bushels. The second most damaged city is Council Bluffs, Pottawattamie County, with a total storage capacity of 8210 bushels, and Davenport (Scott County) will be ranked third, with 4795 bushels of storage capacity.

3.2. County-Based Analysis Results

In this section, based on statewide analyses, the top three most vulnerable counties, namely, Harrison, Linn, and Pottawattamie County, were selected to carry out a detailed analysis using the flood depth maps as additional data. This is because the standard approach for estimating the cost of flood damage usually relies on flood depth as the primary measure of the extent of damage severity [53,54,55,56,57].

The classification of rail lengths according to flood depth originated from Bubeck et al. [22], in a study that used the RAIL method to analyze the railway network in Europe. Using these methodologies, flood damage classification begins with classification of the first damage class when floodwater levels reach up to 20 cm along the track portion. In the standard framework, the second damage class pertains to water heights ranging from 21 to 140 cm, while the third damage class is designated for water levels over 140 cm.

In

Table 5. Affected railway length (km) for the top three counties within the floodplain for different damage classes (cm).

	100-yr Flood Scenario			500-yr Flood Scenario
County	<20 cm	20–140 cm	>140 cm	<20 cm	20–140 cm	>140 cm
Pottawattamie	66.5	135.0	8.1	6.9	175.2	118.2
Linn	18.7	47.3	6.5	22.2	108.2	37.1
Harrison	16.3	53.5	2.5	12.7	86.8	16.0

, we assessed the various damage classes and their potential impact on railway lengths. Our analysis indicates that the second damage class predominates in both considered scenarios, representing the most significant share compared to other classes. Notably, while the low damage class ranks as the second most impactful in the 100-year scenario, it falls to third in the 500-year scenario. A closer examination of the second and third damage classes, which are more prone to severe impacts, reveals that Pottawatomie County exhibits the highest rate of rail length among the three counties analyzed. Consequently, given its leading position in terms of rail length, repair costs in Pottawattamie are anticipated to be substantial. These findings provide a realistic basis for estimating repair costs associated with damage. While quantifying the risk of damage is essential, identifying the precise locations of the highest risk remains a challenge, and this aspect could not be meaningfully addressed in our study.

According to the Iowa DOT [40], some railway companies commonly adopt a practice where they place loaded train carriages on specific bridges when there is high water. This increases the bridge’s stability by adding extra-dense materials such as rocks, ballast, scrap metal, or other heavy and non-reactive substances. In order to identify the location of these bridges, it is important to know which bridges will be inundated, and which will be open.

Table 6. Railway bridges in the top three impacted counties.

		# of Bridges in Floodplain		# of Inundated Bridges
County Name	# of Bridges	100 yr	500 yr	100 yr	500 yr
Pottawattamie	97	59 (61%)	64 (66%)	29 (30%)	37 (38%)
Linn	72	18 (25%)	36 (50%)	0 (0%)	12 (17%)
Harrison	82	56 (68%)	63 (77%)	27 (33%)	39 (48%)

shows how many railway bridges would be in floodplains and inundated based on 100- and 500-year scenarios for the top three impacted counties. Even though almost more than half of the bridges seem to be in the floodplain in both scenarios, the rate of bridges being closed is less than 50% according to the method used in this research. Nonetheless, it should be noted that about half of the bridges in the flood zone will become non-functional. It has been noted that the railway bridges in Pottawattamie and Harrison counties will be affected more than those in Linn County. In particular, in Pottawattamie and Harrison, it was concluded that 30–50% of the rail bridges would be inundated in both flood scenarios.

3.3. Discussion and Limitations

This study shows that Iowa’s rail infrastructure is significantly exposed to flooding. Roughly 9% of rail lines in the state cross the 100-year floodplain; in the 500-year scenario, that number rises to 16%. Almost 58% of rail bridges cross the 100-year floodplain, and 61% do so in the 500-year scenario; they are at particular risk of flooding. This exposure concentration is in line with previous research that emphasizes the susceptibility of rail systems to flooding [21,22], which shows that rail lines are aligned along river corridors and flood-prone lowlands. Historical events such as the 2008 Midwest floods confirm the predictive value of this analysis. The findings underscore the importance of protecting key assets, as also emphasized in global and national assessments [7,14]. With projected increases in extreme rainfall [2] and warming trends, flood risks to railways will likely intensify [22].

County-level patterns further highlight spatial disparities in risk. With up to half of their bridges at risk, Pottawattamie and Harrison, located along the Missouri River, are the most severely impacted. At the same time, Linn County, with the Cedar River running through it, appears less exposed but is still affected. In all three, most inundated rail lines fall in the moderate-depth category (21–140 cm), which can halt operations and damage infrastructure [57]. Notably, Pottawattamie shows the greatest length of track in this depth range, suggesting high repair costs. These regional exposures are consistent with observations from rail systems in other countries, such as China [30], where specific high-risk areas were recommended for targeted reinforcements. This study provides essential insights for mitigation planning at both the state and local levels by identifying particular segments that are susceptible to flooding and warrant prompt attention.

The sharp contrast between the impacts of the 100- and 500-year scenarios (e.g., rail line exposure nearly doubling from 9% to 16%) indicates that current design standards may underestimate risk. Upgrading rail assets should involve design thresholds beyond historical 1% annual chance floods. Strategies may include elevating tracks, improving drainage, and reinforcing bridge foundations. These actions support international calls to increase the resilience of transportation systems [5,8]. Operational resilience is crucial, even beyond design. Although the main and yard lines are the most visible, supply chains depend on them as freight corridors. Wide-ranging economic repercussions could result from disruptions [36]. According to the Association of American Railroads, even short-term shutdowns could have a significant financial impact [11]. Real-time monitoring systems [58], coordinated emergency response, and contingency routing should all be included in mitigation planning. State officials and private rail operators need to collaborate across agencies [17].

This study faces challenges and uncertainties, and, consequently, there is a demand for future research. This research is based on publicly available data, which we cannot rely on to be up to date since many studies cannot afford commercial data sources. Exposure was assessed using static flood maps, which do not capture flow velocity, flood duration, or rise time. How water moves during a flood can make a big difference in terms of the damage it causes. Fast-flowing water can wash away soil and weaken foundations, while standing water can slowly wear down buildings over time. To develop a clearer picture of these effects, future models should rely on both real-world evidence and the physical behavior of water [38]. Moreover, our method treats all intersecting assets as impacted, relying on generalized depth-damage thresholds [22,57], without site-specific vulnerability. Factors including the age, design, and maintenance condition of each bridge or track segment can significantly affect whether it withstands a flood. Advanced tools such as the RAIL model [16,18] could refine these estimates. Moreover, this study does not simulate network effects. While asset-level exposure is shown, cascading delays and detour costs are not modeled. Graph theory and transport simulations could capture such systemic impacts (e.g., input–output models, freight delay cost estimations) [35,36]. Reliable data on flood-induced damage to rail infrastructure and subsequent recovery timelines remain scarce in the United States. A major challenge is the limited availability of empirical data on rail damage and recovery following flood events in the U.S., which hinders both model validation and economic analysis. A national initiative to collect and organize data on flood events, including financial losses and service disruptions, could provide essential support for future planning and more effective risk management [53,56]. Although this study is based on currently available flood extent data, it should be considered an initial scoping assessment. Future research should consider flood scenarios that reflect evolving climate conditions, given the documented rise in the frequency and intensity of extreme weather events. Although the present analysis is preliminary, it provides a useful foundation for setting priorities and enhancing preparedness efforts. By highlighting areas and assets with the greatest exposure, this work supports the development of a more informed and resilient infrastructure planning framework.

4. Conclusions

This research aims to analyze the railway network, railway bridges, rail crossings, and facilities by using 100 (1% chance) and 500 (0.2% chance)-year flood return periods at the state and county levels utilizing spatial analytical software (i.e., GIS). Additionally, the three most affected counties (Pottawattamie, Linn, and Harrison) were selected to carry out more detailed investigations using flood depth maps. The results of the statewide analysis demonstrate that, during 100-year flooding, the percentage of affected rail length, rail crossings, bridges, and facilities could be 9%, 8%, 58%, and 6%, respectively. Under the 500-year flood scenario, these results could reach up to 16%, 14%, 61%, and 13%, respectively. The analysis revealed that key components of the rail system, particularly main and yard lines and bridges, face considerable flood exposure. In addition, in the second analysis stage using the depth maps, we observed that, in both scenarios, the affected railways in the top three counties would suffer mostly second-class damage, and about half of the railway bridges in the flood region would become submerged.

Despite the limitations discussed earlier, such as the use of static flood maps and the lack of detailed damage data, this study provides valuable statewide insights. Future research should build on this baseline to include dynamic flood modeling, economic loss estimation, and asset-level risk assessments. The flood exposure model developed in this study offers useful insights for disaster risk reduction and infrastructure resilience planning. It has the potential to assist specific stakeholders in pinpointing regions of vulnerability and guaranteeing a fair allocation of investments and resources for flood mitigation initiatives within the railroad infrastructure industry. According to our results, improvement projects for rail segments and bridges in exposure areas might be a priority for public agencies (e.g., Iowa DOT) and private organizations. This baseline exposure analysis can be expanded through structure-level risk assessments, enabling more targeted resilience planning. Future studies could build on this foundation by estimating damage probabilities and operational disruptions at individual assets.

In light of our findings, we propose several targeted strategies to mitigate flood risks to railway infrastructure. Constructing flood barriers and enhancing drainage systems in the most vulnerable counties should be prioritized to address both 100-year and 500-year flood scenarios. For critical main and yard railway lines, reinforcing structural integrity through elevated tracks and the use of stronger materials can significantly reduce potential damage. Implementing real-time flood monitoring systems using advanced remote sensing technology will provide timely warnings and facilitate rapid response measures. Additionally, collaboration between public entities, such as the Iowa DOT and private organizations, is crucial for funding and executing these initiatives. A systematic approach to collecting and analyzing damage and repair data will refine our models, allowing for more precise risk assessment and improved strategic planning. These recommendations aim to enhance the resilience of railway infrastructure and ensure a balanced allocation of resources for effective flood mitigation.