100085 Maria Carolina Rogelis, Senior Consultant on Flood Risk Assessment** Introduction Road networks are essential for economic, social, environmental, and security reasons. Road networks are therefore considered critical networks according to the consequences of their disruptions (Tacnet and Mermet 2012). Flooding poses an important threat to roads, and can lead to massive obstruction of traffic and damage to road structures, with possible long-term effects (Buren and Buma 2012). Flooding leads to significant repair costs for road control authorities, access difficulties for emergency services (Versini, Gaume, and Andrieu 2010a), and disruption for road users and the community at large. The consequences for businesses and the economy in general can be very significant (Brabhaharan, Wiles, and Frietag 2006). Because of the time and costs required for rebuilding, sustainable and long-term planning is crucial (Michael, Høegh, and Søren 2010); therefore, the consideration of flood risk constitutes an important input for decision making in planning this type of infrastructure. Flood risk analysis for road networks allows plans to be carried out in an appropriate manner, allocating resources for prevention, mitigation, and restoration (Balijepalli and Oppong 2014; Jenelius and Mattsson 2014). When road networks are disrupted by a hazardous event, the effects can be critical for emergency management. Transportation lifelines are generally considered the most important in an emergency because of their vital role in the restoration of all other lifelines (Cova and Conger 2004). Road network disruptions can threaten the ability to provide medical care and other critical services (Jenelius and Mattsson 2014). This report summarizes the main concepts and methodologies that are used to assess flood risk for road networks. The report presents references and examples, and is intended to be a starting point for practitioners in the field. * This document has been produced under the guidance and supervision of Fernando Ramirez-Cortés and Oscar A. Ishizawa, Senior Disaster Risk Management Specialists, as part of the Technical Notes developed under the World Bank LCR Probabilistic Risk Assessment Program (CAPRA). ** Technical review by Frederico Pedroso, Disaster Risk Management Specialist, and Giovanni Prieto, Flood Hazard Consultant. I – Interactions between Roads and Note: The figure shows five types of interactions involving water between a midslope road parallel to Floods contour and a stream (heavy solid line) Roads can be damaged by floods and also can The failure of a road network is of concern for enhance hazardous flood conditions. The road users (passengers and freight), support flooding of a road induces two levels of infrastructure (road, security equipment, consequences: on the one hand, people may be bridges, etc.), critical infrastructure supported injured and vehicles may be destroyed; on the on elements (such as bridges or tunnels) of the other hand, the disruption of traffic may have road network (power, telecommunications, severe indirect consequences. Road closures etc.), and the transport function (connectivity can have economic, social, and security and accessibility of points connected to each consequences (Tacnet and Mermet 2012). At other by the road) (Tacnet and Mermet 2012). the same time, roads and road development can Flooding causes traffic disruption, which have considerable effects on natural flood disappears as soon as the water subsides and patterns and effects. Roads fragment habitats goes back into its bed. In some cases, after a and interrupt the flow of water, sediments, flood, there is a layer of silt, such as mud or nutrients, and aquatic life, thereby impacting other coarse-grained material. In mountainous the beneficial effects of the natural flood cycle areas or where there is a sufficient bottom (Douven, Goichot, and Verheij 2009). slope, the water has enough energy to produce partial or complete destruction (Bil et al. 2014). Road development in floodplains alters the The types of flooding in a road network can floodplain hydraulics and affects the related be divided into three groups (Michael, Høegh, aquatic ecosystems (Douven, Goichot, and and Søren 2010): Verheij 2009). Figure 1 shows the interactions that can occur between road segments and  If there is insufficient capacity in the flows of water or sediment. Roads may act as drainage system, water on the surface corridors for flows of water on road surfaces (A collects in depressions in low-lying areas. and B in Figure 1) or in roadside ditches (C in The contributing drainage areas can be the figure 1). And roads may be sources of water surrounding areas as well as direct drainage for stream networks through culverts (D in on the road. figure 1) or gullies (E in figure 1). The  Rivers may flood because there is interaction between roads and streams may insufficient downstream capacity. modify the magnitude and direction of flows of  Rising sea level causes flooding of low- water and debris, and water flows may lying areas. transform into debris flows or vice versa (Jones Flooding in a road network can have the et al. 2000). following effects (Buren and Buma 2012):  Water that collects on the road because of the failure of flood defenses leads to traffic stagnation or, if the water reaches a certain depth, traffic stoppage. High water levels on the road or on the sides of the road construction can lead to loss of bearing capacity for the short and long term after flooding. Deep-lying sections, tunnels, as well as roads with a lightweight foundation Figure 1 Types of Water Flow on a Road Source: Jones et al. 2000. can be prone to uplift and heave.  Intense rainfall can increase pluvial favorable because they are longer or more time flooding and instability of the road consuming (Berdica 2002). These extra costs foundation. are part of the indirect costs. Other indirect  Excess groundwater levels can cause costs can be formulated as lost opportunities if uplift and heave of roads in excavation, planned trips are not carried out or other modes loss of bearing capacity, uplift of roads of transportation are chosen (Bil et al. 2014). with a lightweight foundation, and The inaccessibility of inundated roads during leaching of pollution. Possible effects of emergency management activities could cause excess hydraulic heads, in the aquifer indirect damage to the operability of strategic directly below the cover deposits, include structures, such as hospitals and fire stations uplift and heave of roads. (Albano et al. 2014).  The appearance of water on roads during The integration of road planning and design heavy rain can lead to road closures and and flood risk management plays a crucial role safety problems for vehicles. During in developing efficient and sustainable road heavy rain, the development of spray networks in floodplains. Figure.2 presents a behind vehicles results in poor visibility. framework for integrated analysis of road And in the worst case, water on the road planning and design. The figure shows the may cause vehicles to aquaplane. relation between road development design and planning (A) and the various effects (B), which The impacts can be divided into direct and are linked to the use of standards and guidelines indirect ones. Direct impacts include the costs (C). In road development and planning, all of reconstruction of damaged roads and the effects should be taken into account through reconstruction of landslide areas or adjustment the use of economically sound and of erosion entrained banks. Indirect impacts environmentally friendly guidelines for the entail the costs of interruption and logistics planning and construction of roads in disruptions. For example, when a portion of the floodplains (C) (Douven, Goichot, and Verheij road is closed, the detours are always less 2009). Figure 2 Conceptual Framework for Road Design and Planning in a Floodplain Source: Douven et al. 2009 3 Another relevant aspect in road planning and design is climate change. Taking climate The Federal Highway Administration change into consideration requires good flood developed a conceptual model for maps and good planning for water understanding the ramifications of climate management. In addition, consideration of change for transportation infrastructure. The climate change may have an impact on model consists of three interrelated steps. The standard design procedures, since the methods first step is to develop an inventory of assets for calculating and estimating the capacity of and prioritize them based on vulnerability, as drainage works may be insufficient (CEDR shown in the upper left of figure 3. The second 2012). A relevant concept is climate proofing, step is to combine climate data for a region to which involves identifying the risks to a understand the specific drivers of vulnerability, development project as a consequence of as shown in the upper right of figure 3. The climate variability and change, and ensuring third step in the conceptual framework, shown that those risks are reduced to acceptable levels in the center of figure 3, involves quantitative through environmentally sound, economically risk analysis to identify the most vulnerable viable, and socially acceptable changes (Lal transportation assets (VDOT et al. 2011). and Thurairajah 2011). ! Figure 3 Structure of the FHWA Conceptual Framework for Risk Assessment ! and Adaptation of Transportation Infrastructure to Climate Change Source: VDOT et al. 2011. ! ! Figure-1.!Structure!of!the!FHWA!conceptual!framework!for!risk!assessment!and! adaptation!of!transportation!infrastructure!to!climate!change! Description!of!Virginia!model!supporting!the!FHWA!framework! Suarez et al. (2005) provide an example that is called the importance of the element. Many includes climate change in road network flood other terms have been used in various fields for risk analysis. Their approach estimates the the same concept, including “criticality” impacts of flooding on a road network under (Taylor and Susiwalati 2012) and the influence of climate change. The approach “vulnerability” (Jenelius and Mattsson 2014). requires a model that is capable of simulating The main purpose behind the importance road traffic flows under a variety of conditions. measure is to compare and rank different The model is first run under normal elements. This allows, for example, the circumstances to provide baseline values for identification of the parts of the transport traffic volume and travel time. Then a set of system where disruptions would be particularly flooding scenarios is designed to identify those severe. Disruptions of such elements represent areas that are flooded, so that no trips begin or worst-case scenarios and the elements can also end there, and those network links that are be considered potential targets for antagonistic disrupted. The model is rerun and the results attacks on the system. Identifying important are compared with the initial run to determine elements means that targeted measures can be how many lost trips and how much extra travel taken to reduce the risk of disruptions in those time may be attributed to the weather event. locations. This type of analysis provides a basis for estimating the transportation-related costs of more frequent and more extreme weather “In transportation studies, the concept events under various climate change scenarios. of vulnerability is used to recognize To capture the effects of flooding on the performance of the transportation network, that susceptibility is not uniform different flooding scenarios are defined, based across people, vehicles, traffic flow, on combinations of the year of simulation infrastructure, or the environment” (2003 or 2025), area flooded (no flooding, 100- year floodplain, or 500-year floodplain), and type of flooding (coastal, riverine, or both). The combination of importance and disruption probability is called the criticality of the element. Importance can thus be expressed as conditional criticality (Jenelius and Mattsson II – Vulnerability 2014). Another important concept is resilience, which is defined by U.S. Presidential Policy In transportation studies, the concept of Directive 21 (PPD-21) as “the ability to prepare vulnerability is used to recognize that for and adapt to changing conditions, and susceptibility is not uniform across people, withstand and recover rapidly from vehicles, traffic flow, infrastructure, or the disruptions.” Therefore, analysis of the environment. Vulnerability can refer to the resilience of a road network should address the physical vulnerability of the transportation physical characteristics of the road network users or the potential for an incident to decrease and the activities it supports (World Bank the serviceability of the transportation system. 2015). Vulnerability in transportation can also be approached from the point of view of network The definition of vulnerability has not yet been reliability, as a reliable network is less generally accepted (Susilawati and Taylor vulnerable (Cova and Conger 2004) and 2008). Most concepts of vulnerability are based therefore more resilient when a disaster event on reduction in the performance of the road occurs. network. Some definitions of vulnerability include the following: The impact of the disruption of a given element 5  Road network vulnerability analysis can be and the exposure of persons, goods, defined as the study of potential infrastructure, and vehicles. This definition degradations of the road transport system of vulnerability relates to the consequences and their impacts on society through of natural phenomena, and can be modeling the road infrastructure as a decomposed into direct and indirect network with links (road segments) and vulnerability. Direct vulnerability nodes (intersections) (Jenelius and corresponds to physical damage directly Mattsson 2014). linked to the effects of phenomena such as  The expectance (E) of physical impacts physical injury of people or damage to (low, medium, or high1) to assets or infrastructure (caused by road rupture, networks, given different levels of debris flows, avalanches, deposition, exposure (World Bank 2015). rockfalls, etc.). Indirect vulnerability  “A susceptibility to incidents that can result corresponds to the remote consequences of in considerable reduction in road network an event, such as a flood, avalanche, or serviceability” (Berdica 2002). The link, debris flow (Tacnet and Mermet 2012). route, or road serviceability describes the possibility to use that link, route, or road In general, defining vulnerability allows during a given period of time. Furthermore, identifying structural weaknesses in the since accessibility depends on the quality network topology that render the network of the functioning of the transportation vulnerable to the consequences of failure or system, this concept has to do with degradation. Resources can then be targeted at different levels of vulnerability in reducing assessing these weak links (Taylor, Sekhar, and accessibility for various reasons. D'Este 2006).  Taylor, Sekhar, and D’Este (2006) define vulnerability as follows: The following subsections present the main 1. A network node is vulnerable if loss methodologies that are used for analysis of (or substantial degradation) of a few road network vulnerability. links significantly diminishes the accessibility of the node, as measured Multi-Criteria Analysis Techniques by a standard index of accessibility. Multi-criteria analysis establishes preferences 2. A network link is critical if loss (or between options. It makes a comparative substantial degradation) of the link assessment between alternatives or significantly diminishes the heterogeneous measures. For example, accessibility of the network or Benedetto and Chiavari (2010) present an particular nodes, as measured by a analytical model for road vulnerability standard index of accessibility. assessment based on multi-criteria analysis. This definition implies that road For each road element j, vulnerability Vj is vulnerability assesses the weakness of a determined by: road network to incidents and the adverse = ∑ ∙ =1 impacts for the community of degraded road network serviceability. where pis are vulnerability parameters and γis quantify the effect of each parameter on total  Vulnerability is a combination of the vulnerability (they represent the degrees of potential for damage; the associated costs; freedom in this model). The pi are hydraulic, 1For transport networks, levels of impacts are defined as open with minimum loss of road capacity, partially closed, and fully closed. geotechnical, structural, and functional The RIMAROCC method (Bles et al. 2010) parameters (sensitivity). A specific parameter uses several indicators in applying multi- set is defined for each typological element criteria analysis to risk assessment of roads. (adaptability). Each parameter can assume the The methodology conceptualizes the process of value 0, 1, or 2, depending whether the identification of vulnerabilities as looking for parameter implies low/none, medium, or high the vulnerable elements of the road system in vulnerability for the element. Quantitative and the event of the occurrence of an unwanted qualitative vulnerability parameters are (detrimental) event. The study of defined. Assignment of values is based on the vulnerabilities in the RIMAROCC values assumed by the entity considered for the methodology includes the following: (i) quantitative parameters, and on qualitative sensitivity and exposure of an asset (road, assessment categories for the qualitative right-of-way, equipment, maintenance parameters. Map 1 shows the results of the vehicles, etc.) to risk factors and/or an analysis applied to the road network in unwanted event; (ii) traffic; (iii) the age of the Northern Rome in the Tiber floodplain. infrastructure; (iv) design standards; (v) maintenance practice (routine and heavy repairs); (vi) the adaptability of the asset; and (vii) the possibility of upgrading without complete reconstruction of the asset. Analysis Map 1WSEAS of the Road TRANSACTIONS Network on ENVIRONMENT in the Tiber Floodplain in Northern and DEVELOPMENT Rome A. Benedetto, A. Chiavari a. Road vulnerability map b. Road risk map Fig. 9: Road Vulnerability Map Fig. 10: Road Risk Map Source: Benedetto and Chiavari 2010. requirements [10]. The evacuation plan studied this hydrometer. Within the 8 hours and half for residential (38760 inhabitants) and industrial ordinary routes can be used to evacuate and reach (6000 workers) areas in Monterotondo is one the recovery areas. good example of the usefulness of the model in 7 short term strategies planning. The plan has been Hydrometers Propagation time arranged, assuming as a baseline scenario the 200 Orte Scalo – Ponte Felice 2 h and 30 mins years period flood event. Levels of the risk on Ponte Felice – Stimigliano 2 h and 20 mins road network determined by the model shows that 4 h and 40 mins Table 1 shows the vulnerability indicators Serviceability Analysis proposed by the RIMAROCC methodology, which are subsequently used to assess risk in a The serviceability of a link is defined as the multi-criteria analysis framework. Estimation possibility of using that link during a given of the indicators requires collecting data, such period of time, which then relates to the as construction date, standards used, materials, possibility of the partial degradation of the equipment, etc., with the level of precision roads. Finally, if the consequences of a link depending on the scale of the analysis. The being affected are great, then the link is estimation also requires data on actual traffic considered critical to the network (Balijepalli and a comparison with expected traffic for and Oppong 2014). traffic counts, type, origin-destination analysis, Map 2 Vulnerability of a Road Network in etc., as well as data on maintenance (routine Luhacovice, Czech Republic and heavy repairs) and structural defects or existing damages that would likely be worsened by climate factors. The main infrastructure components to be investigated are major hydraulics, minor hydraulics and drainage, engineering structures, equipment, geotechnics, environment, and pavement. Table 1 Vulnerability Indicators Proposed by the RIMAROCC Methodology Source: Bles et al. 2010. Source: Bil et al. 2014. Bil et al. (2014) address vulnerability as the impact of interruption of a specific segment on the serviceability of the whole network (repair costs will be directly proportional to the length of an affected road and will differ according to the types of objects at the location of the interruption; repair costs will be highest in the case of repairs of bridges and tunnels). Map 2 shows the results of the analysis in the Czech Republic. The weakest segments are in the middle part of the territory. If those segments were interrupted, the length of the connection (detour route) between the end nodes would grow substantially. Road networks serve different demands or Another example of the use of the clients and the approaches that have been serviceability concept in the vulnerability described are just a snapshot of the possible analysis of roads is the use of the Network methods to be used when assessing Vulnerability Index, which takes into account vulnerability and criticality. the serviceability and importance of each road link in the network (Balijepalli and Oppong 2014). The serviceability of link i is calculated  Destination accessibility index. This index by dividing the total available capacity of the measures the ability of evacuees to access link by the standard hourly maximum flow rate destinations (such as assembly points or (that is, capacity) per lane for a given type of evacuation centers). If the failure or capacity road. The total available capacity of a link is degradation of a road section affects maximum obtained by summing the capacity of all the reduction in the accessibility index, that road is available operational lanes. identified as a critical location (Luathep et al. 2013). Accessibility Indexes Accessibility is defined by Susilawati and  Accessibility/Remoteness Index of Australia Taylor (2008) as the ease with which people (ARIA). ARIA is a remoteness index that can participate in activities from a specific measures the distance from populated localities location by use of a transport mode. With this to vital service centers. This index can also be definition, accessibility can be used to evaluate defined as the accessibility of a populated the performance of the transport system. Box 1 locality center to the various sizes of vital shows an example, taken form Susilawati and services, such as health, finance, and Taylor (2008), of the application of indexes to education. identify critical links. Several indexes measure  Generalized travel cost. Given the origin- accessibility: destination flows, the difference between the  Hansen index. The Hansen index considers least cost path with the network intact and the not only the generalized cost of travel, but least cost path without the link being evaluated also the attractiveness of the location, is estimated. Therefore, overall increases in which represents the size of activity, such cost in a degraded network can be assessed as the population, the number of theater (Taylor, Sekhar, and D'Este 2006). seats, the number of jobs, as well as the size of shopping centers. The Hansen integral  Network efficiency measure. This index accessibility index for a location can be corresponds to the average number of trips per written as: unit cost and represents the efficiency of the network by the traffic-to-cost ratio. The higher = ∑ ( ) is the traffic handled per unit cost, the more efficient the network is (Balijepalli and Oppong 2014). where Aj is the integral accessibility, B is the attractiveness of location (city) j, and C is the  Importance measure. The importance measure number of opportunities available at j. Often B assumes that all drivers are forced onto a more is taken as the population of city j, and f(Cij) is expensive route when an event causes the the impedance function, which represents the disruption or closure of a link or a group of separation between i and j. The impedance links. More expensive routes not only refer to function f(Cij) in the equation for the index can economics as a cost function in transport be the travel time and travel cost. Thus, the analysis, but also may imply mean travel time higher the impedance function, the lower the and travel distance. The behavior is described accessibility index at the particular area. by the user equilibrium principle, where the Taylor, Sekhar, and D'Este (2006) used the route choice is meant to minimize personal reciprocal of the distance between two cities travel cost. The basis for the measure is the (xij) as the impedance factor, implying that that change in the cost of travel (Balijepalli and for a higher cost of travel between the two Oppong 2014). cities, the accessibility between them is lower. 9 Box 1. Blue spots in The Netherlands  Network robustness index. The network Susilawati and Taylor (2008) studied road network vulnerability in the robustness index is defined as the change in Green Triangle Region in Australia by using two accessibility indices. The travel time cost associated with rerouting first is the Hansen indices, which measure the integral accessibility of all traffic in the system should that segment certain places, and the second is the Accessibility/Remoteness Index of Australia, which is developed by the Department of Health and Aged become unstable. The index is based on the Care (DHAC) in terms of measuring remoteness to access the service capacities of individual links and considers center. The methodology aims at finding out the vulnerability of a road network the rerouting options for the origin- at the regional level by measuring the changes of the Hansen destination pairs that use the link. The accessibility indices and ARIA index after one or more links have been index then uses travel time to measure the degraded. The basic methodology is shown in Figure 4. cost of rerouting traffic should a link be completely removed. The index assumes that the disruption will cause a complete closure of the link and that drivers follow user equilibrium in route choice. The system cost of travel for when all the links are intact is also calculated and the difference is the network robustness index (Balijepalli and Oppong 2014). III – Hazard Analysis Hazard can be defined as a dangerous phenomenon, substance, human activity, or Figure 4 Methodology for vulnerability assessment of a road condition that may cause loss of life, injury, or network (Susilawati and Taylor 2008) other health impacts, property damage, loss of livelihoods and services, social and economic Figure 5 shows the critical links identified in the study. disruption, or environmental damage (UNISDR 2009). In the context of road networks, the damage caused by a dangerous phenomenon—in this case, flooding—can include all the impacts addressed in the section on Interactions between Roads and Floods. In contrast, susceptibility describes the likelihood that a road section will be flooded, given the natural hazard. Susceptibility is the frequency individual link in a network. Detail and spatial of flooding of the considered road section over extent are correlated, but as computer storage a long period of time (Versini, Gaume, and continues to increase, this correlation is Andrieu 2010a). weakening, and soon there may be national (or larger) studies with very fine spatial and There are several important dimensions in temporal detail. The temporal extent and transportation hazard analysis, most notably resolution are also important. A central the spatial and temporal scales. The spatial question is the time horizon of the study, which scale includes the extent of the study and the can range from a single time period (cross resolution or detail. The spatial extent might be sectional) to any duration (longitudinal). Time global, national, regional, local, or an is also important because of the many cycles that affect the potential for hazards (Cova and This method is based on the idea that the higher Conger 2004). the runoff depth is in a sub-basin, the more hazard there will be on roads in that sub-basin. In the existing approaches for assessing flood The flood computations are performed on a hazard in road networks, hazard is assessed sub-basin level. Thus, the runoff depth of each with different levels of complexity; some of the sub-basin may be considered the most effectual approaches are limited to susceptibility. factor that affects the flood impact. Hence, the spatial analysis tools of the Arc GIS software Hazard analysis starts with a hazard are utilized to reclassify runoff depths in 10 identification and impact analysis. A process categories, and each category is assigned a source is considered an area that has a uniform unique number. That number, called the predisposition for hazard formation. For water hazard, or danger factor, is assigned to the road hazards, this area is the water channel and its in a particular sub-basin. By this approach, catchment area. The possible event magnitudes each road in the transportation network gets a are categorized with recurrence intervals for unique hazard factor (in each scenario), which specific years. The following subsections represents the flood hazard level. The danger discuss some approaches and examples to factors obtained by Dawod et al. (2014) are on illustrate the procedure. a scale from 1 to 10, with 10 being the highest hazard. The resulting map for Makkah, Saudi Modeling for Flood Hazard Assessment Arabia, is shown in map 4. To obtain the intensity associated with a return period, flood modeling is used. Map 3 depicts an example of modeling flooding across a Map 3 Floodplain Inundation Map over a Transportation Network transportation network. Map 3 shows the depth of the flood in meters, with the direction and velocity of the flood depicted with a vector field. This example is output from the MIKE 21 flood simulation system for modeling two-dimensional free surface flows. The system can model many conditions that occur in a floodplain, including flooding and drainage of the floodplain, embankment overtopping, flow through hydraulic structures, tidal forces, and storm surge (Cova and Conger 2004). Source: http://www.dhiaust.com/general/m21flood.htm, Figure Other, simpler approaches floodplain 6. A the include use of a inundation taken frommap and Conger flood Covadepicting 2004. depth (m) and velocity (m/s) ov Geographic Information System (GIS), such as network (Source: http://www.dhiaust.com/general/m21flood.htm). a transportation the method presented by Dawod et al. (2014). Note: m = flood depth; m/s = velocity. 11 Author'spersonal copy Map 4 Danger Factors for the Road Arab Network J Geosci in –1156 (2014) 7:1139  Calculate the probability and extent of Makkah, Saudi Arabia potential events.  Create the representation of results and deliverables. The maps that result from the analysis can show the classification of the hazard. As an example of intensity criteria, table 2 shows the intensity classification according to the FEDRO methodology. Table 2 Intensity Criteria for the Flooding Process Source: Dawod et al. 2014. esent and near future urban sub-basins) equals 31.6, 64.6, and 141.9 millionm3 FEDRO Methodology for 1990, 2010, and 2030, respectively. The augmentation of the total flood volume has been increased by almost Source: FEDRO 2009. Note: h = water depth or deposit thickness; v = water low- and moderate- In the 104 % from 1990 to methodology of2010, theandFederal by almost 120 Roads% from velocity; d = average thickness of the erosion. 2010 to 2030. reads out along the Office of Switzerland (FEDRO), the road has & A great part of the road network in Makkah City is ountainous areas forto be taken into account an impacts. asflood element that of Makkah's geology subjected to high dangerous That catego- In the FEDRO methodology, the spatial influences anite, that massively the ry constitutes process 211 roads, (FEDRO representing2009). As 37 % of the net- probability of occurrence is determined for work, with a total length of 481 km. Those roads of high mountainous areas.explained in the section on Interactions each process source, recurrence interval, and flood impacts exist in the holy shrines particularly in the between been identified in the Roads and Floods, the interaction field of the intensity map on a case-by-case 2 Arafat area. The future road network of Makkah City in 4.3 to 360.6 km andbetween roads and 2030 consists 705 roads is of streams determinant whose lengths range infrom basis. This process identifies the area or road ying from 16.50 to hazard conditions. 1.004 to 23.70 km, with a mean of 2.691 km. A major section length that is affected by a hazardous ns have been recog- part of the road network in Makkah City, in 2030, is al of 76 sub-basins, According to the methodology presented in event in comparison with the entire zone that subjected to high dangerous flood impacts. That catego- 90, have been identi- could potentially be affected by that hazard FEDROry(2009), hazard assessment for roads constitutes 563 roads, representing 74 % of the net- basinsextended over includes two work, withstages. Theoffirst a total length 1,398 stage, km. called process for a given scenario. stitute 17.5 % of the & There hazard is a significant is identification, increase in the flood comprised of the on impacts sinsin thestudy area. Makkah's road networks between 2010 and 2030. The Susceptibility analysis following xtended to cover 136 steps: overall length of flood danger-factor roads is increased 00.551 km2. In 2030, from 481 km (with almost 37 %) to 1,398 km (with of Makkah City will 2 Obtain, view, and analyze existing Versini, Gaume, and Andrieu (2010a) present 74 % approximately). areaof 780.388 km . a road susceptibility assessment methodology eased from 165.3 mm information & Urbanization sources. has a direct strong relationship with the based on the following three steps:  Analyze historical events. tal flood volume (for flood hazards. Urbanization leads to change in the land use of a region, changes its geological characteristics,  Carry out its and decreases geological, geomorphic, ability to absorb surface water which, 1. Identification of the set of all road hydrogeological, in turn, increases the and hydrological surface analysis water runoff. sections that could possibly be of the current state. exposed to flooding. In the proposed Based on the attained results, the following concluding  Formulate remarks can bethe basic scenarios (known as drawn: methodology, the points exposed to scenarios of hazard formation). flooding are of three different types: – It is recommended that the attained results should be  d flood impacts on roads Assess intowater the taken accountchannel by decisionand theinexisting makers implementing river crossings, low accumulation measures. new development planning of Makkah metropolitan area. points, and river adjacent points that can be submerged during river The second stage, called impact analysis, is overbank flow events. The comprised of the following steps: identification of all points of the road network belonging to one of these protecting the road network from flooding. The categories is based on analysis of GIS concept involves computer methods executed information. on office personal computers, followed by 2. Identification of the specificities of the targeted field inspections and actions. The road sections. starting point is a screening method that can be 3. Definition of the susceptibility rate. used at the regional scale to find blue spots. Depending on the severity of possible conflicts A key aspect in the methodology is the link between a blue spot and the road, the level of between the susceptibility of a road to flooding investigation can be expanded to analyze the and the dimensions of the river-crossing rain sensitivity of individual blue spots, or even structures (bridges or culverts) and, more an additional step of detailed numerical specifically, the adequacy of the opening modeling of hydraulic processes. The last (cross-section) of the structure and the procedures of the blue spots concept are discharges that may be produced by the inspections at selected local sites, followed by upstream watershed during floods (Versini, the appropriate actions. These actions may Gaume, and Andrieu 2010a). Versini, Gaume, include, for example, upgrading drainage and Andrieu (2010a) propose to compare two systems or improving the monitoring of water discharge values for their case study in France: levels in streams. The blue spots concept is the theoretical maximum free surface discharge intended for use on large and important roads capacity through the crossing structure (Qc), in a nonurban setting (Hansson, Hellman, and which can be estimated with the Manning- Larsen 2010). Strickler formula, and the theoretical 10-year The blue spots method is divided into three return period discharge (Q10) for the upstream levels, as follows (Hansson, Hellman, and watershed, based on a well-established formula Larsen 2010): adapted to small catchments in France. With these two variables, the ratio Q10/Qc is - Level 1. The first level can be described as estimated and used in the susceptibility a screening, where all depressions in the analysis. However, Versini, Gaume, and map material are identified. This is done by Andrieu (2010) found that the road altitude, allowing rain to fall on the model land local slope, and catchment area were the most surface while not allowing for infiltration important factors for identifying susceptible into the ground or evaporation to the road sections. atmosphere. Hence, every drop of rain will flow along the land surface until it reaches Blue Spots Approach a volume of free water collected in a depression. If these volumes are larger than A widely used approach is blue spots analysis. 10 cubic meters and close to a road, they Blue spots are flood-sensitive areas in the road are considered threats and are included in network (Michael, Høegh, and Søren 2010). A the following analysis. blue spot is a location of the road network that - Level 2. The second level is the calculation can be flooded in certain circumstances. A blue of rain sensitivity for each individual spot only refers to the probable cause of depression found in level 1. The calculation flooding and not to the consequences; is done by assuming no drainage from therefore, the identification of a blue spot does depressions and assuming impermeability not by definition mean that the risk of flooding of the catchment of 20, 40, 50, 60, 80, and in that location is unacceptable (Buren and 100 percent. In this way, a map can be Buma 2012). drawn showing the amount of precipitation The blue spots concept is a chain of procedures needed to fill low-lying areas. for systematically analyzing, adapting, and 13 - Level 3. The third level consists of a 2D- (for example, drainage and storage capacity) and 1D hydrodynamic model of surface when setting up emergency plans (Hansson, reservoirs and depressions, which is used to Hellman, and Larsen 2010). find pathways, catchments, and ponds in an area. The calculation of water flow on the Boxes 2 and 3 show examples of the blue spots surface and in the drainage systems is taken methodology applied in the Netherlands and Denmark, into account, giving a more accurate respectively. calculation of flood hazard. Box 2 Blue Spots in the Netherlands The methodology applied in the Netherlands for the identification of blue In addition, the flood hazard caused by sea spots is shown in figure B2.1. A significant amount of data is collected to level rise is mapped by incrementing the sea identify the blue spots. These data deal with the road, climate change, and the level and tracking how far inland the seawater existing modeling results. To anticipate future climate change, the analysis determines what climate change needs to be dealt with in the project. Climate reaches. Dikes act as barriers as long as the change is taken into account for the relevant worst-case scenario for various water level does not exceed the upper limit of types of climate change in 2050. For the analyses, existing knowledge and the dike. modeling results are used as much as possible. In the first phase of the analysis, this knowledge is combined with road information and climate The flood hazard from water level rise in rivers change scenarios to gain a first insight about potential blue spots. Based on can be calculated in the same way as for sea location and the height of the road, locations where water heights are higher level rise. The water level in a river can be than road heights are identified. Subsequently, information about the construction of the roads is used to identify other vulnerable spots, as well as incremented to a given level and the water level locations where water heights do not exceed road heights. rise can be tracked inland (Hansson, Hellman, The results of the first phase are based on existing approximate model and Larsen 2010). calculations, and sometimes assumptions and general information about the road. The calibration was performed by comparing the results of phase 1 with Normally, there will be many blue spots along the experience of road administrators by interviewing the road administrators or near a road stretch and level 2 analysis is in different districts. The calibration also included verification of potential blue spots that were identified in the first phase, as the identification of still probably justified in most, if not all, cases. The 1205568-000-GEO-0007, Version 2, 15 May 2012, final unidentified blue spots. level 2 analysis focuses on pointing out the The identified potential blue spots are not necessarily the actual vulnerable most dangerous depressions. spots in the road sections. For instance, there may be facilities that prevent Two depressions of similar geometry (volume flooding, or the design of the road may be very robust. The last phase of the analysis zooms in on the identified potential blue spots from the previous and shape) do not necessarily pose the same 3 Methodology steps to filter spots that are not vulnerable from the potential blue spots. A problem. It is crucial to determine the list of more likely vulnerable blue spots is the result of this last analysis. These catchment for every depression to estimate the more likely blue spots can later be analyzed to verify whether they are actual Following blue spots. are the general steps that are carried out for all types of flooding. In the specific volume of water available to fill the depression. chapters Figure 5 to 7, the applied methodology for the different types of flooding is elaborated. In B2.1 Methodology Applied in the Netherlands for the Identification A large catchment for a small depression Figure 3.1, one can see a graphical presentation of Blue Spots of this methodology. means a greater threat than a small catchment Existing models I for a large depression. Rainfall depth, in II Climate change millimeters needed to fill the depression, can be calculated by dividing the depression volume by the area of the catchment. III Intense rainfall III III Rise of water levels Land subsidence In conclusion, depressions near the road that Long periods of rain can be filled by relatively moderate rainfall should be targeted first for inspection and Locations vulnerable for III / IV / V preemptive measures (Hansson, Hellman, and (different types of) flooding Larsen 2010). The benefits of implementing level 3 analyses are I / IV that the water flows on the surface and in the Road characteristics Scope of project drainage systems are taken into account, thus providing a more accurate calculation. Level 3 is IV I Databases Rijkswaterstaat an excellent tool to use when looking for a Road administrators DISK / NIS / DTB solution, including more details about the systems Figure 3.1 Methodology I. Collecting data and existing models A lot of data are collected to identify the blue spots. These data deal with the road, climate change and the existing modeling results for the types of flooding A and B. II. Determination of climate change for the different analyses To anticipate for future climate change it is determined what climate change needs to be dealt with in the project. It is agreed with Rijkswaterstaat that climate change will be taken into consequences related to H. The consequences IV – Risk Analysis C are a product of the value of the elements at risk E, and their vulnerability V, such that the In the most general form, risk R can be defined risk equation becomes R = H × E × V. as R = H × C, where H is the probability of a Vulnerability V is a factor between 0 and 1, threatening event (hazard) and C are the indicating the severity of expected loss given a 15 hazard H, and expressed as a fraction of the The hazard and vulnerability data that are used total value of E. In the context of network in the equations shown in Table 3 can be vulnerability, monetary values of road obtained from the application of the segments (pavement, side rails, etc.) can be methodologies described in the previous included to refer to the structural vulnerability sections. An example of the use of these of the elements at risk. Hazard H may express mathematical approaches for the calculation of the probability of occurrence of a potentially flood risk on a road network is presented in damaging phenomenon within a given time Meyer et al. (2014). The method focuses on period and area (Meyer et al. 2014). expenditures on additional traffic loads resulting from road closures, and thus the Most risk calculation methods use so-called functional value of the network links. The static traffic values to assess the risk. The product of traffic volume [vehicles day−1], number of vehicles on a road section is defined excess distance [km], and closure time [days] by an average number of vehicles per time unit gives the total additional average traffic load (daily or annually, for example, annual daily per road closure [vehicles × km]. Assuming traffic) and by assuming that all vehicles travel that characteristic closure times amount to 1 at the same speed. Generally, two types of risk day, the multiplication of link-failure are calculated: (i) object risk, which is the likelihood [1 year−1] by additional traffic load probability that a driver is killed among the results in the annual flood-related link risk total number of persons passing through the [vehicles×kmyear−1]. hazardous area, and (ii) individual risk, which is the probability that a driver passing N times per day in a hazardous area is killed (Voumard et al. 2013). Risk equations for road networks are shown in table 3. Table 3 Risk Equations R is the risk [dead yr−1] or [USDyr−1] with n objects, H is the hazard [yr−1], Expi is the object exposure, that is, the probability that a vehicle is hit in the hazardous area [−], V is the object vulnerability [−] , and W is the potential total loss of persons or costs ([dead] or USD]). Object risk on a road where Rob is the object risk [dead yr−1], Fe is the frequency of occurrence of an event [yr−1], Ps is the proportion of the hazardous section that is affected when a hazard occurs [−], λ is the probability of death when a vehicle is damaged by a hazard [−], β is the average vehicle occupation [persons/vehicle], and Nv is the number of equivalent vehicles permanently exposed in the hazardous area: where Nv_tot is average number of vehicles per day [vehicles day−1], l is the length of the hazardous section [m], v is the average vehicle speed [km h−1], and f is a conversion factor to convert the speed from [km h−1] to mday−1. Fe and Ps represent H, where Ps allows the hazard on a road section to spread. Nv is the sum of exposures (Expi), λ is the vulnerability V, and β is the losses W. Individual risk, where X is the amount of time that a person passes every day through the hazardous road section [day−1]. Dynamic object risk with notation as in the above equations, where tcum is the accumulated time of vehicles observed in the hazardous area and tsim is the simulation time of a dynamic risk model. The use of GIS plays an important role in most methodology uses the envelope curves existing methodologies. For example, Albano developed by Teo et al. (2012), as shown in et al. (2014) propose a framework, integrated figure 7. The curves are shown in three color in a GIS, to estimate the direct and indirect zones (green, yellow, and red), and the damages from a flood event. The objective is to hydraulic stability for each idealized vehicle is understand the strengths and fragilities of a easily identified by color. The stable zone is particular urban area, including main roads, shown in green (on the left in the figure), the secondary and local roads, bridges, etc. The transition zone in yellow (in the center of the methodology proposed by Albano et al. (2014) figure), and the unstable zone in red (on the is shown in figure 6. Accessibility indexes are right in the figure). All vehicles in the red zone used (see the section on Vulnerability) in of the graph are dragged by the water flow; combination with a direct impact estimation to hence, for example, the vehicles could block an obtain an estimate of the maximum impact. emergency vehicle during rescue actions. Figure 6 Framework for Estimating Direct and Figure 7 Critical Threshold Values of Hydraulic Indirect Damages from a Flood Event Instability for Specific Vehicles t al.: Impact estimation and accessibility-operability GIS model for flood emergency management 2849 Source: Albano et al. 2014. Source: Teo et al. 2012. ses of the proposed methodology. As shown in figure 6, the following indexes are The methodology illustrated in figure 6 allows estimated: the analysis of emergency response. Road hase. Concurrently with the occurrence of phys- closures caused by floodwaters, ctional damage to urban areas, the operability urban areas, with the aim of prioritizing actions for flood- estimated on consequence reduction (Fig. 1). Sections 2.1 and 2.2 de-  The inverse reliability index highlights the their ic emergency structures, basis of velocity accessibility, and scribe water and depth the preliminary values, phases needed for the implementation the travel distance reliability of the within the city – or in general the urban area – is of could cause damages and hence could alterthe methodology. Section 2.3 summarizes the proposedpath. Travel distance reliability priority in emergency management. GIS methodology for the estimation of the consequences for emergency nt framework, integrated in a geographic infor- operations travel from an urban population, normal which can also be used to estimate theconsiders the probability that a trip conditions. m (GIS), aims to estimate the direct andAnalysis indirect of structural direct the pathsand economicof damages the for residential, com- between an origin–destination pair can flood event in order to understand the strengths mercial, and industrial buildings. Section 2.4 describes the emergency s of a particular urban area. The scope is to de-activities travel proposed could approachprovide to explore thethedependencies among the be completed successfully via the hy between the variouspossibility to estimate structures (hospitals, fire the operability structures and infrastructure ofof the a city during the emergency shortest distance possible for the n halls, schools, industries, etc.) and infrastruc- strategic emergency structures phase of a flood event (i.e. and highlight during or immediately after a normal functioning of system oads, secondary and local roads, bridges, etc.) flood), in terms of the accessibility of flood-prone areas and weaknesses identific ation of those (for the structures/infrastructure example, thenetworks operability of road most for emergency service. Fi-connections. tion and effectivenessinaccessible area at risk or a strategic are critical in emergency nally, in Sect. 2.5, this latter indirect-consequence estima- . The proposed model can aid in prioritizing connectivity roadbe tion is coupled with direct-consequence estimation through that a is most damaged).  The impedance index is the degree of s on flood mitigation strategies that should maximum-impact index. If the vehicles on s could support the maximization any of the street are dragged by the water benefit inaccessibility of an area that requires vestments by selecting the highest-priority ones 2.1 Data acquisition and harmonization rescue. flow, the road is inaccessible. The y service. In Sect. 2, the overall GIS framework n Sect. 3 the application and results of the pro- The level of epistemic uncertainty in estimating potential on a real flood event are described, and overall damage by the model depends on available data (data col- are provided in Sect. 4. lection, site visits, etc.). An analysis of the data considers 17 land use distribution, data population census, digital eleva- tion terrain models, and buildings and roads categorized on the basis of the function/typology (main roads, local roads, framework industries, resorts, hospitals, etc.). Therefore, the proposed approach requires the characterization of the system during  The hierarchy index is an estimate of The steps and sub-steps of the RIMAROCC the strategic importance of single methodology are shown in table 4. In the first arches. A network link is critical if loss step, objectives are defined as well as the or substantial degradation of the link external and internal parameters to be taken significantly diminishes the into account; the scope; and the risk criteria for accessibility of the network or the remaining steps. The second step entails the particular nodes. identification of sources of risk, areas of impact of unwanted events (including changes in  The inverse redundancy index suggests circumstances), and their causes and potential the number of potential alternative consequences. The third step, risk analysis, connections between one arch and involves developing an understanding of the others related to it considered in the risks. The risk analysis provides input to risk emergency phase. Therefore, the index evaluation and serves as a decision basis for provides information on the number of determining whether risks need to be treated, available and unavailable arches, in the and for selecting the most appropriate risk case of flooding, for emergency treatment strategies and methods. The fourth services if the arc is inoperable. step, risk evaluation, involves comparing the level of risk found during the analysis process The methodology combines these indexes to with risk criteria established when the context produce an influence index that takes into was considered. Based on this comparison, the account the role of each element in the system need for treatment can be considered. Risk in the emergency phase. Finally, estimation of mitigation is the fifth step, which involves the direct economic consequences is coupled identifying, appraising, and selecting one or with the indirect systemic impact in emergency more options for modifying unacceptable risks. management through a maximum-impact In the sixth step, the action plan is developed in index (for details of the analysis, see Albano et detail; responsibilities for implementation are al. (2014)). addressed, resources are allocated, and The use of multi-criteria analysis is proposed performance measures are selected. Since risk by the RIMAROCC methodology (Bles et al. management is a learning process, the seventh 2010). Figure 8 shows the framework for the step aims to monitor and review the methodology. The figure depicts a cyclical implemented actions and capitalize on the process that continuously improves knowledge gained from climatic events and the performance and capitalizes on experience. implementation of action plans. If conditions The process starts with an analysis of the change, re-planning starts within this step (Bles general context where risk criteria are et al. 2010). established, and ends with a reflective step where the experiences and results are documented and made available to the Figure 8 Framework of the RIMAROCC organization. Methodology. Source: Bles et al. 2010. Key steps Sub-steps 1.1 Establish a general context 1. Context analysis 1.2 Establish a specific context for a particular scale of analysis 1.3 Establish risk criteria and indicators adapted to each particular scale of analysis 2.1 Identify risk sources 2. Risk identification 2.2 Identify vulnerabilities 2.3 Identify possible consequences 3.1 Establish risk chronology and scenarios 3.2 Determine the impact of risk 3. Risk analysis 3.3 Evaluate occurrences 3.4 Provide a risk overview 4.1 Evaluate quantitative aspects with appropriate analysis (CBA or others) 4. Risk evaluation 4.2 Compare climate risk to other kinds of risk 4.3 Determine which risks are acceptable 5.1 Identify options 5. Risk mitigation 5.2 Appraise options 5.3 Negotiation with funding agencies 5.4 Formulate an action plan 6. Implementation of 6.1 Develop an action plan on each level of responsibility action plans 6.2 Implement adaptation action plans 7.1 Regular monitoring and review 7. Monitor, re-plan and capitalize 7.2 Re-plan in the event of new data or a delay in implementation 7.3 Capitalization on return of experience of both climatic events and progress of implementation Communication and gathering of information Table 4 Steps and Sub-Steps of the RIMAROCC drivers and trends that have an impact on Methodology. Source: Bles et al. 2010. the objectives of the road authorities; and relationships with and perceptions and A crucial step in the RIMAROCC values of external stakeholders. methodology is to identify the criteria, indicators, and risk evaluation categories to be  Internal context data. These are data on used. These aspects will be transformed into a governance, organizational structure, roles, risk matrix and used in a multi-criteria analysis. and accountabilities; policies and The criteria should correspond to the scope and objectives and the strategies that are in scale of the system under investigation. place to achieve them; capabilities, Since road networks are used by different understood as resources and knowledge players from an economic and social (for example, capital, time, people, perspective, the RIMAROCC methodology processes, systems, and technologies); specifies recommended data collection that relationships with and perceptions and should be adapted to the scale and objectives of values of internal stakeholders and the each specific study and that includes social, organization’s culture; information economic, and political data to establish the systems, information flows, and decision- context of the analysis. The collection of data making processes (formal and informal); and subsequent analysis require the standards, guidelines, and models adopted participation of several stakeholders, normally by the road authority; and the form and entailing intense inter-institutional work. The extent of contractual relationships. data collection includes, but is not limited to, the following (Bles et al. 2010): The internal and external contexts are highly important in assessing the consequences of  External context data. These are data on the events involving risk factors for the multi- social and cultural, political, legal, criteria analysis. The consequences are regulatory, financial, technological, and classified as direct (disruption of the road economic context, as well as the natural system, activities, and/or infrastructure) and and competitive environment context, at indirect (human and socioeconomic impacts). the international and national levels; key Indirect impacts relate to the consequences of 19 the climatic event on the well-being of the users 1. Enhancement of road segment resistance. (including psychological impact, stress, and Road segments endangered by a natural tiredness), safety (casualties), the local or disaster must be adjusted to enhance their regional economy (economic losses), etc. In physical resistance. For this purpose, some cases, specific studies are required to construction adjustments can be carried establish the social and economic costs to out. Examples include converting society. communications on embankments, creating deeper road foundations on slopes, and According to the RIMAROCC methodology, enhancing drainage structures. the impact of risk can be determined in the Vulnerability reduction is sometimes following categories (Bles et al. 2010): carried out together with the reduction of another threat. For example anti-flooding  Integrity of people (users and measures can also reinforce the original employees), that is, persons killed or road (Bil et al. 2014). injured  Damage to infrastructure, that is, cost of 2. Optimization of a road network. It is restoration preferable to ensure that detour routes can be used, in the case of a failure. Bil et al.  Operating losses for road managers (2014) provide an example of optimization (revenue, quality of service, image) and algorithms. users (loss of time, additional cost of using vehicles) 3. Maintenance. The road management staff  Damage to the environment (image and has the responsibility to organize degradation) inspections and maintenance in an  Economic and social consequences for appropriate way. Staff needs to work with the nation, region, or area of influence information retrieval, to evaluate whether (impact on modal choices, impact on the redesign of existing drainage systems is accessibility of local territories, and role necessary, and finally to determine a of transportation in the global economic specific action plan for inspections and system) maintenance. Information-related activities  Cost of palliative solutions should also be should focus on the following (Hansson, determined. Hellman, and Larsen 2010):  Hazard and vulnerability information  Background information at the site V – Risk Reduction  The current drainage system  Feasibility of monitoring (early Reducing risk for a specific segment of a road warning) systems network means influencing the occurrence,  Preservation of information in a frequency, or intensity of disasters, and database. reducing the vulnerability of a given segment. In the case of a decrease in the probability of 4. Monitoring. Local weather stations and, the occurrence of a natural disaster, monitoring for example, water level measurements in a is used for timely warning, flood control retention pond or culvert may provide modifications are made in beds of water flows, valuable information on how the system and appropriate planting of forest cover is responds to specific weather conditions, carried out, among other measures (Bil et al. like heavy rain. Changes in the response 2014). over time also provide good clues to when maintenance is needed. Monitoring Reducing the vulnerability of a road segment systems may include the following can include the following: (Hansson, Hellman, and Larsen 2010):  Local weather stations.  Risk assessment based on weather alerts  Water-level sensors in manholes, and conditions at the site wells, groundwater tubes, culverts,  In case of severe risk, presentation of retention ponds, streams, and reservoirs. information to road users about alternative Water-level readings can be directly routes, for example, by signs, radio, or converted to water flow if a discharge suitable information technology curve is determined for the measurement  Inspection and preparation of the site for site. harsh conditions  Video cameras.  Arrangement of warning signs and lights with adequate information. It is strongly recommended to create a database in which, for example, the following Versini, Gaume, and Andrieu (2010b) present information can be stored (Hansson, Hellman, an example of a flood early warning system for and Larsen 2010): a road network for which warning levels were defined from the distribution of flooding return  Inspection notes, check lists, photos periods and simulated discharges from a of the situation, and other comments distributed hydrometeorological model. about the site  Information about maintenance and repair work  Damage to the road and drainage Concluding Remarks  Flooding events Road networks are critical infrastructure.  Improper dimensions. Therefore, appropriate incorporation of risk information in the planning and operation 5. Flood early warning. The purpose of flood processes for road networks is crucial. Road early warning systems is to give the networks and flood hazards interact in such a responsible persons some time to consider way that roads cannot only be damaged by appropriate measures before the real floods, but also roads can enhance hazardous problems start. Anticipating the state of a conditions. Thus, it is highly important to road network during a flood can be helpful consider the interaction between roads and to prevent traffic from using roads at risk floods during the planning and design process, and identify the safest access routes to the generating integration between road planning affected areas for rescue services (Versini, and design and flood risk management. Gaume, and Andrieu 2010a). The cause of the problem in this case is typically a storm The vulnerability of road networks is an active with heavy rainfall, rapid and massive field of research, and several approaches spring snowmelt, or high river flow. Close currently exist to assess vulnerability. This cooperation with meteorological and report presented multi-criteria analysis hydrological institutes is required to techniques, serviceability analysis, and receive warnings about upcoming accessibility indexes. The choice of method of problems. analysis depends on the purpose, scale, and available data; however, the main purpose of An early-warning system may comprise the all the methods is to identify where and how following: disruptions of road networks may be particularly severe.  A weather alert notification system  Information and data retrieval from The approach to be used in flood hazard monitored sites analysis depends on the scale, type of flood, purpose of the analysis, and available data. 21 This report presented a non-exhaustive number of examples to illustrate some of the methods currently used for flood hazard analysis, with a “Road networks and flood hazards focus on road networks. Complexity can vary interact in such a way that roads cannot from highly detailed hydrologic and only be damaged by floods, but also hydrodynamic modeling to simpler susceptibility analysis techniques to identify roads can enhance hazardous conditions” potentially problematic areas. Hydrologic and hydrodynamic modeling offers the advantage of providing detailed information for planning proposed in the RIMAROCC methodology. and design purposes, with the disadvantage of Due to the high complexity of road networks being data demanding. Susceptibility analysis and the associated risks, risk analysis can be as provides the possibility to address larger areas complex as needed, incorporating aspects such with a limited amount of data. as damage to infrastructure, operating losses, Several approaches for risk assessment were damage to the environment, and economic and presented in this report. These include risk social impacts. equations, the use of GIS incorporating indexes, and the use of multi-criteria analysis Studies." Royal Institute of Technology, Stockholm. http://www.diva- portal.org/smash/get/diva2:9237/FULLTEXT References 01.pdf. Albano, R., A. Sole, J. Adamowski, and L. Mancusi. Bil, M., J. Sedonik, J. Kubecek, R. Vodak, and M. 2014. "A GIS-Based Model to Estimate Flood Bilova. 2014. Road Network Segments at Consequences and the Degree of Risk: Vulnerability Analysis and Natural Accessibility and Operability of Strategic Hazards Assessment. 1–18. Emergency Response Structures in Urban http://www.population- Areas." Natural Hazards and Earth Systems protection.eu/prilohy/casopis/eng/20/87.p Sciences 14: 2847-65. doi:10.5194/nhess-14- df. 2847-2014. http://www.nat-hazards-earth- syst-sci.net/14/2847/2014/. Bles, T., Y. Ennesser, J.-J. Fadeuilhe, S. Falemo, B. Lind, M. Mens, M. Ray, and F. Sandersen. Balijepalli, C., and O. Oppong. 2014. "Measuring 2010. Risk Management for Roads in a Vulnerability of Road Network Considering Changing Climate: A Guidebook to the the Extent of Serviceability of Critical Road RIMAROCC Method. Links in Urban Areas." Journal of Transportation Geography 39: 145–55. Brabhaharan, P., L. M. Wiles, and S. Frietag. 2006. doi:10.1016/j.jtrangeo.2014.06.025. Natural Hazard Road Risk Management Part http://eprints.whiterose.ac.uk/82511/1/An III: Performance Criteria. alysis of Vulnerability of Road http://nzta.govt.nz/resources/research/re Network_v2c_whiterose version.pdf. ports/296/docs/296.pdf. Benedetto, A., and A. Chiavari. 2010. "Flood Risk: Buren, R. Van, and J. T. Buma. 2012. Investigation A New Approach for Roads Vulnerability of the Blue Spots in the Netherlands National Assessment." WSEAS Transactions on Highway Network. Environment and Development 6. http://www.wseas.us/e- library/transactions/environment/2010/89- CEDR (Conference of European Road Directors). 839.pdf. 2012: Adaptation to Climate Change. Oslo. Berdica, K. 2002. "TraVIS for Roads: Examples of Cova, T., and S. Conger. 2004. "Transportation Road Transport Vulnerability Impact Hazards." Handbook of Transportation Engineering 1–50. Luathep, P., P. Suwanno, P. Taneerananon, and C. http://geog.utah.edu/~cova/cova-conger- Engineering. 2013. "Identification of Critical hte-2004.pdf. Locations in Road Networks Due to Disasters." 9. http://easts.info/on- Dawod, G. M., M. N. Mirza, K. A. Al-Ghamdi, and line/proceedings/vol9/PDF/P42.pdf. R. A. Elzahrany. 2014. "Projected Impacts of Land Use and Road Network Changes on Meyer, N. K., W. Schwanghart, O. Korup, and F. Increasing Flood Hazards Using a 4D GIS: A Nadim. 2014. "Roads at Risk: Traffic Detours Case Study in Makkah Metropolitan Area, from Debris Flows in Southern Norway." Saudi Arabia." Arabian Journal of Natural Hazards and Earth Systems Sciences Geosciences 7 1139–56. doi:10.1007/s12517- Discussion 2: 6623–51. doi:10.5194/nhessd-2- 013-0876-7. 6623-2014. http://www.nat-hazards-earth- syst-sci-discuss.net/2/6623/2014/. Douven, W.J.a.M, M. Goichot, and H.J. Verheij. 2009. "Best Practice Guidelines for the Michael, R. A., N. Høegh, and F. Søren. 2010. Integrated Planning and Design of "Development of a Screening Method to Economically Sound and Environmentally Assess Flood Risk on Danish National Roads Friendly Roads in the Mekong Floodplains of and Highway Systems [Développement d’une Cambodia and Viet Nam," synthesis report of méthode d’analyse pour évaluer le risque the “Road and Floods” project. Mekong River d’inondation sur les réseaux routiers et Commision, Off. Secratariate Phnom Penh, autoroutiers danois. 1–10. 35: 143. http://www.mrcmekong.org/assets/Publica Suarez, P., W. Anderson, V. Mahal, and T. R. tions/technical/Tech-No35-Roads-and- Lakshmanan. 2005. "Impacts of Flooding and Floods.pdf. Climate Change on Urban Transportation: A Systemwide Performance Assessment of the FEDRO (Federal Roads Office of Switzerland). Boston Metro Area. Transportation Research 2009: Natural Hazards on National Roads: Part D Transportation Environment 10: 231– Risk Concept. Berne. www.astra.admin.ch. 44. doi:10.1016/j.trd.2005.04.007. http://ced.berkeley.edu/faculty/ratt/readi Hansson, K., F. Hellman, and M. Larsen. 2010. The ngs/ClimateChnage_Sea_LevelRise- Blue Spot Concept. 1–34. references/Suarez_et_al_2005.pdf. Jenelius, E., and L. G. Mattsson. 2014. "Road Susilawati, and M. A. P. Taylor. 2008. "An Network Vulnerability Analysis: Accessibility Approach in Assessing Regional Conceptualization, Implementation and Road Network Vulnerability." 31st Application." Computers, Environment and Australasian Transportation Research Forum Urban Systems. 475–93. https://people.kth.se/~jenelius/JM_2013.p df. Tacnet, J.-M., and E. Mermet. 2012. "Analysis of Importance of Road Networks Exposed to Jones, J. A., F. J. Swanson, B. C. Wemple, and K. Natural Hazards." Proceedings of the AGILE U. Snyder. 2000. "Effects of Roads on 2012 International Conference on Hydrology, Geomorphology, and Disturbance Patches in Stream Networks." Conservation Geographic Information Science 24–27. Biology 14: 76–85. doi:10.1046/j.1523- http://www.agile- 1739.2000.99083.x. online.org/Conference_Paper/CDs/agile_20 http://myweb.facstaff.wwu.edu/wallin/env 12/proceedings/posters/Poster_Tacnet_Ana r435/jones_etal_2000.pdf. lysis_of_importance_of_road_networks_exp osed_to_natural_hazards_2012.pdf. Lal, P. N., and V. Thurairajah. 2011. Making Informed Adaptation Choices: A Case Study Taylor, M. a P., S. V. C. Sekhar, and G. M. D’Este. of Climate Proofing Road Infrastructure in 2006. "Application of Accessibility Based the Solomon Islands. Methods for Vulnerability Analysis of Strategic Road Networks." Networks and Spatial Economics 6: 267–91. 23 doi:10.1007/s11067-006-9284-9. Versini, P. -a., E. Gaume, and H. Andrieu. 2010a. http://lyle.smu.edu/emis/cmmi5/Ibarra/De "Assessment of the Susceptibility of Roads to skTop/PhD_Journals_Used/vulnerability_ap Flooding Based on Geographical Information: p_access_road_Taylor_2006.pdf. Test in a Flash Flood Prone Area (the Gard Region, France)." Natural Hazards and Earth Teo, F. Y., J. Xia, R. A. Falconer, and B. Lin. 2012. Systems Sciences 10: 793–803. "Experimental Studies on the Interaction doi:10.5194/nhess-10-793-2010. between Vehicles and Floodplain Flows." 37 – http://www.nat-hazards-earth-syst- 41. doi:10.1080/15715124.2012.674040. sci.net/10/793/2010/nhess-10-793- http://dx.doi.org/10.1080/15715124.2012.6 2010.pdf. 74040. 2010b. "Application of a Distributed Hydrological UNISDR (United Nations International Strategy for Model to the Design of a Road Inundation Disaster Reduction). 2009: Terminology on Warning System for Flash Flood Prone Areas." Disaster Risk Reduction. Geneva, Natural Hazards and Earth Systems Sciences Switzerland. 10: 805–17. doi:10.5194/nhess-10-805-2010. http://www.unisdr.org/files/7817_UNISDRT http://www.nat-hazards-earth-syst- erminologyEnglish.pdf. accessed May 8, sci.net/10/805/2010/nhess-10-805- 2014. 2010.pdf. VDOT, UVA, VCTIR, HRPDC, and HRTPO (Virginia Voumard, J., O. Caspar, M. H. Derron, and M. Department of Transportation, University of Jaboyedoff. 2013. "Dynamic Risk Simulation Virginia, Virginia Center for Transportation to Assess Natural Hazards Risk along Roads." Innovation and Research, Hampton Roads Natural Hazards and Earth Systems Sciences Planning District Commission, and Hampton 13: 2763–77. doi:10.5194/nhess-13-2763- Roads Transportation Planning 2013. http://www.nat-hazards-earth-syst- Organization). 2011. Assessing Vulnerability sci.net/13/2763/2013/nhess-13-2763- and Risk of Climate Change Effects on 2013.pdf. Transportation Infrastructure: Hampton Roads Virginia Pilot. Funded by: © 2015 The World Bank. This work is a product of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work.