Guidance to Improve Pedestrian and Bicyclist Safety at Intersections (2020)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

35 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS Chapter 3: Analyze Intersection Safety and Identify Issues This chapter focuses on the third step of the process: helping practitioners understand and identify safety issues at intersections. The intent of this section is not to detail safety analysis methods, which are covered in many other manuals (see Table 1 on page 9 in the introduction). Instead, this chapter aims to broadly describe approaches for identifying pedestrian and bicyclist safety issues at intersections to help readers select an approach most suited to their situation and readily locate resources relevant to that approach. 3.1 Approaches to Intersection Safety Assessment There are many approaches to safety analysis, but two are most prominent: hot spot and systemic safety approaches. These approaches are described below. Hotspot Approach (Reactive Approach) Traditionally, many jurisdictions have identified high- priority safety-improvement locations by mapping crashes and identifying locations with large clusters of crashes, referred to as “hot spots” or “black spots” (see Figure 22). A hot spot analysis approach uses information about past crashes to identify locations that seem to have a high risk for future crashes. Hot spot mapping consists simply of mapping locations by their number of crashes, giving it the advantage of being simple to perform and easy to communicate. It is only necessary to assemble crash data for a jurisdiction over a certain time period (e.g., five years) and to plot each crash to the location where it occurred. The resulting list of high-crash locations can be represented as a prioritized list or on a map using symbology, such as larger icons, to indicate high-crash locations, or through spatial tools such as heat maps (Figure 22). Many transportation agencies find that results from a hot spot analysis are easy to communicate to staff, policymakers, and the public. When using this approach, it is important Identify Treatment Options for Creating Safer Intersections Countermeasure Options Final Countermeasure Selection Evaluate Priorities and Assess Trade-Offs and Viability 4566 Frame the Process Identify and Collect the Data for Analysis 1 2Chapter Chapter Chapter Chapter Chapter Chapter Analyze Intersection Safety and Identify Issues 3Chapter

36 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS to plot pedestrian or bicyclist crashes separately from motor vehicle crashes to determine where there are potential safety issues for nonmotorized road users, because a map of total crashes may not identify locations where pedestrian or bicyclist safety improvements are needed. Additionally, in areas with large enough crash numbers, it can be helpful to map certain crash types (e.g., left turns, as in Figure 10 in Chapter 2) to understand specific trends or location characteristics associated with those crash types (see New York City’s Left Turn Pedestrian and Bicyclist Crash Study 2016). However, the simplicity of a hot spot analysis is also its main drawback. Because it looks only at past crashes, it is reactive, not proactive. It may seem counterintuitive, but specific locations where crashes have occurred will not necessarily experience a crash again any time soon, as crashes tend to vary around over time (a phenomenon called “regression to the mean”) (Carter et al. 2012, p.22). This effect is particularly true for less-common crashes like pedestrian and bicyclist crashes, which are heavily influenced by whether someone happens to be walking or bicycling at a certain time in a certain place, or where road user distraction or impairment may be involved. The focus on locations where crashes have occurred—in contrast to a more macro- level focus on common characteristics of those locations—means that hot spot analysis does not easily identify high-risk locations for less-frequent crash types. In addition, the lack of past crashes at an intersection does not necessarily mean that it is safe, particularly for pedestrians and bicyclists. Another limitation of hot spot analysis is that it lacks consideration of exposure, an important predictor of crashes. For example, if there are many crashes at an intersection, but also many pedestrians and bicyclists, the intersection design may actually be safer than others, encouraging more bicycling and walking (increasing exposure) and thus increasing expected and actual crashes. On the other hand, an intersection may have few crashes because there is little walking and cycling activity through it, or just because crashes are somewhat random events and pedestrian and bicycle crashes are rare, such that no crashes happened to occur during the study period. This may correspond to a condition where pedestrians and bicyclists avoid a location because it is seen as too difficult or unsafe to cross the street. These reasons underscore the importance of including at least three years, and preferably five years, of crash data in pedestrian and bicycle hot spot mapping, and of contextualizing that analysis with information about exposure. Additionally, it may be helpful to map public comments or incidents such as near-misses for comparison to reported crash data. Analysis methods presented in the HSM, which use hot spot maps, do account for regression-to- the-mean bias, but may not sufficiently account for pedestrian and bicyclist exposure (see Section 3.3). Systemic Approach (Proactive Approach) Another method for identifying high-risk locations is a systemic analysis. For pedestrian and bicyclist crashes, the systemic analysis methods are more likely to successfully identify safety issues. A systemic analysis identifies high-risk locations by examining the prevalence and combination of site characteristics that have been identified as high- risk indicators in prior research. In more advanced systemic safety analyses with sufficient data, sites can be identified which have a high risk of pedestrian or bicyclist crashes. The goal of systemic analysis is to identify and address these high-risk locations before crashes occur. Source: FHWA Applying Safety Data and Analysis to Performance- Based Transportation Planning, November 2015 Figure 22. Example of hot spot analysis map.

37 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS NCHRP Research Report 893 provides a guidebook for a systemic approach to pedestrian safety, defining a systemic approach as “a data‐driven, network‐wide (or system‐level) approach to identifying and treating high‐risk roadway features correlated with specific or severe crash types. Systemic approaches seek not only to address locations with prior crash occurrence, but also those locations with similar roadway or environmental crash risk characteristics” (Thomas et al. 2018). Figure 23 shows a schematic of the process detailed in NCHRP Research Report 893 Table 11 lists factors from the literature that have been found to influence pedestrian safety. In addition to these factors, Table 3 in NCHRP Research Report 893 lists the following factors of potential interest for pedestrian crash risk at intersections: • Number of traffic lanes • Number of intersection legs • Crosswalk length • Traffic control type • Commercial driveways • Unrestricted/restricted turn phasing (e.g., no vehicular left turns allowed during the pedestrian crossing phase) • Turning lanes • Speed limit • Crosswalk markings and type (high visibility or standard) • Sidewalk coverage • Curb ramps Note that these lists may not be comprehensive, as pedestrian safety research continues to evolve. Table 11. Intersection Factors That influence Pedestrian Safety According to Prior Research Factor Relationship with Crashes Motor vehicle traffic volume + Turning motor vehicle volumes + Roadway functional class + Proportion of truck/bus traffic in traffic stream + Proportion of local streets at intersection – Pedestrian volume* + / – Number of legs > 3 + Total number of lanes + No median/median island + Presence/number of transit stops + Presence of on-street parking + Presence/number of driveways + Presence of signal + (crash frequencies) – (crash severity) Separating turning movements from walk phase (pedestrian scramble phase, or walk and restricted turn phase) – Vehicle speed + with severity Source: Adapted from NCHRP Research Report 893, Table 7 (Thomas et al. 2018). Note: Positive (+) and negative (–) denote correlations with crashes. * Pedestrian volumes have been found to be positively associated with pedestrian crashes up to an extent; however, research such as Thomas et al. 2017a found a breakpoint after which pedestrian volumes were negatively associated with crashes. Figure 23. Steps in a systemic pedestrian safety analysis process. Source: NCHRP Research Report 893 (Thomas et al. 2018) Step 1 Define Study Scope Step 2 Compile Data Step 3 Determine Risk Factors Step 4 Identify Potential Treatment Sites Step 5 Select Potential Countermeasures Step 6 Refine and Implement Treatment Plan Step 7 Evaluate Project and Program Impacts Systematic Pedestrian Safety Analysis Process

38 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS Unfortunately, there is currently no similar systemic safety guide for bicycle safety. However, Table 12 lists factors identified in prior research in North America that have been found to affect bicyclist safety. NCHRP Project 17-84, “Pedestrian and Bicycle Safety Performance Functions for the Highway Safety Manual,” is also yielding useful tables of factors that affect walking and bicycling safety (Torbic et al. 2018). Table 12. Intersection Factors That Influence Bicyclist Safety at Intersections According to Prior Research in North America Factor Relationship with Crashes Reference Motor vehicle traffic volume + Nordback et al. 2014; Wang et al. 2017 Right-turn motor vehicle traffic volume + Strauss et al. 2015 Left-turn motor vehicle traffic flow + Strauss et al. 2015 Bicycle traffic volume* + Nordback et al. 2014; Strauss et al. 2015; Thomas et al. 2017b; Wang et al. 2017; Harris et al. 2013 Crosswalk width + Strauss et al. 2015 Commercial properties + Thomas et al. 2017b Roadway functional class + Thomas et al. 2017b; Strauss et al. 2015; Harris et al. 2013 Bicycling in the opposite direction of motor vehicle traffic + Harris et al. 2013 Speed limit + Chen and Shen 2016 Proportion of local streets at intersection – Thomas et al. 2017b Pedestrian volume + Thomas et al. 2017b Number of legs >3 + Thomas et al. 2017b; Strauss et al. 2015 Total number of motor vehicle lanes on largest approach leg (>4 lanes) + Thomas et al. 2017b Presence/number of transit stops + Thomas et al. 2017b; Strauss et al. 2015 Presence of on-street parking + Thomas et al. 2017b Raised median – Strauss et al. 2015 Presence of signal + Thomas et al. 2017b Two-way center turn lane + Thomas et al. 2017b Trail crossing + Wang et al. 2017 Land use entropy Wang et al. 2017 Street slope + Thomas et al. 2017b Note: Positive (+) and negative (–) denote correlations with crashes. *Bicyclist volumes have been found to be positively associated with bicyclist crashes, but that relationship is nonlinear, with fewer additional crashes expected per additional bicyclist at higher bicycle volumes.

39 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS 3.2 Using Crash Typing to Identify Treatment Options If pedestrian- and bicycle-specific crash typing is available, countermeasures can be selected specific to the crash types observed, using resources such as those listed below. For example, potential countermeasures for the “motorist left turn–parallel paths” crash type include: • Signs and signal modifications—pedestrian signal timing, leading pedestrian interval, left-turn phasing • Intersection design changes—smaller curb radius, curb extension, raised pedestrian crossing, roadway lighting • Traffic control device additions—medians, rectangular rapid flash beacons • Other measures—pedestrian and driver education, police enforcement These potential countermeasures are a starting point in the process of figuring out how to address safety issues. However, the context of the crash problem is highly important in determining how to implement these changes. For example, if the intersection handles large amounts of vehicle traffic at moderately high speeds, then a raised pedestrian crossing is likely inappropriate, since this countermeasure is intended for lower-speed locations. Instead, countermeasures such as implementing a leading pedestrian interval would be more suitable for this situation. Chapters 5 and 6 provide additional detail about the importance of considering context when selecting countermeasures. Chapter 4 contains tables of recommended countermeasures for prevalent pedestrian and bicyclist intersection crash types, with a complementary Countermeasure Glossary provided in the Appendix. The glossary entries provide additional information about contextual considerations for each countermeasure, as well as complementary and alternative countermeasures for consideration. This information can be used in conjunction with other countermeasure resources, such as the FHWA’s PEDSAFE, BIKESAFE, and CMF Clearinghouse, and NHTSA’s Countermeasures that Work, to determine the most appropriate and effective countermeasures for identified safety problems. Not all countermeasures listed are currently MUTCD compliant. 3.3 Safety Analysis Tools Safety practitioners can also use advanced tools and resources for considering the safety of design features. The AASHTO HSM is a comprehensive resource for data-driven safety analysis methods (AASHTO 2010). It presents a process to consider crash, traffic volume, and roadway data to identify locations that are high priority for safety improvements. The procedures allow a practitioner to diagnose crash problems and develop solutions. A significant part of the HSM presents a series of models which can be used to predict the number of crashes that would be expected to occur at an intersection given certain intersection characteristics. However, the models provided in the HSM 1st Edition are currently limited. For example, while the HSM includes prediction models for pedestrian crashes at four-leg signalized intersections, the HSM method for predicting bicycle crashes amounts to simply calculating them as a proportion of the predicted motor vehicle crashes, with the factor determined by the intersection type. Consequently, this method should be used with caution, as there are factors (listed in Table 12) that may increase or decrease the expected number of bicycle crashes that are not considered by the HSM method. Future updates to the

40 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS HSM are expected to improve methods for estimating the number of bicycle and pedestrian crashes. CMFs are another resource for exploring the safety effect of intersection characteristics. The HSM 1st Edition contains several pedestrian- and bicycle-related CMFs, but the most comprehensive and current repository is the CMF Clearinghouse, available at www.cmfclearinghouse.org (FHWA 2019). Although the CMF knowledge related to pedestrian and bicycle safety is not as comprehensive as that for motor vehicle safety, there are a growing number of CMFs related to the effect of intersection features on pedestrian and bicycle crashes. 3.4 Engineering Studies In addition to analyzing the pedestrian or bicycle crash type and other crash features, deciding which countermeasure is best suited for a particular location requires finding out more about the site’s characteristics. Determining the underlying reason(s) for the crashes, or the potential for crashes, is essential to selecting the right countermeasure(s) to improve safety. For example, if a “dart-out” crash trend occurs at a two-lane residential street crossing and a similar problem is found at on multilane, arterial street, the best-suited countermeasures will likely differ between the two sites. For the two-lane residential crossing site, some type of traffic calming measure may be appropriate to slow down vehicles, in addition to enhancing an established marked crosswalk with appropriate amenities (e.g., advance yield line and markings, overhead lighting, raised crossing), depending on site conditions and crash causes. In contrast, the multilane arterial street might best be treated with one or more measures such as a road diet, median island, and/or some type of traffic control measure (e.g., RRFB or pedestrian hybrid beacon), depending on site conditions. In either the two-lane street or the multilane arterial situation, the final project selection decision requires first gaining a better understanding of the behaviors of road users (drivers, pedestrians, bicyclists), and an understanding of site details which may be contributing to unsafe behaviors. Examples of some of the types of engineering studies that may be needed to understand the likely crash cause(s) and, therefore, which countermeasure(s) will be best suited for the site include: • Road Safety Audit (RSA). This is a detailed investigation of all of the physical characteristics of the location, including the lighting levels, number of lanes, bike lane presence, median or median island presence, signs, signals, markings, curb radius, pavement condition, and many of the characteristics described below. • Vehicle speed study. Collecting the average and 85th percentile motor vehicle speeds may be helpful in identifying whether speed is a factor and what, if any, speed reduction measures may be appropriate. • Traffic signal study. At signalized intersections and crossings, it is important to understand why pedestrians, bicyclists, and motorists behave the way they do. Useful information includes data on cycle length, amount of WALK time (assumed walking speed), facilities for bicyclists (pavement markings, signing, bike signal phasing), placement of stop lines, crosswalk type and condition, size and visibility of the signals for all users, presence of all of the proper signal features (e.g., pedestrian signals present), etc. It is also important to observe vehicle, pedestrian, and bicyclist signal violations. • Geometric condition studies. Whether an intersection or a midblock location is under review, it is important to collect such details as the number of lanes, intersection legs, driveway location and design, stopping sight distance, intersection sight distance, intersection turn radii, sidewalk presence, shoulder type and condition (if any), among others. • Sign and marking studies. Information collected includes presence, type and location of marked crosswalks; crosswalk-related signs and marking, such as advanced STOP/ YIELD signs with markings; other warning and regulatory signing and marking (e.g., speed limit signing); and bikeway type or sharrows. • Delay studies. Because many pedestrian and bike crashes result from risky behavior in response to a lack of road crossing opportunities caused by high traffic volumes, high traffic speed, and/ or high traffic signal delay, it is important to understand the delay incurred by road user type. This activity may involve observing the delay for pedestrians, bicyclists, or motorists caused by a lack of safe gaps in traffic, the existing signal timing, and/or the lack of proper facilities.

41 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS • Bus stop/bus route features study. Because many pedestrian crashes involve people going to or from transit stops, it is important to closely observe features at and near these stops. Is the transit stop on the near side or far side of the intersection? Are there adequate features for pedestrians to cross the street to get to or from a transit vehicle? Are there sidewalks or walkways in the surrounding area that allow pedestrians to access the transit stop? Does the transit shelter block the site lines or the sidewalk? • Accessibility studies. An inability to safely access facilities (sidewalks, paths, etc.) due to a lack of accommodation leaves people vulnerable in the street. Are there curb cuts at each corner? Are the curb cuts well designed, or do they unsafely deposit pedestrians into the intersection? Are the sidewalks in poor condition or obstructed in some way that might discourage their use? A careful review of known crash characteristics, combined with a preliminary site visit, will help in determining which engineering studies are needed to support countermeasure identification. After such studies are conducted, one or more potential countermeasure(s) may be identified. In some cases, however, no obvious site deficiencies will be found, and prior crashes may be the result of obvious road user error (e.g., drunk driver or pedestrian/bicyclist, road user distracted by smart phone or other cause, motorist speeding or running red light). At this point, the final selection of the best treatment(s) may be made, in some cases, with the help of cost effectiveness analysis (e.g., benefit–cost ratio), as described in Chapter 5. 3.5 Assessing Comfort and Safety Thus far, this chapter has discussed the usefulness of gathering and examining safety data, such as crashes, high-risk factors, and conflicts. However, there are other metrics that can be important in quantifying the perspectives those who are walking or bicycling, such as their level of stress, comfort, or perceived safety. While this topic involves safety (both actual and perceived), it also crosses over somewhat into an operational realm, analogous to a motorist level of service. A lack of appropriate features and facilities for bicyclists and pedestrians may impact these road users’ comfort and deter them from bicycling and walking. For example, an intersection with numerous intersecting travel lanes, but no medians or refuge islands, may not only be unsafe for pedestrians, it may dissuade some from walking. How does the latter impact safety? In general, more crashes would be expected with increased numbers of pedestrians and cyclists using an intersection. However, the phrase “safety in numbers” was coined to describe the phenomenon that, as pedestrian and bicyclist volumes increase, assuming no change in motor vehicle volume, fewer crashes can be expected per pedestrian and per bicyclist (Elvik and Bjørnskau 2017). In this way, the level of comfort of pedestrians and bicyclists can impact the expected crashes at an intersection and also improve mobility for those who walk and bicycle. Pedestrian crossing at intersection with truck approaching in Miami Beach. Source: www.pedbikeimages.org/Dan Burden

42 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS Various tools have been developed that rate an intersection or road based on the level of stress or comfort that a pedestrian or bicyclist experiences while interacting with motorists. These tools primarily provide a comfort rating based on the roadway’s characteristics, such as the roadway’s lane configuration, and the proximity and speed of motor vehicles. One of the uses of these metrics is to identify high-stress locations in a network that create barriers to nonmotorized travel. These barriers to walking and bicycling discourage pedestrian and bicycle travel from occurring as it might otherwise naturally occur, leading to pent-up or “latent” demand for safe places to walk and bicycle. For example, this demand can be observed when a new pedestrian or bicycle path is installed and nonmotorized traffic which previously did not exist begins using the path (Krizek et al. 2005). For pedestrians, level of stress can be measured by the Pedestrian Level of Service (PLOS) metric found in the HCM 6th Edition (Transportation Research Board 2016); other metrics, such as the Pedestrian Environmental Quality Index (Center for Occupational and Environmental Health 2011) are also available to assess pedestrian comfort. The list below summarizes information needs (most of which can be defaulted if not readily available) for computing PLOS on segments and at intersections (Dowling et al. 2016). For segment PLOS: • Sidewalk width • Free-flow pedestrian speed • Segment length • Signalized intersection delay walking along street(s) • Signalized intersection delay crossing street(s) • Outside lane width • Bicycle lane width • Shoulder/parking lane width • Street trees or other barriers • Percentage of segment with occupied on-street parking • Landscape buffer width • Curb presence • Median type (divided/undivided) • Number of travel lanes • Directional vehicle volume • Vehicle running speed • Average distance to nearest signal • Intersection PLOS For intersection PLOS: • Traffic signal cycle length • Major street walk time • Minor street walk time • Number of lanes crossed on minor street crosswalk • Number of channelizing islands crossed on minor street crosswalk • 15-minute volume on major street • Number of major street through lanes in the direction of travel • Midblock 85th percentile speed on major street • Right-turn-on-red flow rate over the minor street crosswalk • Permitted left-turn volume over the minor street crosswalk The HCM also offers a Bicycle Level of Service (BLOS) measure. However, since the BLOS metric was developed to measure bicyclist comfort operating in shared lanes and bike lanes (separated bicycle lanes and sidepaths are not included in the model, although a different model for off-street facilities is available), other metrics have since been developed to better capture bicyclist comfort across networks. The Bicycle Compatibility Index (BCI) was first proposed in the late 1990s (Harkey et al. 1998). More recently, Level of Traffic Stress (LTS) (Mekuria et al. 2012) has become popular, and many versions of it have been adopted by different organizations and agencies, including by the original authors (Lowry et al. 2016). Table 13 presents a summary of these various tools for calculating bicyclist stress, comfort, or level of service, and the road characteristics that each one requires.

43 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS Table 13. Data Needs of Selected Rating Systems for Bicyclist Level of Stress and Comfort Attribute BCI BAM BSS MM BLOS WisDOT LTS LTS v2 Travel lane width Bike lane width Shoulder width On-street parking Presence of curb Vehicle volume Vehicle LOS Number of lanes Vehicle speeds Functional class Heavy vehicles Pavement condition Driveways Land use Turn lanes Type of bicycle facility Medians Traffic signals Nonmotorized user volumes Passing restrictions Transportation demand management Notes: • An outlined box indicates availability in Open Street Map data. • BCI (Harkey et al. 1998). • BAM (Bicycle Adaptive Model—original BLOS) (Landis et al. 1997). • BSS (Bicycle Scoring System) (Dixon 1996). • MM (MMLOS)—(Mozer 1998); note this is not the same MMLOS as in the HCM. • BLOS—as adapted for the HCM (Transportation Research Board 2016). • WisDOT (Wisconsin Rural Bike Planning Guide) (Williams et al. 2006). • LTS (Mekuria et al. 2012). • LTSv2 (Lowry et al. 2016).

44 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS Figure 24. Example intersection diagram from Pedestrian and Bicycle Intersection Safety Index. These types of stress or comfort metrics are usually not validated in terms of how they increase walking or bicycling or how they relate to actual safety. The purpose of many of these metrics is to get a bird’s-eye view of potential problem spots where “interested but concerned” bicyclists may be trapped on an “island” of low-stress streets (such as neighborhood streets) separated from the rest of the transportation network by high-stress barriers, often in the form of high- stress roads and intersections. The “interested but concerned” bicyclists, as described by Roger Geller and codified by researchers at Portland State University, represent the majority of the public and are those who “want to ride a bicycle, or are at least ‘curious about bicycling,’ but generally require comfortable facilities, and will not want to ride on streets with heavy motor vehicle traffic” (Dill and McNeil 2016, p. 90). Some research has found a correlation between these measures of comfort and safety, but more research needs to be done to understand the strength of the relationship. However, the indirect relationship goes back to safety in numbers: the more people that walk and bicycle, the safer each person tends to be. This relationship has been found in multiple studies and crash prediction models, and is particularly strong when there are fewer walkers and riders (Elvik and Bjørnskau 2017). There are many hypotheses to explain the “safety in numbers” effect, including that drivers become more accustomed to seeing and looking out for pedestrians and bicyclists when there are more of them; that drivers themselves are more likely to walk or bicycle where there is more opportunity, thus being more careful when driving; and that these places are naturally safer, which is why people walk and bicycle there (although longitudinal research from Seville suggests that this latter explanation is not comprehensively true) (Marqués et al. 2015). While research is still investigating these relationships, one thing is clear: if people do not feel safe, they will avoid bicycling and walking on a given facility or through a given intersection if they have a choice. The result is that locations that feel unsafe are usually more dangerous for those who do walk or bicycle through them. For this reason, it is important to understand perceived safety for bicyclists and Source: Carter et al. 2006

45 GUIDANCE TO IMPROVE PEDESTRIAN AND BICYCLIST SAFETY AT INTERSECTIONS pedestrians when assessing the overall safety of an intersection for bicycling and walking. Some tools have been developed that relate intersection characteristics more directly to safety measures. One set of tools includes the Pedestrian and Bicycle Intersection Safety Indices (Carter et al. 2006). The Pedestrian and Bicycle Intersection Safety Indices (Ped ISI and Bike ISI) are models that take observable characteristics of an intersection crossing or approach leg, such as number of lanes and traffic volume (see Figure 24), and produce a safety index score. Higher scores indicate greater priority for an in-depth safety assessment. The tool provides a rating of the safety of an individual crossing (Ped ISI) or approach leg (Bike ISI). A practitioner can use the tool to develop a prioritization scheme for a group of pedestrian crossings or bicyclist approaches. This method enables the practitioner to prioritize and proactively address sites that are the most likely to be a safety concern for pedestrians or bicyclists without having to wait for crashes to occur. The safety measures used in developing the Ped ISI and Bike ISI included avoidance maneuvers and expert ratings. Ultimately, data on comfort, observed behavior (such as through an RSA), and perceptions of safety, walkability, etc. should be considered along with the safety data whenever possible to inform countermeasure selection that will address multiple needs. Of these tools, the LTS, BLOS, and PLOS are the most widely used and accepted by agencies. 3.6 Conclusion This chapter discussed several approaches to intersection safety assessment for pedestrian and bicyclists and presented common tools and measures of comfort. Using a wide range of available data, agencies can choose to identify sites and risk factors using traditional hot spot (reactive) approaches or systemic (proactive) approaches. In many cases, the availability and quality of data may determine the approach an agency takes. Expanding these analysis methods to capture more complex measures of perceived safety and comfort is a challenge for transportation agencies, but one that is worthwhile to ensure that safety priorities reflect the experience of pedestrians and bicyclists. While it is not possible in this brief overview to detail the methods, we hope that this Guide provides a general overview and points to other resources for deeper study of the topics. The next chapter continues the discussion of intersection safety by focusing on how operation and roadway context impact pedestrian and bicyclist safety. Key sources cited in this chapter: NCHRP Report 893 PEDSAFE BIKESAFE CMF Clearinghouse FHWA Road Safety Audits Highway Safety Manual Highway Capacity Manual Level of Traffic Stress Methods Pedestrians and bicyclists crossing at an intersection in New York City. Source: www.pedbikeimages.org/Laura Sandt