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.
57  Case Studies This chapter presents three case studies demonstrating the use of context classification. The first case study focused on central Kentucky and applies context classification at the regional level and describes step-by-step procedures for performing automated, GIS-based classification and manual review. By following the instructions in this section, practitioners can replicate automated classification for their areas of interest. The second case study examines a roadway corridor in Orlando, Florida, and explores the use of context classification to develop transpor- tation expectations. The final case study investigates circumstances that warrant the adoption of a special context in Toledo, Ohio. 9.1 Case Study on Application This case study reviews GIS-based context classification at the regional level using the measures discussed in Chapter 2. It focuses on classifying all roadways in the Kentucky Trans- portation Cabinetâs (KYTC) District 7. Located in central Kentucky, the district encompasses 12 counties; 15 incorporated cities, including the Lexington area MPO; and contains 7,234 miles of roadway. Each subsection provides concise instructions that practitioners can follow to automate their classifications. 9.1.1 Road Network Identification The first step is to retrieve data for the roadway network being classified. Information for statewide and regional classification can be downloaded from nationwide GIS databases (e.g., TIGER/Line) or agency-maintained databases. All roads in KYTC District 7 were identified using the TIGER/Line âAll Roadsâ shapefile (Figure 41). 9.1.2 Collect Required Data Using the data sources noted in Table 9, retrieve data that will be used to calculate listed measures (see Section 2.1.1 for explanations of each measure). GitHub houses the building footprint database for all states at the following link: https:// github.com/Microsoft/USBuildingFootprints. U.S. Census Urbanized Areas Boundaries are available from the Tiger/Line Shapefiles website maintained by the U.S. Census Bureau. C H A P T E R 9
58 Context Classification Application: A Guide 9.1.3 Classify Urbanized and Non-Urbanized Areas 1. Import the roadway network file and U.S. Census Urban Areas Boundaries into ArcGIS. 2. Based on the U.S. Census Urban Areas Boundaries, split the roadway network into: (1) areas inside urbanized areas and urban clusters and (2) areas outside urbanized areas and urban clusters. 3. Assign roadways inside urbanized boundaries to the category Urbanized Contexts. This encompasses the following contexts: suburban, urban and urban core. 4. Assign roadways outside of urbanized boundaries to the category Non-Urbanized Contexts. This encompasses the following contexts: rural and rural town. 5. Assign roadways located on the boundary line of an urbanized area/cluster to Urbanized Contexts. Figure 42 depicts KYTC District 7âs urbanized boundaries. Figure 41. KYTC District 7 roadway network. Building Density Microsoft Maps U.S. Building Footprint database Building Area Density Microsoft Maps U.S. Building Footprint database Intersection Density Block Length TIGER/Line Street Network Measure Data Source Table 9. Summary of data sources.
Case Studies 59 Urbanized boundaries are updated every 10 years in conjunction with the U.S. Census. Practitioners should periodically review urbanized boundaries to ensure development has not progressed beyond the boundary since the most recent update. Figure 43 maps development on the east side of Lexington, Kentucky, that has occurred outside of the urbanized boundary and may be considered for an urbanized designation. 9.1.4 Roadway Segmentation 1. In ArcGIS, use the Intersect tool to create a geodatabase of points located where roadway line segments intersect the urban boundary polygon. 2. In ArcGIS, use the Generate Points Along Lines and Split Lines at Point tools to segment road- ways. Urbanized roadways are split into 0.50-mile segments. Non-urbanized roadways are split into 0.25-mile segments. Roadways are segmented to provide a smaller unit for classification and define the influence area for recommended context measures. This method assumes that roadway segments are continuous along a route, as is native to the TIGER/Line AllRds shapefile. If roadways were previously segmented along jurisdictional boundaries, first create a continuous route in ArcGIS using the Create Route tool. This tool requires identification of Figure 42. Urbanized boundaries in KYTC 7.
60 Context Classification Application: A Guide a unique route identified in the attribute table. Roadways vary in length and as a result some remainder road segments will be created that are shorter than the 0.25- or 0.50-mile road segments. Include these segments in the classification. 9.1.5 Buffer Analysis 1. In ArcGIS, use the Buffer (Analysis) tool to create buffers around each line segment. For Urbanized areas, use a 0.25-mile buffer. For Non-Urbanized areas, use a 0.125-mile buffer. 2. In ArcGIS, use the Calculate Geometry tool to compute the area of each buffer. Polygons generated through buffer analysis are used to calculate associated measures through a spatial join analysis (Figure 44). Figure 43. Urban growth outside of urbanized boundary in Lexington, Kentucky.
Case Studies 61 9.1.6 Calculate Intersection Density 1. In ArcGIS use the Intersect (Analysis) tool to analyze TIGER/Line data and create a geo- database of points at each intersection in the AllRds shapefile. 2. Intersect (Analysis) generates a point at every location where two or more lines intersect for each line segment at an intersection. At a four-way intersection, the tool generates four unique point features at the same location. 3. In ArcGIS, use the Delete Identical tool to remove redundant intersection points. 4. In ArcGIS, use the Spatial Join (Analysis) tool to perform a spatial join on the intersection point features and roadway segment buffer. 5. Calculate intersection density by dividing the count of intersection point features by the area of the roadway segment buffer. 9.1.7 Calculate Building Density 1. Import the building footprint database into ArcGIS. 2. In ArcGIS, use the Spatial Join (Analysis) tool to perform a spatial join on the roadway segment buffer and the building database (Figure 45). 3. Calculate building density by dividing the count of building features by the area of the roadway segment buffer. 9.1.8 Calculate Building Area Density 1. Import the building footprint database into ArcGIS. 2. In ArcGIS, use the Calculate Geometry tool to compute the area of each building footprint. 3. In ArcGIS, use the Spatial Join (Analysis) tool to perform a spatial join on the roadway segment buffer and the building database. 4. Calculate building area density by dividing the sum of building areas within the segment buffer by the area of the roadway segment buffer. Building area density captures the denser building footprint typical of urban and urban core context development patterns. Figure 44. Buffer analysis of single roadway segment.
62 Context Classification Application: A Guide 9.1.9 Automated Context Classification Once all measures are calculated for the entire network, segments are classified according to the established thresholds. Thresholds for urbanized and non-urbanized areas are summarized in Table 10. Based on these thresholds, in ArcGIS use Select By Attributes to select all segments that have a building area density greater than 5.6 million sq. ft/sq. mi. The Field Calculator is used to manually assign the urban core designation to the context field in the attribute table. The same procedure is completed for urban, suburban, and rural town segments, based on the thresholds in Table 10. Figure 46 presents context classifications for KYTC District 7. 9.1.10 Refinement of Context Classification Once automated context classification is complete, practitioners should refine the results through a manual review. This review can be undertaken using aerial photography, land-use data, Google Street View, and local knowledge. Where questionable classifications are identified, roadway segments are compared to context definitions. After completing their review, practi- tioners can manually edit segment classifications in ArcGIS by entering the appropriate context Figure 45. U.S. Building Footprint database. Urbanized Areas Context Context Building Area Density Intersection Density Urban Core > 5,600,000 sq. ft/sq. mi. â Urban < 5,600,000 sq. ft/sq. mi. > 110 intersections/sq. mi. Suburban < 5,600,000 sq. ft/sq. mi. < 110 intersections/sq. mi. Non-Urbanized Areas Building Area Density Intersection Density Rural Town > 710 buildings/sq. mi. > 185 intersections/sq. mi. Rural All remaining non-urbanized segments Table 10. Context classification thresholds.
Case Studies 63 in the attribute table. Manual review of this study area (with over 7,000 miles of roadway) was completed in under 3 hours by a two-person team. Urban/Suburban Manual Context Classification Figure 47 shows the results of automated classification for a portion of KYTC District 7 (southwest Lexington). The area includes a mix of urban and suburban contexts. A review of land uses (Figure 48) indicated that Area 1, which automated classification designated urban, is residential with limited commercial services and thus does not meet the criteria for an urban context. The lack of service land uses near these residential areas also reduces the appeal and feasibility of nonmotorized transportation options. Figure 47 captures the land uses in Area 2. This area is more thoroughly discussed in Section 2.6.1. Its intersection density is artificially high due how the roadway network is coded in the roadway shapefile. Figure 49 shows the final context classification after completing a manual review. Rural Town Manual Context Classification Rural town classification attempts to identify at least one roadway segment in each rural town (see Section 2.1.1). This requires a manual review of relevant segments to identify adjacent Figure 46. Automated context classification.
64 Context Classification Application: A Guide Area 1 Area 2 Legend Urban Suburban Figure 47. Urban/suburban automated context classification results. Figure 48. Urban/suburban land-use map (note: red is residential).
Case Studies 65 segments in the rural town road network. Roadways that clearly contribute to the rural town network and primary roadways that have developed are included in the rural town. Figure 50 maps the segments in KYTC District 7 automated classification identified as rural town. Several roadways in the area contribute to the townâs overall network. Figure 51 shows the final iden- tification of all rural town roadways. Once all urbanized and non-urbanized segments have undergone manual review, the final context classification is complete (Figure 52). Area 1 Area 2 Figure 49. Final urban/suburban classification after manual review. Legend Rural Rural Town Figure 50. Rural town automated context classification.
66 Context Classification Application: A Guide Figure 51. Final rural town after manual review. Figure 52. Final KYTC District 7 context classification after manual review.
Case Studies 67 9.2 Use of Context Classification to Develop Transportation Expectations 9.2.1 Confirming Project Intended Outcomes The second case study examines a hypothetical project in Orlando, Florida. The focal point is a road that traverses a community in which many low-income households have no access to a personal vehicle. Adequate and continuous pedestrian and bicycle facilities are absent. Key project goals include creating pedestrian and bicycle facilities, offering safe access to transit, and giving community members access to a range of transportation options. 9.2.2 Identifying Context Classification for Project Area First, determine whether previous analysisâat any scaleâhas produced a context classification for the project area. ⢠Statewide. If a preliminary context has been established through statewide analysis, examine the typical characteristics of that context. At the projectâs outset, verify the correctness of this context designation. ⢠Regional. Sometimes projects fall within the jurisdiction of a local or regional agency that also has defined the preliminary context. At the projectâs outset, verify that this context designation is correct. ⢠Project Level. If no preliminary context has been defined, gather, review, and analyze avail- able data to establish the context. For the Orlando roadway, a statewide or regional designation would have established a suburban designation. Without a statewide classification, information on land-use types, intersection densities, building setbacks, and building area densities should be reviewed (see Section 1.5.1 for details). Analyzing these data returns a classification of suburban. Given the variability in suburban form, local/project-level information is used to refine transportation expectations. Information presented in the next section offers insights into the corridor, its users, and transportation expectations. 9.2.3 Defining Transportation Expectations Users/Vehicles The corridor mostly serves shorter-distance trips or trips internal to neighborhoods/ communities (districts) along the roadway (Figure 53a). This holds true for transit trips as well (Figure 53b). U.S. Census Bureau data indicate dense concentrations of residents along some portions of the corridor who live at or below the poverty line and do not rely on single-occupant vehicles for their commutes. Transit ridership is relatively high, with some stops serving over 600 trips per day (Figure 54). Consistent with this, levels of pedestrian and bicycle activity are higher than is typical of sub- urban corridors. Heavy freight traffic (vehicles with more than two axles) accounts for less than three percent of traffic volumes throughout the corridor. Intended project outcomes highlight the need for pedestrian and bicycles facilities to pro- vide feasible choices for traveling along and across the road. Installing pedestrian and bicycle facilities will improve access to transit and enhance the overall experience within the community. A design oriented around pedestrians and bicyclists should be a priority.
68 Context Classification Application: A Guide Movement Vehicle volumes are between 20,000 and 50,000 vehicles per day, and average travel speeds during the peak hour are relatively high. Nonpeak travel speeds range from 31 mph to over 50 mph (Figure 55). Despite the strong demand from pedestrians, bicyclists, and transit users, vehicle volumes and speeds along the corridor are not supportive of nonmotorized modes. The corridor averages seven vehicle crashes per day, with a significant number of crashes involving pedestrians and bicyclists (Figure 56). People depend on travel options other than motor vehicles to move between origins and destinations. A key project goal is to provide continuous pedestrian and bicycle facilities along the corridor as well as easy access to transit. Design solutions need to account for inter- actions between fast-moving motor vehicles and slower, nonmotorized and vulnerable users. (a) (b) Figure 53. Travel patterns: (a) trips among zones and (b) transit trips. Source: LYNX SR 436 Transit Corridor Study.
Case Studies 69 Percent Figure 54. Transit ridership along the corridor. Source: LYNX SR 436 Transit Corridor Study.
70 Context Classification Application: A Guide Figure 55. Travel speeds along the corridor. Source: LYNX SR 436 Transit Corridor Study. Permeability Block sizes are larger than typical for a corridor with similar characteristics to this one. Intersection spacing exceeds the distance pedestrians may be willing to tolerate to cross the road. Design solutions should provide appropriate spacing between pedestrian crossings that are safe and comfortable. Midblock crossings may be needed. Transit stops should coincide with enhanced crossings so that vulnerable users have protected crossing opportunities. Intersections that provide access to the neighborhood for all users and should be designed accordingly.
Case Studies 71 Figure 56. Crashes along the corridor. Source: LYNX SR 436 Transit Corridor Study.
72 Context Classification Application: A Guide Network The road network is a disjointed grid system. Design solutions should deliver improve- ments along the road and at intersections as well as numerous and evenly spaced crossing opportunities, including midblock crossings. A limited grid network may require having intersections with additional turn lanes for operational purposes. One design option is to allow congestion at intersections by minimizing intersection footprints. This solution integrates pedestrians and bicycles by minimizing their exposure and reducing crossing distances. Speed The posted speed limit is 35â50 mph (Figure 57). The design speed is 45â60 mph. Considering community needs, potential safety challenges, and the requirement to provide pedestrian and bicycle facilities, design solutions should incorporate a slower target speed to minimize speed differentials between motorized users and nonmotorized users. Design solutions also should integrate speed management treatments as these can channel motorists into the community at a slower, more appropriate speed. If slower speeds are not feasible, design solutions should include separated pedestrian and bicycle facilities and frequently protected crossings (e.g., a red traffic control device). 9.3 Case Study of a Special Context The final case study examines Front Street in Toledo, Ohio, and illustrates the utility of a special context designation (Figure 58). Context definitions dictate the corridor should be classified as suburban. Despite being inside the cityâs urbanized boundary, the road network and levels of development are less dense than is typical of urban contexts. Community service and retail outlets are absent. On the west side of Front Street, land uses are industrial and port, while on the east side residential and commercial land uses dominate. 9.3.1 Refining Transportation Expectations Users/Vehicles The roadway connects the port and industrial areas to I-280 and serves significant through and regional trips. Residential areas to the east could use the facility as a connector to other areas in Toledo. However, limited local traffic is anticipated. Transit services are available and sidewalks are present along the corridor. The residential area creates the potential for pedestrian and bicyclist traffic, especially with the presence of an elementary school. Data from the Ohio DOTâs Traffic Monitoring Management System suggest that over 50 percent of traffic consists of heavy vehicles. Accordingly, pedestrian and bicyclist integration in the vicinity of transit stops should be evaluated as well as connections between residential areas and transit stops. Movement Vehicle volumes are between 6,000 and 8,000 vehicles per day, with a reported 85th percentile speed of 45 mph due in part to the wide five-lane cross section. Residential and commercial land uses could generate pedestrian, bicyclist, and transit demand. However, vehicle volumes and high vehicle speeds discourage nonmotorized users. Transit stops may require special
Case Studies 73 Figure 57. Posted speeds along the corridor. Source: LYNX SR 436 Transit Corridor Study.
74 Context Classification Application: A Guide attention to ensure pedestrians can safely cross the roadway and access residential and com- mercial land uses. Permeability Industrial and port land uses to the west have resulted in large blocks, while the east side has neighborhood block sizes typical of residential areas (500Â ft average). The residential area provides adequate connections to key facilities (e.g., elementary school, library, churches) that could be served with various modes (driving, walking, or bicycling). Significant heavy vehicle traffic associated with the industrial area may be seen as a barrier for nonmotorized users. Transit stops should be reviewed to ensure they coincide with enhanced crossings that provide protected crossing opportunities for vulnerable users. Intersections that offer access to the neighborhood for all users and should be designed accordingly. Figure 58. Front Street, Toledo, Ohio. Source: Google Earth.
Case Studies 75 Network The road network has a well-defined grid system in the residential and commercial areas. Adequate connections to Front Street exist. Midblock crossings, especially near transit stops, should be a key design consideration and provided as needed. Because the roadway serves a mix of land-use types and freight vehicles are common, it is important for design solutions to integrate all users. Freight traffic is concentrated along the roadway because it is connected to I-280. This may pose risks to pedestrians and bicycles. Speed The posted speed limit is 35Â mph. The Ohio DOT Traffic Monitoring Management System lists the 85th percentile speed as 45Â mph. Front Streetâs characteristics and transportation expectations indicate the proper classification is special context: industrial/port/warehouse. This designation enables additional design flex- ibility and underscores the importance of balancing the needs of freight vehicles and residential and commercial traffic. Keeping speed differentials between motorized users and nonmotorized users to a minimum is a key design consideration. Areas near transit stops should be reviewed to ensure they can properly and safely integrate nonmotorized users.
