Assessing and Comparing Environmental Performance of Major Transit Investments (2012)

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.

E-1 Appendix E – Level of Service and Other Measures for Assessing Pedestrian and Bicycle Access to Transit  E.1 Introduction This white paper discusses multimodal level of service (LOS) measures and other metrics that can be used to indicate the quality of pedestrian and bicycle access in and around transit station areas. The paper has been prepared in support of TCRP Project H-41, the goal of which is to identify environmental performance metrics for transit projects. Walkability and bikeability metrics were identified in the Phase 1 Interim Report for this project (September 2010) as potential proxy measures for the physical activity and asso- ciated public health benefits of a project (Table H.1, Metric No. 107). While such metrics do not measure physical activity directly, they do provide an indicator of how likely the project is to support additional levels of walking and bicycling based on the presence of a supportive local (station area or corridor) environment. Walkability and bikeability metrics, as well as transit LOS metrics, also could potentially be used as indicators of transportation choice, which may be considered a community and quality of life benefit (Table H.1, Metric No. 109). While it was determined to exclude community and quality of life metrics from further consideration in H-41 and to focus instead on more traditional environmental benefits, such metrics may still be of interest to FTA and project sponsors. Finally, the metrics presented also could potentially be used as quantitative indicators for the transit-supportive land use criterion which FTA currently evaluates from a qualitative standpoint. This paper includes the following sections: • An overview of level/quality of service measures, including walking, bicycling, tran- sit, and multimodal measures; • A discussion of design guidelines and how they may inform the development of metrics; • An overview of area-level walkability indicators; • A summary of previous research for FTA on pedestrian accessibility measures, con- ducted in support of the land use assessment process;

E-2 • Technology applications for level of service measures; and • Conclusions regarding application of these metrics to transit environmental benefits evaluation.  E.2 Level/Quality of Service Measures LOS originated as a concept to measure the quality of automobile travel. The primary methodology for determining transportation LOS measures in the U.S. is presented in the Highway Capacity Manual (HCM). The HCM was first created in 1950 by the Federal government. LOS is represented using a score of A to F (A being the best), and was at first primarily a measure of automobile vehicle speed and levels of congestion along roadway segments and at intersections. The Manual has gone through numerous updates and the most recent version, 2010, includes a multimodal LOS measure (automobile, transit, bicycle, and pedestrian) to compare service changes for different modes along single transportation facilities or corridors. Whereas nonautomobile modes were in earlier versions characterized, often inaccurately, based on travel speeds and capacity, new measures more accurately reflect user perceptions of transit, walking, and bicycling environments. Multimodal Level of Service The concept of multimodal level of service includes a number of methodological frame- works that integrate LOS measures for automobile, transit, bicycling, and walking. While measures for each of these modes have been developed as individual methodologies from various sources, the frameworks have recently been combined to integrate calculations to allow for comparison of multimodal LOS measures on segments of road. Multimodal analysis has been incorporated into the 2010 HCM, due to be released in April 2011. The methodology creates LOS measures for each mode that can be compared and used to measure the effects of changes in transportation infrastructure and transit service. The methods in the 2010 HCM are based on NCHRP Report 616, Multimodal Level of Service Analysis for Urban Streets (and NCHRP Web-Document 128 – Users Guide). This report presents a framework and methods for determining levels of service for the four modes on urban streets. The LOS models were intended to evaluate “complete streets” and “context-sensitive” design strategies and are sensitive to street design (e.g., number of lanes, widths, and landscaping), traffic control devices (signal timing, speed limits), and traffic volumes. For example, improved signal timing increases car and bus speeds which increases car and bus LOS. However, the higher speeds reduce the LOS perceived by bicyclists and pedestrians. Similarly, planners can test the effects on both motorists and bicyclists of reducing a four-lane street to three general travel lanes with bicycle lanes. Table E.1 shows the data required to develop the multimodal LOS.

E-3 Table E.1 Multimodal Level of Service Data Needs Street Geometry Number of through lanes (#) No default Travel lane widths (feet) 12 feet of local default Median width (if present) (feet) 12 feet or local default Bike lane width (if present) (feet) 5 feet or local default Shoulder width (if present) (feet) No default Parking lane width (of present) (feet) 8 feet or local default Presence of barrier in planter strip (yes/no) No default Sidewalk width (if present) (feet) 5 feet or local default Presence of left turn lane(s) at intersection (yes/no) No default Length of analysis segment (feet) No Default Presence of right turn channelization islands at intersections (yes/no) No Default Cross-street through lanes at intersections (#) No Default Cross-street width curb to curb (#) No Default Number of transit stops (#) No Default Percent of transit stops with shelters (%) Use local defaults Percent of transit stops with benches (%) Use local defaults Unsignalized intersection and driveways (#/mile) Use local defaults Pavement condition (1-5) 3 for satisfactory condition Demand Intersection vehicle turning moves (vph) No Default Vehicle right turn on red volume (vph) No Default Vehicle peak hour factor (PHF) 0.92 or local default Percent heavy vehicles 5% or local default Local bus volume (vph) No Default On-time performance of transit (%) 75% or local default Peak passenger load factor for transit (passenger/seat) 0.80 or local default Pedestrian volume (pph) No default Percent of on-street parking occupied (%) 50% or local default Intersection Control Saturation flow rate through lanes (vphgl) 1,800 or local default Green time per cycle for through move (g/c) 0.40 or local default Green time per cycle for cross-street (g/c) 0.40 or local default Cycle length (sec) 100 seconds or local default Quality of progression (1-5) Use 3 for random progress Speed limit (mph) Use local defaults Cross street speed limit (mph) Use local default Source: Dowling, R. (2008), Multimodal Level of Service Analysis for Urban Streets (2008). NCHRP Report 616, Transportation Research Board, Washington, D.C.

E-4 Bicycle and Pedestrian Levels of Service The HCM, and NCHRP Report 616, have incorporated methodologies from past studies to develop walking and cycling LOS measures. Key studies and reports are summarized below. One of the first reports to apply an LOS framework to pedestrian and bicycling facilities was by Linda Dixon for the Transportation Research Record. The report illustrated how a basic LOS rating for both pedestrian and cycling facilities could be constructed.1 Table E.2 Bicycle and Pedestrian Level of Service The score was based on seven criteria each, presented in Table E.2, with the score total resulting in an LOS rating of A through F. Bicycle Score Range Pedestrian Score Range Facility Provided: lane width, off- street 0-6 Facility Provided: type, width, off street 0-6 Conflicts: driveways, barriers, parking, visibility, intersections 0.5-1 Conflicts: driveways, streets, signal delay, crossing dist., road speed, medians 0.5-1 Speed Differential 0-2 Amenities: buffer, benches, lights, trees 0.5-1 Motor Vehicle LOS: lanes 0-2 Motor Vehicle LOS: lanes 0-2 Maintenance (-1)-2 Maintenance (-1)-2 TDM/Multimodal Support 0-1 TDM/Multimodal Support 0-1 Segment Weight: Based on corridor length Segment Weight: Based on corridor length Source: Dixon, L. (1996). The FHWA report Capacity Analysis of Pedestrian and Bicycle Facilities provided a metho- dology to evaluate and implement new types of pedestrian and cycling facilities and incorporate up-to-date information on transportation facility design.2 1 Dixon, L. (1996). Bicycle and Pedestrian Level of Service Performance Measures and Standards for Congestion Management Systems. Transportation Research Record 1538 pages 1–9, Transportation Research Board, Washington, D.C. The methodology was developed for and included in the HCM. 2 Rouphail, N., et al. (1998). Capacity Analysis of Pedestrian and Bicycle Facilities. Report No. FHWA- HRT-98-107, Federal Highway Administration, Washington, D.C.

E-5 The Florida Department of Transportation (DOT), with Bruce Landis, contributed to the pedestrian LOS measures, most recently in 2001 with a quantitative LOS model.3 Landis’ earlier work also included a bicycle LOS model (Table E.3). The study identified roadway and traffic variables describing pedestrians’ perception of safety and comfort and expands on the methodology in the HCM. 4 Ann Vernez Moudon has authored reports empirically studying the attributes of walkable routes and walkable neighborhoods, identifying areas such as grocery stores, parks, and particular land uses as key elements. The model estimates the suitability of a roadway to accommodate cyclists safely based on a similarly quantita- tive process. The model can be used to both evaluate existing facilities and evaluate roadways to identify good locations for future bicycle investments. 5 In a study by Michael Ianoco, researchers note the difficulties of calculating nonmotorized accessibility measures, citing issues with data availability, data quality, the zonal structure of transportation planning models (versus the small-scale areas associated with walking), and the adequacy of models and travel networks for describing nonmotorized travel. 6 The U.S. Department of Transportation developed bicycling-specific LOS measures with the Bicycle Compatibility Index (BCI). The authors present practical strategies for addressing some issues and also suggest that a high degree of accuracy in measurements is not needed for identifying places needing design applications. 7 3 Landis, Bruce, et al. (2001). Modeling the Roadside Walking Environment: Pedestrian Level of Service. Transportation Research Record 1773 pages 82-88, Transportation Research Board, Washington, D.C. The BCI provides common methods to evaluate existing facilities, identify possible improvements, and determine operational and geometric require- ments for new facilities. The index incorporated research on bicyclists’ comfort levels on roadways, reflected in operational conditions. 4 Landis, Bruce W. et al. (1997). Real-Time Human Perceptions: Toward a Bicycle Level of Service. Transportation Research Record 1578, Transportation Research Board, Washington, D.C. 5 Moudon, Anne Vernez, et al. (2006). Operational Definitions of Walkable Neighborhood: Theoretical and Empirical Insight. Journal of Physical Activity and Health, 3, S99-S117. 6 Iacono, M., et al. (2010). Measuring Nonmotorized Accessibility: Issues, Alternatives, and Execution. Journal of Transport Geography 18, pages 133-140. 7 Landis, Bruce, et al. (1998). Development of the Bicycle Compatibility Index: A Level of Service Concept. Final Report, No. FHWA-RD-98-072, Federal Highway Administration, Washington, D.C.

E-6 Table E.3 Landis (1997) Methodology for Bicycle Level of Service Bicycle LOS = a1ln (Vol15/Ln) + a2SPt(1+10.38HV)2 + a3(1/PR5)2 + a4(We)2 + C Where: Vol15 = Vol1 = Where: ADT = D = Kd = PHF = Ln = SPt = SPt = Where: SPp = HV = PR5 = We = Volume of directional traffic in 15 minute time period (ADT x D x Kd)/(4 x PHF) Average Daily Traffic on the segment or link Directional Factor (assumed = 0.565) Peak to Daily Factor (assumed = 0.1) Peak Hour Factor (assumed = 1.0) Total number of directional through lanes Effective speed limit 1.1199 ln(SPp – 20) + 0.8103 Posted speed limit (a surrogate for average running speed) Percentage of heavy vehicles (as defined in the 1994 Highway Capacity Manual) FHWA’s five point pavement surface condition rating Average effective width of outside through lane: Where: We = We = We = Where: Wt = OSPA = Wl = Wps = Wv = And: Wv = Wv = Wv – (10 ft x % OSPA) and Wl = 0 Wv + Wl (1 – 2 x % OSPA) and Wl > 0 and Wps = 0 Wv + Wl – 2 (10 x % OSPA) and Wl > 0 and Wps > 0 and a bikelane exists Total width of outside lane (and shoulder) pavement Percentage of segment with occupied on-street parking Width of paving between the outside lane stripe and the edge of pavement Width of pavement striped for on-street parking Effective width as a function of traffic volume Wt if ADT > 4,000veh/day Wt (2-0.00025 x ADT) if ADT £ 4,000veh/day, and if the street/road is undivided and unstriped. And: a1: 0.507 a2: 0.199 a3: 7.066 a4: – 0.005 C: 0.760 Where: a1- a4 are coefficients established by the multivariate regression analysis. Source: Landis (1997), ibid.

E-7 Variables required for the BCI include: • Lane configuration (number of lanes); • Curb lane width; • Bicycle lane width; • Pavement condition; • Motor vehicle speed; • Traffic volume; • Heavy truck volume; • Right turn volumes; • On-street parking; • Parking time limits; and • Land use types (residential or other). In 2003, Professor Anne Vernez Moudon evaluated 31 bicycling and walking LOS tools, in a broad evaluation of transportation-related “environmental audits.”8 The study identified nearly 200 variables used in the models to describe the travel envi- ronment, which may serve as a catalog of applicable variables for LOS evaluations. The author notes that the large number of variables shows that there is a dearth of research to establish which variables will be best able to reflect relationships between nonmotorized travel and the user’s environment. The author also notes that the models are specific to professional fields, noting that “instruments from the health field tend to undervalue the transportation components of walking and bicycling, whereas those from the transporta- tion field disregard the physical activity aspects of travel.” This would leave most tools falling short of a comprehensive LOS evaluation tool. The document syn- thesized instruments and evaluation methods, including both behavioral and spatial models, and summarized key topics. The inventory included categorization by purpose, user types, professional field, and scale, as well as the year published and a brief descrip- tion. Nine instruments were categorized as “Level of Service” tools (Table E.4). Nearly all of these applied to the route quality, while one evaluated the surrounding area. Five measures were designed for pedestrian facilities and four were for bicycling facilities. The Bicycle Environmental Quality Index by the San Francisco Department of Public Health was developed in 2009. It created a modified collection of 22 indicators, expanding on the BCI by incorporating indicators such as lane marking, slopes, bicycle parking, sur- rounding uses, and traffic calming features. 8 Moudon, A.V., and C. Lee (2003). Walking and Bicycling: An Evaluation of Environmental Audit Instruments. American Journal of Health Promotion, Vol. 18, No. 1.

E-8 Table E.4 Summary of Environmental Audit Instruments Reviewed Level of Service Name Modes Description Botma Bicycle Ranking of bicycle paths and trails based on bicyclist and pedestrian behaviors. Method used to validate audit and ranking not explained. Dixon Bicycle and Pedestrian Ranking of road segments based on roadway characteristics and traffic conditions. Method used to validate audit and ranking not explained. Eddy Bicycle Simple formula to rank road segments based on roadway characteristics and traffic conditions. Method used to validate audit and ranking not explained. Florida DOT Pedestrian Ranking of road segments based on roadway characteristics and traffic conditions. Model calibrated and tested on 75 subjects for perception of safety and comfort. Fort Collins Pedestrian Simple assessment of roadway characteristics, visual interest of environ- ment, and sense of security. LOS for a given area yielded from the ranking. Target LOS provided for different types of pedestrian planning areas and corridors. Khisty-PM Pedestrian Qualitative performance measures of pedestrians perception of safety, security, comfort, convenience, attractiveness, way finding and continuity. Landis Bicycle Ranking of road segments based on roadway characteristics and traffic conditions. Model calibrated and tested on 150 subjects for levels of ‘‘perception’’ in real time. Washington DOT Pedestrian Level of service based on design, location, and user factors; designed to audit road segment. Source: Moudon and Lee (2003), ibid. Private consulting firms have developed methodologies and tools to measure the quality of walking and bicycling infrastructure and experience. As the details for these methods are not publicly available, they are not covered here. General information, however, indi- cates that they are based on the principles laid out by previous research, and are often integrated with land use analysis tools and developed through geographic information systems.

E-9 Transit Level of Service The Multimodal LOS measure incorporated into the HCM and presented in NCHRP Report 616 incorporates public transit LOS measures developed in the Transit Capacity and Quality of Service Manual.9 The transit LOS measures incorporate both service availability and comfort and conveni- ence measures, as shown in Table E.5. The manual emphasizes the importance of the transit user’s perspective in establishing quality of service and presents an A through F scale. The measure is not based on specific national standards, as the authors suggest individual agencies to set values based on their unique settings. Recognizing the difference between fixed-route and demand-responsive service, a separate demand responsive transit LOS was developed with scores of 1 through 8. Table E.5 Summary of Transit Level of Service Measures Fixed Route Transit Demand Responsive Transit Service Availability Headway in minutes Response time Number of hours of service Days and hours available Share of transit-supportive areas covered by transita On-time percentage Percent trips not served Comfort and Convenience Difference between transit and automobile travel times (or only transit travel time) Difference between transit and automobile travel time Passenger load (persons per seat) Standing area (sq. ft. per person) On-time performance Headway adherence Missed trips Mechanical breakdowns Source: Kittleson & Associates, et al. (2003). a Transit-supportive areas is defined in the TQSM as the portion of the area being analyzed that has a house- hold density of at least three units per gross acre or an employment density of at least four jobs per gross acre. 9 Kittleson & Associates, et al. (2003). Transit Capacity and Quality of Service Manual, 2nd Edition. Transit Cooperative Research Program Report 100, Transportation Research Board, Washington, D.C. http://www.trb.org/Main/Blurbs/Transit_Capacity_and_Quality_of_Service_Manual_ 2nd_153590.aspx.

E-10  E.3 Design Guidelines Design guides are available from some municipal governments and can offer features that could be helpful in developing walking and bicycling LOS metrics. The cities of Los Angeles, Portland, and New York City (as well as others) have developed guides to inform the development of new streets. Minnesota DOT has developed a checklist that produces a generalized “score” of facility quality. The United Kingdom has developed a national guide for roadway policy and design, as well as a source of case studies that local governments can apply to a broad range of roadway types. From the sources listed below, metrics include the number of signalized crosswalks, the number or presence of traffic calming design such as bulb outs, posted speed limits, presence of bike lanes, and other nonmotorized safety features. The availability of design guidelines also could be used as a proxy to represent a comprehensive approach to transit supportive planning in an area. • City of Los Angeles, California: http://www.urbandesignla.com/walkability/Crosswalks.pdf. • City of Portland, Oregon: http://www.portlandonline.com/shared/cfm/image.cfm?id=84048. • City of New York, New York: http://www.nyc.gov/html/dot/downloads/pdf/sdm_lores.pdf. • Minnesota DOT: http://www.dot.state.mn.us/bike/pdfs/ Bicycle_and_Pedestrian_Toolbox_2008_04.pdf. • United Kingdom Department for Transport: http://www.dft.gov.uk/pgr/sustainable/manforstreets/. • Transport Canada (Tools for Measuring Roadway Suitability for Bicycles): http://www.tc.gc.ca/eng/programs/ environment-utsp-casestudy-cs44e-bikeindex-270.htm.  E.4 Area-Level Walkability Indicators The LOS metrics described above are primarily intended for application at a facility level, although they can be aggregated across facilities to produce an average LOS for an area. Metrics also have been developed to describe the pedestrian environment and accessibility at an area level. Examples described here include pedestrian environment factors (PEF) and “3D” metrics of the built environment.

E-11 Pedestrian Environment Factors Pedestrian environment factors have been defined as area-level metrics for the purposes of improving mode choice prediction in regional travel demand models. The factors are applied at a traffic analysis zone (TAZ) level, which is a similar geographic scale as the one-half-mile radius used by FTA in New Starts evaluation. The factors are designed to measure the quality of the pedestrian environment, and therefore to relate to the likeli- hood of walk trips occurring within the TAZ. The specific factors that are included in a PEF metric vary, but can generally be measured through some combination of existing GIS data and field and/or aerial surveys. One example of a PEF is that developed by Portland Metro. The factor is based on four criteria:10 • Sidewalk availability; • Ease of street crossing; • Connectivity of street/sidewalk system; and • Terrain. Each of these is rated on a 0 to 3 point scale for a total of up to 12 points. They can be assessed qualitatively or quantitatively, if data are available. Montgomery County, Maryland developed a similar Pedestrian and Bicycle Environment Factor that is based on: • Amount of sidewalks; • Land use mix; • Building setbacks; • Transit stop conditions; and • Bicycle infrastructure. Each factor is assigned fractional points on a qualitative basis for an overall rating of between zero and one for each zone. 3D Metrics The U.S. Environmental Protection Agency’s (EPA) Smart Growth INDEX model included an approach that used elasticities of travel with respect to 3 “D’s” – density, diversity, and design – to predict reductions in vehicle trips and VMT as a result of pedestrian design 10 Schwartz, W.L., et al. (1999). Guidebook on Methods to Estimate Nonmotorized Travel: Supporting Documentation. Prepared for Federal Highway Administration, publication No. FHWA-RD-98-166.

E-12 factors. The “3D” methodology has since been incorporated in other sketch planning tools based on D-factors measured at a TAZ or neighborhood level, like the PEFs described above. The same factors would presumably relate to increases in walking and bicycling trips, although the 3D methodology has not been used explicitly for that purpose. The “design” factor in the Smart Growth INDEX model was specified as either the percent change in locally calibrated PEF, or the percent change in the “design index,” which was computed as follows:11 Design Index = 0.0195 * street network density + 1.18 * sidewalk completeness + 3.63 * route directness where: street network density = length of street in miles/area of neighborhood in square miles sidewalk completeness = length of sidewalk/length of public street frontage route directness = average airline distance to the neighborhood center/average road distance to the neighborhood center The design index for a TAZ or neighborhood can be computed based on data on street centerlines, sidewalks, and location of neighborhood centers if stored in a GIS database.  E.5 Previous Research for FTA on Pedestrian Accessibility Measures Research was undertaken for FTA in 2006 through 2008 to develop an enhanced set of indicators of the potential economic development benefits of transit projects.12 To inform the research, a literature review was conducted to identify key metrics used to describe the walkability or pedestrian-friendliness of a neighborhood in research studies. Metrics were then tested on real-world data. Metrics were tested in three specific evalua- tion categories that relate to pedestrian accessibility: These indi- cators included quantitative metrics describing the existing and planned pedestrian environment in proposed transit station areas. The pedestrian environment indicators were proposed as part of the evaluation factor, “Land use plans and policies encouraging transit-supportive development.” As of this date, the rating system developed through this research has not been adopted by FTA. 11 Criterion Planners/Engineers and Fehr & Peers Associates (2001). Smart Growth INDEX Reference Guide. Prepared for U.S. Environmental Protection Agency. 12 Cambridge Systematics, Inc. for Federal Transit Administration. Guidance for Evaluating the New Starts Economic Development Criterion. Draft working document, September 25, 2008.

E-13 • Pedestrian network coverage and directness; • Sidewalk availability; and • Street crossings. Additional metrics were tested relating to urban design (building setbacks and parking design), mix of uses, residential and commercial densities, and parking constraint. The following specific evaluation subfactors and metrics were proposed for pedestrian net- work coverage and sidewalk availability, and for building setbacks and parking design (which impact the pedestrian environment, if not directly affecting connectivity). It proved too difficult to determine a fair and quantitative metric for plans to provide ade- quate street crossings. Subfactor 1 – Pedestrian Network Coverage and Connectivity Metric: There exists, or plans specify, a continuous pedestrian network in the station area, with an average spacing of pedestrian connections of no more than 600 feet. Ratings: • Required (2):13 − Area plan includes public network connections meeting spacing criteria and/or requirements for accessible connections within private developments; or − A network meeting the criteria already exists (and there are no major redevelop- ment plans that would eliminate blocks). • Recommended (1): − Adopted policies recommend a continuous pedestrian network meeting spacing criteria. • Neutral (0): − Network connectivity not required or recommended. • Not Allowed (-1): − Existing street/parcel layout precludes network connectivity; or − Area plan shows network not meeting spacing requirements. Comments: • For undeveloped areas, refer to area master plans or development policies. For devel- oped areas, use GoogleEarth, a GIS program, or a map and ruler to measure typical block lengths in the vicinity of the transit station. A “block” can be defined by 24-hour publicly accessible pedestrian passages, as well as streets. Parking lots do not count unless there is a defined pedestrian route, primarily separated from traffic, which tra- verses the lot. 13 Numbers in parentheses represent points assigned.

E-14 • If there is a mix of block lengths, some less than and some exceeding the 600-foot threshold, use the following approach: With a path measurement tool, measure the perimeter of the four blocks located closest to the transit station (i.e., those with any part of the block closest to the station). Compute the average block face length by dividing the total perimeter of all four blocks by the total number of block faces (usually 16). Subfactor 2 – Sidewalk Availability Metric: Sidewalks (minimum eight feet wide in commercial areas containing street- fronting retail uses, five feet elsewhere) provided along all street frontage. Ratings: • Required (2): − Sidewalks required for new development. • Recommended (1): − Adopted policies recommend sidewalks for new development. • Neutral (0): − Sidewalks not required or recommended. • Not Allowed (-1): − Sidewalks discouraged or prohibited (not likely to be assigned). Comments: • If the area is already covered by a publicly maintained sidewalk system and there is clear evidence that the city either provides or requires sidewalks in conjunction with new development, a (+1) rating may be assigned even if sidewalk provision is not explicitly addressed in the zoning code or other municipal ordinances. • Google’s Streetview program allows for two-dimensional viewing of some metropoli- tan areas at street level, in effect allowing one to drive the streets. This tool may be helpful in identifying the existence of sidewalks and pedestrian connections. Subfactor 3 – Building Setbacks Metric: Setbacks along street frontages (distance from the front of the building to the lot line) are no more than 15 feet for commercial or mixed-use properties and no more than 20 feet for residential properties. Ratings: • Required (2): − Maximum setbacks (as specified in zoning or binding design guidelines) are less than thresholds. • Recommended (1): − Setbacks may be less than or greater than thresholds; setbacks less than the threshold are recommended in adopted policy or plan documents or design guidelines.

E-15 • Neutral (0): − Setbacks may be less than or greater than thresholds; no guidance specified in pol- icy or plan documents or design guidelines. • Not Allowed (-1): − Minimum setback requirements are greater than the thresholds. Comments: • If the setback condition is met for some uses but not others, see the guidance above under “spatial extent.” • Setback requirements will generally be found in the section of zoning pertaining to a specific type of use (residential, commercial, etc.). Different setback requirements may also be specified for overlay districts (e.g., pedestrian or transit overlay). • It may not be possible to rate this factor for institutional areas (e.g., college or hospital campuses) as the traditional concept of a setback from the street may not be meaning- ful in a campus environment. Subfactor 4 – Parking Design Metric: No more than 30 percent of the street-fronting parcel length is for parking or automobile access/egress. Ratings: • Required (2): − Zoning code establishes this or a functionally similar requirement (e.g., parking must be in structures or behind building). • Recommended (1): − Design guidelines adopted for this area include this or functionally similar recommendation. • Neutral (0): − Location and design of parking not specified. • Not Allowed (-1): − Parking is required in front of buildings (not likely to be assigned). Comments: • This metric is intended to focus on parking for newly built commercial, mixed-use, or multifamily structures. Except for districts with special design standards, such as transit or pedestrian overlay districts, most zoning codes will not specify the location of parking for these types of uses. Some zoning codes prohibit parking in the front yards of residential lots, but this alone should not justify a positive rating for this factor. • Institutional master plans may be rated for this factor based on the extent to which parking is planned to be in structures versus surface lots. For example, a 2 rating could be assigned for master plans that call for all new future parking supply to be accommodated in structures and for redevelopment of surface lots with buildings.

E-16  E.6 Technology Applications for Level of Service Measures Various technical tools (in addition to general statistical and GIS software) have been developed to assist in the evaluation of walking and cycling paths. • Walk Score is available through a web site at no cost to the user.14 • Ped INDEX is a GIS application developed by Fehr & Peers to assess a community’s pedestrian needs. The application is based on a GIS using Open Street Map and other data sources representing facilities, amenities, and other factors. The application produces a score from 1 to 100, based on distances to different types of amenities and destinations and road quality, with the latter, including intersection density, link/node ratios, and average block lengths. However, Walk Score is not available to produce scores for batches of locations, nor for overall corridor accessibility. The developers also have created Transit Score, which applies a similar process to evaluate neighborhood access to public transportation. 15 • WBC Analyst is a GIS application developed at the University of Washington College of Architecture and Urban Planning. It has been considered theoretical and somewhat difficult to operate. The process can identify key pedestrian locations through a process developed for the U.S. Environmental Protection Agency’s Smart Growth INDEX. (Smart Growth INDEX is a sketch planning GIS tool for comparing alterna- tive land use and transportation scenarios.) The product is an overall index of an area’s walking potential and pedestrian facilities, identifying locations where pede- strian improvements can have the greatest safety benefits and encourage walking. The application evaluates demographics and socioeconomic data, distance to amenities, the pedestrian environment, policy areas, and the condition of blocks, traffic, and intersections. 16 14 http://www.walkscore.com/. Research to develop the tool included quantitative analysis of land use and transportation data for King County, Washington, as well as a telephone survey to assess residents’ propensity for walking and cycling. 15 http://www.smartgrowthplanning.org/PDFs/PedIndexBrochureWeb.pdf, accessed January 20, 2011. 16 http://proceedings.esri.com/library/userconf/proc05/papers/pap1040.pdf, accessed January 20, 2011.

E-17  E.7 Conclusions Regarding LOS and Walkability Metrics in Transit Project Evaluation Multimodal LOS measures and other walkability metrics represent a potential way of assessing the existing or planned pedestrian and bicycle-friendliness of transit station areas, and therefore the potential extent to which the transit project may support increased physical activity, reductions in vehicle-travel, transportation choice, and other community quality-of-life factors. (They also can help inform local planning activities by identifying deficiencies in pedestrian and bicycle access.) The multimodal level of service measure, incorporated into the 2010 HCM, combines LOS measures for automobile, transit, bicycling and walking into one widely applicable methodology that allows for basic com- parisons between modes. The LOS measures are primarily intended to inform facility design, and are typically computed at a level of an individual roadway segment, rather than an entire station area. However, the segment-level scores can be averaged across an area to yield one LOS score by mode. Other metrics, including pedestrian environment factors, the walkability metrics developed for FTA, and WalkScore, are designed to be applied at an area level. Computing multimodal LOS or areawide walkability measures requires a fair amount of detail that would need to be collected for all major streets in a station area, although the detail can be combined using a software package such as the HCM methodology or Ped Index. In most cases, project sponsors or their local partners are unlikely to have collected all of the required data, unless they have already conducted areawide bicycle and pede- strian planning studies. Therefore, if such metrics are to be applied in New Starts project evaluation, it may be desirable to rely on simplified assessment methods such as the metrics proposed for FTA’s land use assessment process. In the future, as local jurisdic- tions expand and improve their GIS data to more fully encompass bicycle and pedestrian- related variables, it may be possible to use more sophisticated tools to compute multi- modal LOS or walkability metrics for different transit projects.