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.
77 A P P E N D I X D Given that the NCHRP Project 17-56 data collection consisted of short (1-hour or 2-hour) pedestrian counts, it was necessary during the analysis phase to convert these short counts into day-long pedestrian volumes. To address this need, the pedestrian volume data for the Charlotte intersections were used to calculate expansion factors to use during analysis. These data were also previously used to determine the optimum time of day for pedestrian counting and the most efficient length of count. Expanding Short Counts to Daily Volumes The calculations below are based on pedestrian volume from 204 unsignalized intersections in Charlotte. The Charlotte DOT provided historical pedestrian crossing volumes for these inter- sections, collected during its regular turning movement counts. These counts ranged in length from 12 to 17 hours, but were typically 12-hour counts between the hours of 7:00 a.m. and 7:00 p.m. The data used in this analysis were only from the period of 7:00 a.m. to 7:00 p.m., even though more hours were available at some sites. None of these intersections were in the Central Business District (uptown Charlotte). Charlotte pedestrian counts were provided for a full day (typically 12 hours) in 15-minute increments. This provided the ability to know the âground truthâ (the actual full 12-hour count) for any intersection as well as the counts by specific times of the day (e.g., 4:00 to 5:00). Based on previous analysis, the afternoon and early evening hours were the target time periods counted during the NCHRP Project 17-56 data collection effort. The expansion factors were calculated as the total pedestrian count (âground truthâ) divided by the number of pedestrians counted in time period x. Table D-1 provides an example of how the expansion factor calculation for particular time periods would work for a single intersection with 135 pedestrians counted for the full day. However, to calculate general expansion factors based on the entire group of 204 inter sections, the count data were aggregated first and then used to determine expansion factors for each time period. The expansion factors in Table D-2 were calculated using a citywide group of unsignalized intersections in Charlotte. Since this amount of detailed pedestrian volume information was not available in the other cities used in NCHRP Project 17-56, these factors were used to expand the 1- and 2-hour counts in those cities. For example, for a particular site in another city, if the field data collection of pedestrian volume was done from 5:00 to 6:00 p.m. and resulted in a count of 10 pedestrians, the estimated daily total for that intersection would be 10 Ã 9.22 = 92 pedestrians. Analysis of Charlotte Pedestrian Volumes to Determine Method for Adjusting Pedestrian Volume Counts
78 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments The factors presented in Table D-2 are general factors for any count taken, regardless of time of year. They are however weekday-only factors, since all pedestrian counts done by the City of Charlotte were performed on weekdays only. This synchronizes well with the NCHRP Proj- ect 17-56 counts, which were also taken only on weekdays. It is reasonable to expect that weather variation throughout the seasons of the year will affect pedestrian activity. Table D-3 presents expansion factors by season. These factors were calcu- lated in the same manner as for the entire group of intersections but using only intersections with counts in the specific season. To be as accurate as possible, the process for expanding short- term counts in NCHRP Project 17-56 used the seasonal expansion factors shown in Table D-3. Comparison to Other Studies and Cities Table D-4 compares the expansion factors calculated from the Charlotte data to factors based on Seattle data used in a previous FHWA study, Safety Effects of Marked versus Unmarked Cross- walks at Uncontrolled Locations: Final Report and Recommended Guidelines, FHWA-HRT-04-100 (Zegeer et al., 2005) and factors calculated using data obtained in NCHRP Project 7-19 research. The Seattle data are based on a group of sites in Seattle; the NCHRP Project 7-19 data were from a limited number of sites (five) that were geographically far apart. The Charlotte expansion fac- tors match fairly closely to those calculated from Seattle data. The expansion factors from the five sites included in NCHRP Project 7-19 are mostly higher than the Charlotte and Seattle fac- tors, but generally not by a large amount. This comparison validates the use of the expansion factors calculated from Charlotte data for the NCHRP Project 17-56 analysis. Ground Truth (whole day pedestrian count) Time Period Pedestrian Count Expansion Factor (ground truth/short count) 135 4:00-5:00 11 12.27 5:00-6:00 3 45 6:00-7:00 13 10.38 4:00-6:00 14 9.64 5:00-7:00 16 8.44 Table D-1. Example of expansion factor calculation for a single site. Ground Truth (whole day pedestrian count, total of all intersections) Time Period Pedestrian Count Expansion Factor (ground truth/short count) 34,824 4:00-5:00 3,248 10.72 5:00-6:00 3,776 9.22 6:00-7:00 3,223 10.8 4:00-6:00 7,024 4.96 5:00-7:00 6,999 4.98 Table D-2. Expansion factors calculated from entire group of intersections. Table D-3. Expansion factors by season. Season 4:00-5:00 5:00-6:00 6:00-7:00 4:00-6:00 5:00-7:00 Winter (Dec-Feb) 10.72 8.92 12.92 4.87 5.28 Spring (Mar-May) 10.27 9.36 10.03 4.90 4.84 Summer (Jun-Aug) 9.62 8.59 12.78 4.54 5.14 Fall (Sep-Nov) 9.77 8.31 8.26 4.49 4.14
Analysis of Charlotte Pedestrian Volumes to Determine Method for Adjusting Pedestrian Volume Counts 79 Adjusting Pedestrian Volume through the Years Another question for the analysis is if and how pedestrian volumes in past years should be esti- mated based on present day counts. Essentially, most treatment and comparison sites included in NCHRP Project 17-56 only have pedestrian volume from 1 yearâtypically 2014 when the field data collection was done. A before-after analysis would need to have annual pedestrian volumes for the past years at the intersection. Again, Charlotte pedestrian volumes were used to examine this issue. Using the same group of intersections as listed above, certain intersections were identified as having two or more pedestrian volume counts in different years. For example, an intersection may have been counted in 2004 and again in 2009. This provides the basis to examine annual trends in pedestrian volumes. From the above group of intersections, there were 56 intersections which had two or more counts and thus could be used in this analysis. For each of these, the growth rate was calculated as ( )= â1r FPn where r = growth rate F = future year (i.e., the later year of the pair) P = present year (i.e., the earlier year of the pair) n = time between the two years (in years) For example, if an intersection had a pedestrian count of 40 peds/day in 2004 and 60 peds/day in 2009, the growth rate would be calculated as: 60 40 1 0.084 or 8.4% 2009 2004r r ( )= â = ( )â Table D-4. Comparison of expansion factors from other cities. Time Period Expansion Factor from Charlotte Data Expansion Factor from Seattle Dataa (range is from fringe to residential) Expansion Factors Calculated from NCHRP Project 7-19 Washington D.C. â Key Bridge Davis, CA â Sycamore Park Minneapolis, MN â 15th and Como Portland, OR â 5th Ave San Francisco, CA â Fell St 4:00- 5:00 10.72 7.9 - 9.3 18.45 15.58 12.89 12.27 15.82 5:00- 6:00 9.22 8.1 - 11.4 10.46 15.11 13.33 12.80 12.79 6:00- 7:00 10.8 n/a 8.74 19.25 16.92 33.00 11.27 4:00- 6:00 4.96 n/a n/a n/a n/a n/a n/a 5:00- 7:00 4.98 n/a n/a n/a n/a n/a n/a a Zegeer, C. et al. Safety Effects of Marked versus Unmarked Crosswalks at Uncontrolled Locations: Final Report and Recommended Guidelines, Federal Highway Administration, FHWA-HRT-04-100, 2005.
80 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments Using this formula, growth rates were calculated for each intersection, specific to the combi- nation of the original year and the later year (i.e., 2004 to 2009). Intersections that were counted in the same two years were grouped, and the average of their growth rates was taken. Table D-5 presents the growth rates in terms of the âfromâ year and âtoâ year. Unfortunately, the group of 56 intersections that had at least two counts through the years was a small group to begin with. Thus, the numbers of intersections used to calculate the growth rates in each cell of Table D-5 were very small. Table D-6 provides the number of intersections used for each cell in Table D-5. Based on the small amount of data available and the wide range of growth rates shown above, there is no basis on which to make a recommendation for adjusting present day pedestrian volumes to estimate past year volumes. Table D-5. Growth rates by year. To From 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2002 62% 18% 68% 7% 9% -1% 38% 27% -4% 2003 250% 53% 32% 6% 7% 11% 13% 4% 1% 2004 -73% -12% 3% -29% 19% 1% 9% 2005 27% 10% 2006 129% 11% 35% 12% -3% 2007 26% -14% -18% -26% -3% 2008 17% 71% -3% 2009 -14% 2010 34% 2011 -9% 2012 28% Table D-6. Number of intersections used to calculate each growth rate. To From 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2002 1 1 1 1 1 1 1 2 1 2003 1 3 1 4 1 3 5 2 6 2004 1 3 1 1 2 2 1 2005 1 1 2006 1 2 2 1 2 2007 2 7 1 1 5 2008 2 2 1 2009 1 2010 3 2011 1 1 2012 3