National Academies Press: OpenBook

Speed Reduction Techniques for Rural High-to-Low Speed Transitions (2011)

Chapter: CHAPTER TWO Literature Review

« Previous: CHAPTER ONE Introduction
Page 8
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 8
Page 9
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 9
Page 10
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 10
Page 11
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 11
Page 12
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 12
Page 13
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 13
Page 14
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 14
Page 15
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 15
Page 16
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 16
Page 17
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 17
Page 18
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 18
Page 19
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 19
Page 20
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 20
Page 21
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 21
Page 22
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 22
Page 23
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 23
Page 24
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 24
Page 25
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 25
Page 26
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 26
Page 27
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 27
Page 28
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 28
Page 29
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 29
Page 30
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 30
Page 31
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 31
Page 32
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 32
Page 33
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 33
Page 34
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 34
Page 35
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 35
Page 36
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 36
Page 37
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 37
Page 38
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 38
Page 39
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 39
Page 40
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 40
Page 41
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 41
Page 42
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 42
Page 43
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 43
Page 44
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 44
Page 45
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 45
Page 46
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 46
Page 47
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 47
Page 48
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 48
Page 49
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 49
Page 50
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 50
Page 51
Suggested Citation:"CHAPTER TWO Literature Review." National Academies of Sciences, Engineering, and Medicine. 2011. Speed Reduction Techniques for Rural High-to-Low Speed Transitions. Washington, DC: The National Academies Press. doi: 10.17226/22890.
×
Page 51

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.

6 CHAPTER TWO LITERATURE REVIEW Literature in the field of speed and speed management is abun- dant. However, the applicability of this literature to rural speed transition areas is limited. Even in the literature about rural traffic calming or traffic calming in small towns and villages, it is sometimes difficult to identify measures that are specific to transition zones as researchers do not always distinguish between speed maintaining measures (located in the village) and speed changing measures (located in the transition zone). The literature may be broken down into the follow- ing broad categories: effectiveness/evaluation studies and guidance/standards/recommended practices. The literature research focused on the former category but includes some important documents from the latter category as well. In this section, the effectiveness/evaluation studies are presented first, followed by the recommended practices and guidelines. The effectiveness studies are organized by measure of effec- tiveness (i.e., effect on operating speed or crash risk). The manuals and guidelines are organized by continent. EFFECTIVENESS STUDIES Speed Studies Van Houten and Van Houten (1987) Van Houten and Van Houten (1987) conducted an experi- ment to examine the efficacy of a roadside sign with the leg- end “BEGIN SLOWING HERE” on vehicle speeds at the start of a new speed zone. The study site was an approach to an urban area where a four-lane facility with a speed limit of 50 mph (80 km/h) transitioned to a two-lane facility with a speed limit of 30 mph (50 km/h). Speeds were recorded at the start of the 30 mph (50 km/h) speed zone. The test sign was a rectangular sign with a white back- ground and black lettering, and a black arrow that pointed down at 45 degrees to the roadway. The sign was placed 280 ft (86 m) upstream of the start of the 30 mph (50 km/h) speed zone. The standard advance warning sign for a reduced speed limit was also present, upstream of the experimental sign. The measures of effectiveness included the percentage of vehicles traveling above 35 mph (60 km/h). The study employed baseline speed measurements, followed by speed measurements with the subject sign in place, followed by speed measurements with the experimental sign removed, and finally a fourth measurement when the sign was reintro- duced. The results were as follows: • During the baseline measurement, the percentage of vehicles exceeding 35 mph (60 km/h) was 45.8%. • After implementing the sign for the first trial, the above percentage dropped to 37.4%. • During the interim baseline, the percentage of vehicles exceeding 35 mph (60 km/h) rebounded to pretest levels. • During the second implementation of the sign, the sub- ject percentage dropped to 33.8%. The researchers concluded that the effectiveness of the sign is the result of the clear direction provided to the motor- ist, whereas the standard reduced-speed-ahead sign simply provides nonspecific advice on where to slow down. Herrstedt et al. (1993) Herrstedt et al. (1993) examined and documented the effects of a wide variety of traffic calming measures on operating speeds in Denmark, France, and Germany. The examples that are presented refer to several locations where measures were used on a highway that runs “through a town,” with associated data provided. Table 1 presents examples where the text mentioned a “gateway” or “portal” of a highway tra- versing a town, suggesting some treatment at the high-to- low speed transition. The measures that were implemented clearly demon- strated effectiveness in managing speeds—an 11% reduc- tion in mean speeds, and a 15% reduction in the percentage of motorists traveling faster than 35 mph (60 km/h). How- ever, although the data showed impressive improvements in speed management, the information is of limited value. First, it is uncertain where the speed measurements were taken. Second, it is not possible to tease out the effects of the high-to-low speed transition treatment from the effects of the “in town” measures that were also employed. However, these examples underscore the effectiveness of providing a complete solution to the speed management issues of small towns—that is, providing a well-designed rural-urban tran- sition and carrying the measures through the town for a sus- tained speed reduction.

7 was that there needs to be a distinct and evident relationship between the change to the village speed limit and a change in the road environment. TABLE 2 TRAFFIC CALMING EFFECTS ON 85TH PERCENTILE SPEED IN THE UNITED KINGDOM Inbound at Gateway In Village No Measures in the Village Measures in the Village No Measures in the Village Measures in the Village a) Up to 3 mph (4.8 km/h) — a) Up to 4 mph (6.4 km/h) d) Up to 3 mph (4.8 km/h) b) Up to 7 mph (11.2 km/h) e) Up to 9 mph (14.4 km/h) b) Up to 5 mph (8.0 km/h) e) Up to 10 mph (16 km/h) c) Up to 10 mph (16 km/h) f) Up to 12 mph (19.2 km/h) c) Up to 6 mph (9.6 km/h) f) Up to 12 mph (20.8 km/h) a. Gateway signing, minor marking. b. Gateway signing with significant marking/colored surface and minor narrowing. c. Significant physical restriction at gateway. d. No gateway treatment. e. Gateway with significant signing/marking at it or in advance of it. f. Significant physical restriction at gateway. With respect to resident opinions, the five surveys revealed generally strong support for the speed management measures. One-third to one-half of the residents believed that TABLE 1 TRAFFIC CALMING EFFECTS FOR TRANSITION ZONES IN DENMARK, FRANCE, AND GERMANY Town Before After Change Speed Limit Speed Crash Rate+ Speed Limit Speed Crash Rate Speed Crash Rate mph (km/h) Mean mph (km/h) % > 35* Injury All mph (km/h) Mean mph (km/h) % > 35* Injury All Mean % > 35* Injury All Åbenrå 30 (50) 35 (58) 41 1.64 4.48 30 (50) 32 (54) 21 1.21 2.40 –7% 20 –26% – 46% Vinderup 35 (60) 31 (51) 20 1.87 3.46 25 (40) 25 (42) 3 0.56 1.83 –18% 17 –70% – 47% Tinglev 35 (60) 44 (73) 92 1.48 3.11 30 (50) 35 (58) 80 1.40 2.46 –21% 12 –5% –21% Vipperød 35 (60) — — 1.02 1.17 30 (50) — — 0.50 0.76 — — –51% –35% Tarm 30 (50) 23 (39) — 0.60 0.80 20 (30) 22 (36) 0.50 1.00 –8% — –17% 25% Skægkær 35 (60) 38 (63) 26 1.32 1.93 30 (50) 35 (59) 14 0.76 1.03 –6% 12 – 42% – 46% Nøvling 30 (50) 32 (54) — 0.70 1.00 25 (40) 29 (49) — 0.00 0.50 –9% — –100% –50% Arnage (France) — — — — 0.63 25 (40) — — — 0.21 — — — –67% Average –11% 15 – 44% –36% *% > 35 = Percentage of vehicles traveling faster than 35 mph (60 km/h). +Crash rates are reported as annual crashes per million vehicles. County Surveyors’ Society (1994a) The County Surveyors’ Society (1994a) of the U.K. reported the results of an investigation into the costs, benefits, and effectiveness of measures for managing vehicular speeds in villages. The analysis consisted of before-after studies in 24 villages, using 85th percentile speed and resident opinion surveys as measures of effectiveness. The number of sites was limited by the project budget, and sites were selected to provide a range of village size, road classification, traffic volume, and calming measures. Of the 24 sites, 11 sites had traffic calming on the approach to the village, four sites had traffic calming within the village, and nine sites had traffic calming on the approach and within the village. Ten of the sites were on roads of “high” priority. The researchers did not collect crash data, as the small numbers available in the short after period were thought to be unreliable indicators of effectiveness. Instead, they col- lected data on speed, volume, and public opinion as mea- sures of effectiveness. Speed and volume measurements were taken continuously for at least 1 week in the before and after periods; the after period was generally 1 or 2 months after installation. The results indicated that small expenditures (i.e., simple gateways) provide small reductions in speed [about 3 mph (5 km/h) or less], whereas significant speed reductions [6 mph to 10 mph (10 to 16 km/h)] accompanied more compre- hensive/expensive treatments (i.e., gateways and measures in the village). Table 2 shows the results presented by the researchers. One of the main conclusions on effectiveness

8 traffic had slowed down. The majority of residents wanted more to be done. The study concluded that there is no inexpensive and sin- gular device that will have a significant and lasting effect on speeds, and that traffic calming measures and other engi- neering changes to encourage speed compliance are appro- priate for villages. County Surveyors’ Society (1994b) The County Surveyors’ Society (1994b) of the U.K. also col- lected and summarized data on numerous traffic calming installations in the U.K. The intentions of this effort were to provide practitioners with a snapshot of traffic calming in the U.K., and to detail a number of case studies so that prac- titioners might better understand what plans and measures work and do not work in different situations. In all, 85 case studies were documented in the report, with 25 cases classi- fied as traffic calming on rural roads/areas. Table 3 summarizes the effects of traffic calming on the 25 rural case studies. It is noted that all of the traffic calming was implemented for one or more of the following objec- tives: reduce speed, reduce crash occurrence, and reduce through traffic. However, the measures implemented at each of the sites were not implemented for the same reasons and are not of the same form (i.e., humps versus narrowings). Therefore, this analysis is necessarily general. This research indicated that traffic calming on rural roads has a positive impact on speeds. The average speed reduction was almost 5.4 mph (9 km/h). Twenty-three of the 25 sites reported before-after speed data, and all sites experienced a speed reduction following the introduction of traffic calm- ing. However, the speed limit at each site was not reported, so it is perhaps more appropriate to indicate that there was a 21% reduction in speeds. The data contained in the report do not indicate whether the rural road is a primary route or a local road. If the “before” traffic volume is used as a surrogate for road classi- fication, for the eight sites with daily traffic volumes greater than 10,000 (which are assumed to be primary routes or arte- rial roads) the average speed was reduced by 4.5 mph (7.5 km/h) or 16%. These data indicate that traffic calming on the higher volume roads was effective at speed control, but not as effective as traffic calming placed on lower volume roads. Pyne et al. (1995) Pyne et al. (1995) conducted driving simulator experiments to test the effectiveness of a variety of measures for reduc- ing speeds on undivided, rural, arterial roads, including on the approaches to villages. The measures of effectiveness included effect on speed (at the start, in the middle, and at the end of the village) and lateral position. The experiment was conducted in two phases: Phase 1 tested a variety of transition area measures, and Phase 2 examined variations on and combinations of the most effective treatments. Figure 2 shows the high-to-low speed transition measures evaluated in Phase 1, and Figure 3 shows the treatments evaluated in Phase 2. Table 4 shows the range of speed changes caused by the various treatments examined in the Phase 1 research. The changes in the mean and 85th percentile speed were relatively small for all measures tested. Using a system for ranking treatments based on a combination of mean speed and 85th percentile speed, Pyne et al. determined that at the start of the village transverse lines were the most effec- tive measure, followed by Wundt illusion, countdown speed signs, central hatching with no road narrows sign, the chi- cane, and the speed limit on the road surface, in that order. The only treatment that produced speeds that were statisti- cally significantly different from speeds in the control run was the transverse lines. Table 5 shows the range of speed changes caused by the various treatments combined and examined in the Phase 2 research. The combined treatments used in the Phase 2 research produced lower speeds than those treatments used in Phase 1. Again, using the ranking system that combines effectiveness in reducing mean speed and 85th percentile speed, the mea- sures that were most effective at the start of the village are V2.17, V2.19, and V2.20. The hazard marker posts (V2.03) and the speed limit on the road surface (V2.05) did not have a significant effect on speed at any of the three points. Also, the chicane without hatching (V2.08) produced speeds sig- nificantly different from the control at all three locations, but the chicane with hatching (V2.07) did not. The researchers posit that this is probably because the travel lane was less obvious without hatching. Pyne et al. concluded that the most effective combination of treatments for the high-to-low speed transition zones was the chicane without hatching, associated with countdown speed limit signs on the approach to the village, followed by transverse markings throughout the village. This combina- tion of measures produced the “best” reduction in operating speeds at all three locations through the village. Barker and Helliar-Symons (1997) Barker and Helliar-Symons (1997) examined the effects of two different low-cost engineering measures designed to alert drivers to a downstream lower speed limit, in hopes of effecting greater compliance with speed limits. The first

9 TABLE 3 THE EFFECTS OF TRAFFIC CALMING ON RURAL ROADS IN THE UNITED KINGDOM Case Descriptions Traffic (ADT) Speed in mph (km/h) Crashes (injury crashes/year) Crash Rates (annual injury crashes/TVD)* Before After Before After Change Before After Before After 1 Narrowings, temporary material on a trial basis 4,000 5,000 47 (75) 35 (56) –12 (–19) 1.8 1 0.45 0.20 2 Gateway features, traffic islands, markings, and speed roundels 2,000 2,000 44 (70) 36 (58) –8 (–13) 0.7 1.5 0.35 0.75 3 Gateway signing 6,900 6,900 40 (64) 38 (61) –2 (–3) 2.3 2 0.33 0.29 4 Speed camera with a gate- way added later 17,500 17,500 47 (75) 41 (66) –6 (–10) 1.3 1 0.07 0.06 5 40 km/h zone — — — — — — — — — 6 Give way narrowings, series of round and flat top humps, bus route 1,700 1,500 37 (59) 27 (43) –10 (–16) 2.7 0.3 1.59 0.20 7 Traffic islands and chicanes with bypasses for cyclists 10,600 11,300 47 (75) 41 (66) –6 (–10) 5.7 2.7 0.54 0.24 8 Gateway effect with central island located on entry to village 2,900 2,900 38 (61) 35 (56) –3 (–5) 0.3 0 0.10 0.00 9 Narrowings, speed tables, throttle humps, mini roundabout 7,300 6,850 35 (56) 24 (38) –11 (–18) 2.8 1 0.38 0.15 10 Narrowings, castellated sur- facing, gateways 2,000 1,800 42 (67) 39 (62) –3 (–5) 0.7 0 0.35 0.00 11 Mini roundabout, road humps, narrowing 7,100 7,100 31 (50) 16 (26) –15 (–24) 2 — — — 12 Mini roundabouts, raised speed tables, narrowings 7,100 7,100 31 (50) 22 (35) –9 (–14) 2 — — — 13 Gateways, refuges, chicanes, mini, closure 10,000 10,000 47 (75) 34 (54) –13 (–21) 3.7 0 0.37 0.00 14 Reduced number of lanes, refuges, cycle lanes 9,400 9,400 60 (96) 53 (85) –7 (–11) 5 2.4 0.53 0.26 15 Flat top humps, minis, ref- uges, landscaping 8,500 9,800 45 (72) 27 (43) –18 (–29) 5 0 0.59 0.00 16 Gateways, chicanes, textured surfaces, and street lighting 13,500 13,500 40 (64) 35 (56) –5 (–8) 3 2.4 0.22 0.18 17 Narrowings, gateways, light- ing improvements, signs, and markings 10,200 — 51 (82) 44 (70) –7 (–11) 10.7 8 — — 18 Road humps, mini roundabouts 8,400 6,300 37 (59) 28 (45) –9 (–14) 4 1.7 0.48 0.27 19 Gateways, refuges, rumble strips, signs, markings 550 550 49 (78) 41 (66) –8 (–13) 3.7 1 6.73 1.82 20 Jiggle bars — — 48 (77) 39 (62) –9 (–14) — — — — 21 Narrowing, cycle track, hatched central reserve 11,350 11,350 56 (90) 44 (70) –12 (–19) 4.3 4.6 0.38 0.41 22 Road humps, narrowing, chicane 5,000 3,600 — — — 0.3 0 0.06 0.00 23 Road humps, mini round- abouts, 30 mph speed limit 1,300 1,080 40 (64) 25 (40) –15 (–24) 0.3 0 0.23 0.00 24 Gateways, mini roundabouts, road humps 10,100 10,100 42 (67) 39 (63) –3 (–5) 1.3 0.8 0.13 0.08 25 Edge and center markings and road studs to narrow traveled way 10,100 10,100 40 (64) 32 (51) –8 (–13) 1.3 0 0.13 0.00 Average –8.7 (–14) 0.70 0.24 *TVD = thousand vehicles per day.

10 FIGURE 2 Transition zone treatments evaluated in Phase 1 research [Source: Pyne et al. (1995)]. FIGURE 3 Transition zone treatments evaluated in Phase 2 research [Source: Pyne et al. (1995)].

11 • Roundels used on the approach to 40 mph (65 km/h) speed limits produced a 3 mph (5 km/h) mean speed reduction. • Roundels used on the approaches to 30 mph (50 km/h) speed limits did not have a significant effect on mean speeds. • The countdown speed signs did not have a significant effect on mean speeds. • Owing to the small sample size, the effects of either measure on crashes are inconclusive. Farmer et al. (1998) In a trial of speed feedback signs placed at entrances to six rural villages in Norfolk, England, Farmer et al. (1998) used mean and 85th percentile speeds recorded at the signs and in the center of the village to determine effectiveness. Speeds were assessed 1 month, 6 months, and 12 months after installation. The results show that both the mean and 85th percentile speeds show a sustained reduction at the signs, with the mean speed reduction over all sites of 4.3 mph (6.9 km/h). A smaller effect was noticed in the center of the village. Berger and Linauer (1998) Berger and Linauer (1998) examined the effects of gateway treatments on five two-lane roads at the transition from rural to urban areas in Austria. The treatments were raised islands placed in between the two travel lanes that were supple- mented with appropriate signs and markings. Four different island designs were used, each intending to provide some degree of road narrowing (by dividing the two lanes) and deflection. The island shapes are shown in Figure 5. The study methodology was a before-after analysis of mean, 85th percentile, and maximum observed speeds recorded near the town sign (which presumably is proximate to the island). The results are shown in Table 6. Not surprisingly, the incidence of speed reduction increased as the deflection increased. These results were used to develop the following regression models relating the island shape and expected speed: Metric: V85 = 14.797Ln(L/2d) + 19.779 R2 = 0.9098 Vm = 12.907Ln(L/2d) + 17.753 R2 = 0.9693 Where: V85 = 85th percentile speed (km/h) Vm = mean speed (km/h) L = length of island + length of both tapers (m) d = lateral deflection of lane (m) measure was a speed limit roundel marking—elongated cir- cles with the speed limit in the center, laid in the white mark- ings on the road surface in the middle of the travel lane. The second measure was countdown signs—a standard speed limit sign together with three, two, and one black diagonal bars on a white background mounted below the speed limit sign (see Figure 4) and respectively placed 985, 655, and 330 ft (300, 200, and 100 m) from the start of the reduced speed limit. TABLE 4 RANGE OF SPEED REDUCTIONS IN SIMULATED TRANSITION ZONES (PHASE 1) Mean Speed mph (km/h) 85th Percentile Speed mph (km/h) Start of Village –2.3 to +1.2 (–3.7 to +1.9) –3.0 to +2.1 (–4.8 to +3.4) Middle of Village –1.4 to +0.3 (–2.2 to +0.5) –2.8 to +1.6 (–4.5 to +2.6) End of Village –1.7 to +0.1 (–2.7 to +0.2) –9.4 to +5.0 (–15.0 to +8.0) TABLE 5 RANGE OF SPEED REDUCTIONS IN SIMULATED TRANSITION ZONES (PHASE 2) Mean speed mph (km/h) 85th Percentile Speed mph (km/h) Start of Village –4.2 to –1.1 (–6.7 to +1.8) –7.2 to –2.9 (–11.5 to +4.6) Middle of Village –3.0 to +0.5 (–4.8 to +0.8) –4.1 to +1.1 (–6.6 to +1.8) End of Village –3.4 to –0.6 (–5.4 to +1.0) –6.7 to +2.0 (–10.7 to +3.2) FIGURE 4 Countdown speed signs [Source: Barker and Helliar-Symons (1997)]. The roundels were field-tested at 12 villages, and the countdown signs were field-tested at five villages. After studies were conducted from 1 week to 12 months after installation. The research conclusions were as follows:

12 • Requiring motorists to reduce speed just prior to leav- ing the urban area (which may promote a uniform, lower speed throughout the urban area), and • Preventing motorists entering the urban area from using the opposing lane (anecdotal observations). U.K. Department of the Environment, Transport and the Regions (undated) Subsequent to a 1994 report on speed control in villages, the U.K. Department of the Environment, Transport and the Regions (undated) commissioned a study on comprehensive traffic calming schemes in villages, particularly on trunk (i.e., major) roads. To qualify for the study, sites required FIGURE 5 Raised islands at city limits [Source: Berger and Linauer (1998)]. TABLE 6 SPEED EFFECTS OF GATEWAY TREATMENTS IN AUSTRIA Speed* Period Island Design Type (see Figure 5) 1 2 3 4 5 Mean Before 32.4 (54.0) 34.8 (58.0) 36.0 (60.0) 39.0 (65.0) 39.0 (65.0) After 32.5 (54.1) 29.0 (48.4) 26.5 (44.1) 28.3 (47.2) 24.1 (40.1) Change 0% –17% –27% –27% –38% 85th Percentile Before 37.2 (62.0) 40.2 (67.0) 42.0 (70.0) 45.6 (76.0) 46.2 (77.0) After 36.6 (61.0) 32.7 (54.5) 30.3 (50.5) 33.1 (55.2) 26.8 (44.6) Change –2% –19% –28% –27% –42% Maximum Before 42.0 (70.0) 52.8 (88.0) 51.6 (86.0) 57.0 (95.0) 58.2 (97.0) After 45.7 (76.2) 35.6 (59.3) 33.7 (56.1) 39.5 (65.8) 28.1 (46.9) Change +9% –33% –35% –31% –52% *All speeds in mph (km/h). U.S. Customary: V85 = 9.194Ln(L/2d) + 12.290 Vm = 8.020Ln(L/2d) + 11.031 Where: V85 = 85th Percentile speed (mph) Vm = mean speed (mph) L = length of island + length of both tapers (ft) d = lateral deflection of lane (ft) The authors note that Island Design No. 3 has additional advantages of—

13 • Speed reductions were achieved at all of the gateways. • The mean after speeds were close to the speed limits, but the 85th percentile speeds remained considerably above the speed limit. Alley (2000) Alley (2000) assessed the effectiveness of different gateway treatments typically found on New Zealand roads using a driving simulator, with mean speed as the measure of effec- tiveness. He categorized the treatments as narrowings, raised traffic islands, and oversized speed restriction signs. The report also specified that the implementation of these measures across New Zealand has met with varying levels of success, including at least one situation where operating speeds actually increased. Alley used a human factors approach to developing tran- sition zone measures and hypothesized that the effectiveness of the speed reduction is related to an increase in the local edge rate presented to a motorist passing through the tran- sition zone. The local edge rate is the number of elements (edges) that pass a stable reference point within the optical field in 1 second, such as the rate at which utility poles pass the A-pillar as one drives down the road. Translating this hypothesis to transition zone treatments, Alley proposed that the more elements that are introduced into the transition zone, the more effective the speed reduction, because the increased edge rate will increase the perception of speed. Two experiments were undertaken to test the hypothesis. (The Alley research actually includes four experiments, but the first two experiments were used to validate the driving simula- traffic volumes of at least 8,000 vehicles per day, of which 10% were heavy vehicles. The objective of the study was to assess whether traffic calming could be implemented that would reduce the 85th percentile speed of vehicles to the speed limit. Nine sites were reviewed, with six of the sites receiving a speed limit reduction concurrently with the installation of traffic calming. Table 7 presents the site characteristics. Speed studies were conducted before, 1 month after, and 12 months after the installation of traffic calming measures. Table 8 shows the effects of the traffic calming on mean and 85th percentile speeds at the gateways and within the village. The following conclusions were reached: TABLE 7 CHARACTERISTICS OF VILLAGE TRAFFIC CALMING ON MAJOR ROADS Site Daily Volume Proportion of Heavy Vehicles (%) Village Population Speed Limit, mph (km/h) Original New 1 11,500 18 400 60 (100) 40 (65) 2 5,500 10 5,400 30 (50) 20 (30) 3 9,000 15 1,900 40 (65) 30 (50) 4 9,000 16 350 30 (50) 30 (50) 5 8,000 16 1,200 40 (65) 30 (50) 6 17,000 10 3,900 30 (50) 30 (50) 7 17,000 15 150 60 (100) 40 (65) 8 13,000 20 2,200 30 (50) 30 (50) 9 16,500 18 3,370 50 (80) 40 (65) TABLE 8 MEAN AND 85TH PERCENTILE SPEEDS—BEFORE AND AFTER SCHEME INSTALLATION (MPH) Site New Speed Limit (mph) N/W Gateway (inbound) S/E Gateway (inbound) In Village (mean of both directions) Before After1* After2* Before After1 After2 Before After1 After2 Mean 85th %ile Mean 85th %ile Mean 85th %ile Mean 85th %ile Mean 85th %ile Mean 85th %ile Mean 85th %ile Mean 85th %ile Mean 85th %ile 1 40 43 48 40 44 46 52 42 47 40 45 38 42 2 20 33 39 23 29 25 30 36 43 24 32 25 31 30 36 21 27 22 28 3 30 41 49 33 39 33 40 42 49 33 40 33 41 34 39 28 32 27 30 4 30 41 48 28 33 33 39 39 46 31 36 32 37 31 36 29 32 28 31 5 30 38 43 31 37 31 35 37 42 34 40 32 37 39 44 33 37 34 39 6 30 39 45 39 49 37 44 43 49 39 45 37 44 35 41 36 41 34 39 7 40 48 56 42 49 40 50 53 63 48 53 50 55 51 58 41 46 41 47 8 30 46 53 37 44 36 41 42 49 33 41 31 37 39 44 32 37 31 34 9 40 42 48 37 41 38 43 46 52 40 45 38 43 41 46 38 43 39 44 Avg. 41.2 47.7 34.1 40.3 42.7 49.4 34.8 40.6 37.8 43.2 32.0 36.5 Change 17% 16% 19% 18% 15% 16% * “After1” are speeds measured about 1 month after installation, and “After2” are speeds measured about 12 months after installation. %ile = percentile.

14 FIGURE 6 Effect of perceptual measures in New Zealand [Source: Alley (2000)]. Phase 1 Phase 2 2. Oversized signs placed on both sides of the road; 3. The same as Alternative 2, but reducing the width between the curbs from 28.1 ft to 22.8 ft (8.56 m to 6.94 m); 4. The same as Alternative 2, but with a different sign shape; and 5. The same as Alternative 4, but reducing the width between the curbs from 28.1 ft to 22.8 ft (8.56 m to 6.94 m). In the second experiment, the researcher removed infor- mation on the speed limit from the scenarios to investigate the effects of gateway shape and size and lateral placement on operating speed in the vicinity of the transition zone. The same threshold alternatives from the first experiment were examined. Figure 6 also shows the results of the simulator testing on speeds upstream, downstream, and at the gateways. Alternatives 3, 4, and 5 showed small but significant speed reductions [less than 1.5 mph (2.5 km/h)] at the gateway. However, at 820 ft, 1,640 ft, and 2,460 ft (250 m, 500 m, and 750 m) downstream of the gateway, speeds measured tor as a viable representation of real life driving.) In the first experiment, 31 participants of varying age and driving expe- rience were asked to drive through five different urban/rural thresholds, each treated with different engineering measures on a two-lane rural road. The five thresholds are as follows: 1. Conventional speed-restriction signs placed on the right and left sides of the road, denoting the reduced speed limit of 45 mph (70 km/h). 2. All measures in Alternative 1, plus painted hatching on the paved shoulder, and a “half-pinch” (the edge lines on both sides of the road are deflected toward the centerline to slightly narrow the travel lanes). 3. All measures in Alternative 2, plus a narrow painted median to further reduce the lane width. 4. All measures in Alternative 3, plus roadside traffic islands consisting of raised curbs on the previously paved shoulder. 5. All measures in Alternative 4, except the speed- restriction signs are replaced with oversized speed- restriction signs. The presentation order of the five alternatives was random. The urban area had a 45 mph (70 km/h) speed limit and was 0.6 mile (one km) long. Speed measurements were reported for five locations: 820 ft (250 m) upstream of the threshold (in the rural zone), at the threshold, and 820 ft (250 m), 1,640 ft (500 m), and 2,460 ft (750 m) downstream of the threshold (in the urban area). The results are shown in Figure 6. There was no significant difference between the approach speeds at 820 ft (250 m) upstream of the threshold for all five alternatives, demonstrating that the participants were approaching each threshold at the same speed. At the thresh- old and 820 ft (250 m) downstream of the threshold, Alterna- tive 5 produced the lowest speeds of any of the alternatives (as expected). The alternative that produced the next lowest speeds at these locations is Alternative 1: the conventional speed signs. This aforementioned result was not expected, as the hypothesis suggests that the recorded speeds should decrease as the alternative number increases. It is interesting to note that the full reduction in speed is not realized until 1,640 ft (500 m) past the gateway/threshold. In the second experiment, the researcher removed infor- mation concerning the speed limit from the scenarios to investigate the effects of gateway shape and size and lateral placement on operating speed in the vicinity of the transition zone. The threshold alternatives examined were: 1. One oversized sign placed on the right side of the road;

15 the SPEED LIMIT SIGN (R2-1), that was blanked out and activated by motorists exceeding the speed limit. Some of the roundels were supplemented with flashing beacons. All of the VAS were located 66 ft to 164 ft (20 m to 50 m) downstream of the beginning of the speed limit change (i.e., within the village speed limit). With respect to speed data, the first after data were col- lected about 1 month after the VAS became operational. To establish whether any speed changes were sustained, fur- ther data collections were made 1 year after the VAS were placed in operation, and for some sites additional data were collected 3 years after installation. A minimum of 7 days’ continuous data were collected at each site in both the before and after periods. Speeds were collected at or immediately downstream of the VAS. The primary measure of effective- ness for speed is the percentage of vehicles exceeding the speed limit. Table 9 shows the study results. VAS on the approach to reduced speed limits for villages appeared to be very effective in reducing speeds; in particu- lar, they were capable of reducing the number of drivers who exceed the speed limit, without the need for enforcement such as safety cameras. Moreover, there was no evidence that drivers became less responsive to the signs over time, even over 3 years. Hildebrand et al. (2004) Hildebrand et al. (2004) examined the effectiveness of tran- sitional speed zones in rural areas in New Brunswick, Can- ada. The intent of the study was to determine if tempering large changes in the posted speed limit [25 mph (40 km/h)] with an intermediate posted speed limit was more effective than an abrupt change alone. Six sites with transitional speed limits were compared with seven sites without transitional speed limits, using mean speed, percentage exceeding the speed limit, and speed dispersion as measures of effective- ness. The study concluded that the transitional speed limits did not have any significant impact on speed. Agustsson (2005) Agustsson (2005) reported on the effectiveness of “envi- ronmentally friendly through-roads,” which are streets where traffic is managed by using different forms of speed- reducing measures. Twenty-one of these newly developed roads were implemented to reduce speed, increase safety, and improve road design. The measures used included gates, roadside reservations, medians, roundabouts, raised areas, changes in road surface, road markings, signing, lighting, road closures, rumble strips, and bicycle lanes. The average length of the through-road is about 0.6 mile or 1 km (it is not clear if this was the total length of the road or the length of the road that was treated, although it is expected that it is the latter). for all threshold designs were not significantly different than the upstream speeds. In comparing the measured speeds between alternatives, there were no consistent patterns or trends to suggest any of the gateway designs was more effec- tive than the others. However, the removal of the urban speed limit information seems to have eliminated the gateways’ downstream effects on speed. Although the order of presentation for threshold alterna- tives was counterbalanced in both experiments, the research- ers examined and found that the likelihood of a participant decelerating for the gateway decreased as the presentation order increased. In other words, repeated exposure to gate- ways seems to lessen the desired effect of the gateway on speed. This habituation effect is consistent with an unref- erenced study cited by Alley, where a gateway at the high- to-low speed transition on a state highway in New Zealand achieved a 6.2 mph (10.4 km/h) reduction in the average speed 6 months after installation, but only a 3.3 mph (5.2 km/h) reduction 12 months after installation. Winnett and Wheeler (2002) Winnett and Wheeler (2002) conducted an evaluation of the effectiveness of vehicle-activated signs (VAS) in reduc- ing speeds and crashes in the U.K. One type of VAS evalu- ated was a dynamic speed limit sign (a “speed roundel”) for speed limit changes, employed mainly at village sites on rural, undivided roads (see Figure 7). FIGURE 7 Vehicle-activated speed signs [Source: Winnett and Wheeler (2002)]. Sites were selected for VAS implementation if they had a recent history of crashes in which inappropriate speed was a contributory factor or a record of excessive speed for the conditions was believed to be a potential problem. Sites selected for evaluation also required suitable sight lines to the VAS by the approaching driver, as well as traditional traffic control devices (i.e., fixed signs and markings) that were in compliance with the applicable standards. The VAS implemented was a speed roundel, the U.K. equivalent to

16 Department for Transport (2005) The U.K. Department for Transport (2005) developed and tested a quieter alternative to conventional transverse rumble strips known as a rumblewave surface. The rumblewave sur- face delivers the auditory and tactile stimulus to vehicle occu- pants in an attempt to elicit a slower travel speed, but does not generate as much ground vibration or noise for the surround- ing community. The recommended profile for a rumblewave is shown in Figure 9. A profile with a wave length of 1.1 ft (0.35 m) and a wave height of about one quarter of an inch (6 or 7 mm) is recommended, as this profile produces the largest increases in interior noise and vibration for a range of vehicle types but creates little increase in exterior noise levels. The final layout of the surface would be determined by local conditions. Rumblewave surfaces were piloted at seven locations, including a high-to-low speed transition, as shown in Table 10. The measures of effectiveness for the pilot program were mean and 85th percentile speeds, and personal injury crashes. The speed results are shown in Table 11 and indicate that both the mean and the 85th percentile speeds exhibited reductions at all pilot locations. In the worst case, the reduction was only 1%; in the best case, reductions of 5% and 6% were measured. No statistical tests of significance were presented. TABLE 9 SUMMARY OF SPEED REDUCTIONS AT ROUNDEL SIGNS Speed Limit No. of Sites Mean Speed, mph (km/h) Change in Percent Exceeding Speed Limit Average Before Average After Average Change Maximum for Group Minimum for Group 30 mph (50 km/h) 17 34.5 (55.2) 30.0 (48.0) –4.5 (–7.2) –51 –15 40 mph (65 km/h) 5 38.2 (61.1) 35.1 (56.2) –3.1 (–5.0) –12 –1 20 mph (30 km/h) where a 30 mph (50 km/h) speed limit existed before the trial 6 31.1 (49.8) 24.9 (39.8) –6.2 (–9.9) –58 –38 30 mph (50 km/h) where a 40 mph (65 km/h) speed limit existed before the trial 7 39.7 (63.5) 30.8 (49.3) –8.9 (–14.2) –80 –50 50 mph (80 km/h) 1 52.0 (83.2) 47.9 (76.6) –4.1 (–6.6) –51 –15 All 36 39.1 (62.6) 33.7 (53.9) –5.4 (–8.6) –80 –1 The evaluation indicated that the average speed was reduced in the near-term and remained lower by 6 mph (10 km/h), a 17% reduction. However, Agustsson noted that there was a large variation in effectiveness depending on where and which type of traffic measure was used. The effect of the environmentally friendly through-roads on the speed pro- file was equally impressive (see Figure 8). The percentage of motorists traveling in excess of the speed limit was more than halved, from 75% to 36%. FIGURE 8 Effect of environmentally adapting roads on speed profile [Source: Agustsson (2005)]. FIGURE 9 Rumblewave surface [Source: Department for Transport (2005)].

17 ter emphasize the start of the village. Speed measurements taken proximate to the gateways revealed that mean speeds were reduced by 4 to 8 mph (6.5 to 13 km/h) to 37 mph (59 km/h), with similar decreases in the 85th percentile speeds. Forbes (2006) Forbes (2006) conducted a retrospective, observational before-after study of the safety effects of traffic manage- ment in rural settlements located on rural arterial roads. The study involved a literature review and some original research using data collected from several small developed areas in Ontario, Canada. The new research documented the types of measures implemented and their efficacy related to crash risk and speed management. Twelve treatment sites were selected by convenience. Speed of travel was evalu- TABLE 10 RUMBLEWAVE PILOT LOCATIONS Site Problem ADT Speed Limit mph (km/h) Rumblewave Sections No. of Sections Length feet (m) Spacing feet (m) 1 Horizontal curve 4,200 30 (50) 1 66 (20) N/A 2 40 to 30 mph speed limit change on a semi- urban, major road 9,200 40–30 (65–50) 1 initially, 1 8 months later 72 (22) 328 (100) 3 Low speed limit with no other calming 1,300 20 (30) 4 39, 56, 2x72 (12, 17, 2x22) 184–302 (56–92)* 4 High vehicular, pedestrian, and cyclist vol- umes, and a high crash rate 21,500 30 (50) 9 7x39, 49, 59 (7x12, 15, 18) 0–643 (0–196)* 5 Residential area with schools and shops, shortcutting, and a high crash rate 4,100 30 (50) 6 2x33, 39, 49, 2x56 (2x10, 12, 15, 2x17) 272–686 (83–209)* 6 High to low speed rural transition 3,600 60–40 (100–65) 2 72 (22) 164 (50) 7 Residential area close to road, high crash rate 9,600 30 (50) 5 26, 2x39, 2x72 (8, 2x12, 2x22) 302–804 (92–245)* *Depending on intersections. TABLE 11 EFFECTS OF RUMBLEWAVE SURFACES ON OPERATING SPEEDS Site Mean Speed 85th Percentile Speed Before mph (km/h) After mph (km/h) Change (%) Before mph (km/h) After mph (km/h) Change (%) 1 31.0 (49.6) 30.6 (49.0) 1 36.3 (58.1) 35.8 (57.3) 1 2 37.1 (59.4) 36.3 (58.1) 35.7 (57.1) 2 4 42.1 (67.4) 41.5 (66.4) 41.3 (66.1) 1 2 3 26.8 (42.9) 25.1 (40.2) 6 33.5 (53.6) 32.4 (51.8) 3 4 29.5 (47.2) 27.6 (44.2) 6 34.7 (55.5) 32.8 (52.5) 5 5 28.5 (45.6) 27.1 (43.4) 5 34.2 (54.7) 32.7 (52.3) 4 6 38.9 (62.2) 38.2 (61.1) 2 44.4 (71.0) 44.8 (71.7) 1 7 32.1 (51.4) 31.9 (51.0) 1 37.3 (59.7) 36.7 (58.7) 2 Kennedy (2005) As part of a larger study concerning psychological traffic calming, Kennedy (2005) experimented with village traf- fic calming in a small village in the U.K. The main street through the village was a wide and straight two-lane road with just under 2,000 vehicles per day, a 40 mph (65 km/h) speed limit, and a crash-free history (the rural speed limit was not mentioned). The measures applied at the transition zone were stone gateways with a village name plate and speed limit signs showing a new speed limit of 30 mph (50 km/h), build-outs at the side of road preceded by hatching, and removal of the directional dividing line. Complementary measures were installed throughout the village to assist in managing speeds downstream of the gateway. Additionally, the gateway was moved closer to the built-up area, to bet-

18 operating speeds. This research only marginally relates to speed management in rural areas through two reported projects. First, the village of Starston in the U.K. demon- strated that removal of the directional dividing line resulted in a 7 mph (11 km/h) speed reduction. Second, a similar line removal plan for 30 mph (50 km/h) zones in Wiltshire, Eng- land, resulted in a 5% reduction in travel speeds. Charlton and Baas (2006) The goal of a study by Charlton and Baas (2006) was to identify research findings that could be used to develop an approach to speed management for New Zealand roads that is commensurate with the self-explaining road concept. The literature review identified two distinct types of speed man- agement treatments: • Measures that are implemented at a location where a change in speed occurs; and • Measures that encourage drivers to maintain an appro- priate speed within the zone. The former type of speed management incorporates rural high-to-low speed transition zones and is of interest in this synthesis. Charlton and Baas categorized engineer- ing measures intended to effect a speed change into three types: gateways, physical measures, and visual measures. The synthesized information contained in the Charlton and Baas report is reproduced in Figure 10. With respect to gate- ways, the researchers noted that the effects are somewhat variable, but reiterated the desire for a downstream setting that reinforces the need for the lower speed. In fact, they stated that when the downstream effects of various gateways are compared, all types produce speed reductions of 2 to 3 mph (3 to 5 km/h). However, when the gateway is combined with in-village road narrowing, the downstream effects are more pronounced—about a 9 mph (14 km/h) reduction at the gateway and 10 mph (16 km/h) in the village. The researchers drew some firm conclusions from their review of the literature and their survey of experts and knowledgeable practitioners concerning speed transition zones: • The location of the measure is directly related to effec- tiveness. It is better to site the treatments at a loca- tion that appears to warrant a speed change than at a jurisdictional boundary or other seemingly arbitrary location. • If the measures are not accompanied by downstream changes in road conditions, increases in urban density, or similar conditions, the speed reductions produced by the gateway may dissipate within 820 ft (250 m). • Visual treatments (e.g., signing and perceptual narrow- ing) appear to be more appropriate for higher-speed environments, whereas physical treatments produc- ated using 85th percentile speed, mean speed, percentage of vehicles exceeding the speed limit, and speed variance as the metrics. Table 12 shows the sites used in the analysis and the data available at each site. TABLE 12 SITES SELECTED FOR ANALYSIS Settlement Treatment Data Available Ashburn Edge line Speed Claremont Edge line Speed Epsom No passing zone Speed Sandford CSZ, no parking in front of school, and gateway signing Speed Kilbride Pavement narrowing through mark- ings, roadside posts, sections of tex- tured pavement Crash Alberton Speed limit reduction Crash Kirkwall Speed limit reduction Crash Belfoun- tain Lanes narrowed by pavement mark- ings, and “SLOW” marked on pave- ment in 5 locations Speed Palgrave Gateway signing, and some minor signing Speed Terracotta Lanes narrowed by pavement mark- ings, “road watch” program initiated, and raised reflective markers added Speed Laskay Gateway signing and posted speed reduction Crash & speed Udora Gateway signing Crash & speed At the nine sites where speed data were available, the traf- fic management measures were successful in reducing mean and 85th percentile speeds, and increasing speed compliance, with a few explainable exceptions. The average reduction in the 85th percentile speed was 3.6 mph (6 km/h). Overall, the results indicated that traffic management in rural settle- ments in Ontario was effective in managing speed. However, the results should be viewed with some caution as the sample sizes were small, the treatments applied at each site varied considerably, and there may have been co-interventions that contributed to the improvements (i.e., increased awareness and enforcement that are unaccounted for in the analysis). Quimby and Castle (2006) Quimby and Castle (2006) investigated the effects of simpli- fied streetscapes on operating speeds. As the name implies, simplified streetscapes are projects that reduce visual clutter and complexity, usually through the removal of unneces- sary street furniture and signs. In the most radical projects, the simplified streetscape includes the removal of signs and markings as well as physical separations between modes of transport in order to create driver uncertainty and reduce

19 Data were collected at each treatment site at a location upstream of the dynamic display sign, at a point where the sign is not visible, and adjacent to the dynamic sign where the lowered speed limit begins. After results were collected 1 week, 2 months, 7 months, and 1 year after installation. The study sites with the dynamic displays experienced reduc- tions in the 85th percentile speed averaging 6.9 mph (11.0 km/h), which was sustained over 12 months. Knapp and Rosales (2007) An increasingly popular form of traffic management that may be useful for rural settlements located on multilane roads is “road diets,” or lane reductions. Typically, road diets involve converting a four-lane road to a two-lane road, with a center ing deflection or discomfort (e.g., chicanes, and speed humps) may be unsuitable for speeds higher than 20 to 30 mph (30 to 50 km/h). Sandberg et al. (undated) Sandberg et al. (undated) studied the long-term effects of permanent dynamic speed feedback displays on operating speed where rural highways transition into urbanized areas. The study design was a before-after observational study with a control group. The sites were located on county roads with a speed limit of 45 to 55 mph (70 to 90 km/h), with a 10 mph (16 km/h) or greater reduction in the posted speed limit, and a history of speed-related safety concerns. There were four treatment sites and one control site. FIGURE 10 Effects of engineering measures on speed change [Source: Charlton and Baas (2006)].

20 R3 = path radius on exit from the roundabout (m) a, b = see Table 13 d3 = distance between the midpoint of the path on the circulating roadway and the point of interest on the exit (m) Venter = the predicted entry speed for the roundabout (km/h) d1 = distance between the point of interest on the entry and the midpoint of the path on the circulating roadway (m) U.S. CUSTOMARY: Where: Vexit = the predicted exit speed for the round- about (mph) R1 = path radius on entry to the roundabout (ft) R2 = path radius on the circulating roadway (ft) R3 = path radius on exit from the roundabout (ft) a, b = see Table 13 d3 = distance between the midpoint of the path on the circulating roadway and the point of interest on the exit (ft) Venter = the predicted entry speed for the roundabout (mph) d1 = distance between the point of interest on the entry and the midpoint of the path on the circulating roadway (ft) TABLE 13 SPEED PREDICTION PARAMETERS FOR ROUNDABOUTS Metric U.S. Customary e = +0.02 e = –0.02 e = +0.02 e = –0.02 a 8.7602 8.6164 3.4415 3.4614 b 0.3861 0.3673 0.3861 0.3673 turn lane and bicycle lanes on either side of the road. Knapp and Rosales (2007) aggregated the results from a number of different road diet studies, analyzing the effects on both crash risk and operating speed. The impact of road diets on operating speeds is available but is not very detailed. The authors summarized before-after studies from 13 four-lane to three-lane conversions implemented in a number of states. The converted roadways had average daily traffic volumes between 8,400 and 24,000 vehicles per day, and experienced the following impacts on speed: • Average or 85th percentile speed reductions usually less than 5 mph (8 km/h); and • Up to a 70% reduction in excessive speeding (i.e., the number of vehicles traveling 5 mph (8 km/h) faster than the posted speed limit). The settings for these studies are not provided in the article, so the results are not necessarily directly applicable to transition zones. Nonetheless, road diets might be imple- mented at the downstream end of a transition zone and car- ried through the developed area as a means of reinforcing the lower speed limits. Rodegerdts et al. (2007) Rodegerdts et al. (2007) sought to develop a set of opera- tional, safety, and design tools that were based on U.S. experience with roundabouts. Although this research does not specifically mention the use of roundabouts in transition zones, roundabouts may be used as gateways to small towns (if an intersection is present within the transition zone), and the tools are therefore useful in this regard. With respect to operating speed, the following prediction models were developed for vehicles entering and exiting roundabouts: METRIC: Where: Vexit = the predicted exit speed for the round- about (km/h) R1 = path radius on entry to the roundabout (m) R2 = path radius on the circulating roadway (m)

21 Arnold and Lantz (2007) As part of a larger experiment with traffic control devices in rural areas, Arnold and Lantz (2007) assessed the effect of transverse optical bars (shown in Figure 11) on the approach to a small town, where the speed limit drops from 55 mph to 45 mph (90 km/h to 70 km/h). They were placed on both approaches to the town, and speeds were assessed before, 7 days after, and 90 days after installation to assess any novelty effect. The results are shown in Table 14. Stations 1 through 3 were located well in advance of the bars in the 55 mph (90 km/h) speed zone, at the end of the bars just at the start of the 45 mph (70 km/h) speed zone, and approximately in the center of town, respectively. The optical speed bars had an overall positive impact, as vehicle speeds recorded at the 45 mph (70 km/h) speed FIGURE 11 Optical speed bars in Virginia [Source: Arnold and Lantz (2007)]. TABLE 14 RESULTS OF OPTICAL SPEED BAR EXPERIMENT Time Period Westbound Eastbound Data Collection Period Before After* After 90** Before After* After 90** Station 1 All days 56.71 59.16 59.54 59.69 Weekday 56.71 58.78 59.58 59.64 Weekend 56.71 59.42 59.45 59.79 Day (6 a.m. to 6 p.m.) 56.86 59.12 59.73 59.87 Night (6 p.m. to 6 a.m.) 56.56 59.23 59.35 59.5 Station 2 All days 54.42 49.31 51.02 56.77 55.57 47.22 Weekday 54.49 49.59 50.85 56.75 55.68 47.18 Weekend 54.25 48.74 51.45 56.79 55.3 47.34 Day (6 a.m. to 6 p.m.) 54.74 49.86 51.85 57.02 56.16 47.71 Night (6 p.m. to 6 a.m.) 54.09 48.85 50.19 56.51 54.97 46.74 Station 3 All days 45.56 39.98 46.91 37.25 41.99 47.04 Weekday 44.95 39.96 46.95 37.43 42.02 47.09 Weekend 45.68 40.06 46.63 36.79 41.94 46.7 Day (6 a.m. to 6 p.m.) 46.05 40.19 47.23 36.93 41.79 46.75 Night (6 p.m. to 6 a.m.) 45.05 39.72 46.53 37.57 42.2 47.37 [Source: Arnold and Lantz (2007)]. *After = within 7 days of installation. **After 90 = Approximately 90 days after installation. Note that the predicted entry/exit speed is independent of the speed on the rural approach and is determined by the characteristics of the circulating roadway and the distance from the roundabout.

22 limits outside the transition areas were 55 mph and 60 mph (90 and 100 km/h), and speed limits in the villages were 25 mph to 35 mph (40 km/h to 55 km/h). An extensive literature review was used to assemble a list of potential treatments. Treatments for implementation were selected in consultation with the local government, and were generally low cost, had the ability to accommodate farm vehicles and large trucks, and were compatible with the rural setting and driver expectations. Seven different treatments were tested, as shown in Table 15. Speed and volume data were collected from 1 month to 1 year after implementation, with the 85th percentile speed being the primary measure of effectiveness. The results, shown in Table 16, revealed that the speed feedback signs, speed table, median island using tubular markers, and speed limit markings with red background are the most effective. The converging chevrons and transverse pavement mark- ings were moderately effective, producing speed reductions of less than 3 mph (5 km/h). The remaining treatments were either ineffective or only marginally effective. Dixon et al. (2008) Dixon et al. (2008) conducted a literature review on rural speed transition zones leading to driving simulator testing of several alternative transition zone treatments. Because the study uses a driving simulator, the alternatives investigated were limited to either physically or perceptually narrowed roads at the transition locations. The following transition zone treatments were included in the full-scale simulation: • Layered landscape • Gateway with lane narrowing • Median treatment only • Median with gateway treatment • Medians in series with no pedestrian crosswalks • Medians in series with pedestrian crosswalks The speed transition was in all cases from 55 mph to 35 mph (90 km/h to 55 km/h), with a 45 mph (70 km/h) stepped-down speed limit in the transition zone. Speed mea- surements were taken at several locations starting in the 55 mph (90 km/h) section and throughout the transition zone up to and including the gateway. Table 17 shows the results, which indicated that all speed reductions were minimal but the most effective alternatives were the median treatments, particularly the medians in a series or combined with a gateway. The layered landscape treatment and the gateway with lane narrowing treatment did not result in statistically significant speed reductions. Also, the speed reductions generally occurred vicinal to the transition treatments—the reductions generally did not per- sist downstream of the treatment. limit sign at the downstream end of the bars decreased for all time periods 90 days after installation at both ends of town. Based on the experiment results, the magnitude of the speed decrease from optical speed bars in transition zones is expected to be 3 to 9.5 mph (5 to 15 km/h). No conclu- sions can be drawn about the habituation effect, as speeds increased with time at one end of town but decreased with time at the other end of town. Fitch and Crum (2007) Fitch and Crum (2007) also tested optical speed bars on the approaches to several small villages in Vermont with less than impressive results. Speeds were measured on Saturdays and midweek to examine the effects of the striping on both local and tourist traffic. The researchers concluded that the striping is ineffective at reducing 85th percentile speeds, producing only a 1.0 mph (1.6 km/h) reduction 4 months after installation. Curi- ously, the speed-reducing effect seemed to be stronger with drivers who were exposed to the markings on a daily basis. Department for Transport (2007) The U.K. Department for Transport (2007) reported that 85th percentile speed reductions of 3.5 mph (5.6 km/h) and 6 mph (9.6 km/h) for “before” speed limits of 40 mph and 60 mph respectively are achievable by implementing local 30 mph (50 km/h) speed limits. This result comes from a 1994 program implemented in Suffolk County Council where 30 mph (50 km/h) speed limits were enacted in 450 villages and commu- nicated by standard speed limit signing and entry roundels. However, it is noted that the new limits were complemented by a policy of implementing physical traffic calming where a crash problem was evident, mobile speed cameras, and a high- profile anti-speeding campaign. Hence, it is not possible to attribute the speed reductions to any one particular measure. Hallmark et al. (2007) Hallmark et al. (2007) set out to examine the effects of selected transition zone treatments at several rural commu- nities in Iowa. Site selection criteria included the following: • Through, paved, major county or state highway; • No traffic calming currently in place or planned; • No construction, reconstruction, or significant main- tenance activities planned along the route during the study period; • No access control; and • No adverse geometry such as sharp horizontal curves or steep vertical curves where treatments would be placed. Of the 18 sites that met the inclusion criteria, the five locations with the most significant speeding problems (determined by the difference between the posted speed and prevailing travel speed) were selected for treatment. Speed

23 TABLE 15 DESCRIPTION OF TRANSITION ZONE TREATMENTS IN IOWA City (population) Treatment Roadway AADT (veh/day) Cross section (all are two-lane) Union (427) Transverse pavement makrings’ with speed feedback sign D-65 (west edge of City) 830 Asphalt (22.4 ft), unpaved shoulders Transverse pavement markings’ with speed feedback sign S-62/SH (from intersection with D-65 to north city limit) 1,680 Concrete (40.0 ft), curb and gutter Lane narrowing using painted center island and edge line markings Transverse pavement markings’ SH 215 (near south city limit) 1,000 Asphalt (22.4ft), unpaved shoulders Roland (1,324) Converging chvrons with “25 MPH” pavement legend E-18 (near east and west city limits) 2,300 Asphalt (22.6 ft), unpaved shoulders Lane narrowing using shoulder widen- ing and “25 MPH” pavement legend E-18 (from intersection with R-77 to east city limit) 2,300 Concrete (36.0 ft), curb and gutter “25 MPH” pavement legend E-18 (from intersection with R-77 to west city limit) 2,300 Asphalt (22.6 ft), unpaved shoulders Gilbert (987) Speed table E-23 (center of community) 1,480 Asphalt (22.0 ft), shoulders Slater (1,306) Lane narrowing with center island using tubular markers channelizing markers R-38 (from intersection with SH210 to south city limit) 2,060 Concrete (25.8 ft), curb and gutter Speed feedback sign R-38 (near north city limit) 2,870 Asphalt (22.6 ft), unpaved shoulders “SLOW” pavement legend SH 210 (west from intersection with R-38 to west city limit) 2,940 Asphalt (22.5 ft), unpaved shoulders Dexter (689) “35 MPH” pavement legend with red background F-65 (near east and west city limits as well as at curve before west city limit) 1,000 Asphalt (25.4 ft), unpaved shoulders [Source: FHWA (2009)]. A request for experimentation was submitted to and approved by FHWA for this treatment. TABLE 16 RESULTS OF THE IOWA TRANSITION ZONE TREATMENTS Treatment Change in 85th percentile speed (mi/h) Cost Maintenance Application Transverse pavement markings’ -2 to 0 $ Regular painting community entrance Transverse pavement markings’ with speed feedback signs -7 to -3 $$$ Regular painting community entrance Lane narrowing using painted center island and edge marking -3 to +4 $ Regular painting entrance or within community Converging chevrons’ and “25 MPH” pavement markings -4 to 0 $ Regular painting community entrance Lane narrowing using shoulder markings and “25 MPH” pavement legend -2 to 4 $ Regular painting entrance or within community Speed table -5 to -4 $$ Regular painting within community Lane narrowing with center island using tubular markers -3 to 0 $$$ Tubes often struck needing replacement within community Speed feedback sign (3-months after only) -7 $$$ Troubleshooting electronics entrance or within community “SLOW” pavement legend -2 to 3 $ Regular painting entrance or within community “35 MPH” pavement legend with red background -9 to 0 $ Background faded quickly; accelerated repainting cycle entrance or within community $ = under $2,500 $$ = $2,500 to $5,000 $$$ = $5,000 to $12,000 ‘Experimental approval required per Section 1A, 10 of MUTCD [Source: FHWA (2009)].

24 Measures of effectiveness were as follows: • Absolute change in speed within the impact zone (km/hΔ) • Δ per meter (Δ/m)—this standardizes the Δ enabling comparison across treatments • The lowest travel speed observed in the impact zone— Speed at maximum Δ (v@max Δ)—this refers to vehi- cle speed at the point at which the change in speed is greatest. • Speed at the downstream end of the impact zone • Percentage of speed change across the impact zone The impact zone is treatment specific, starts at the point where the driver first perceives the treatment, and ends at the point where the treatment is reached. Figure 13 presents the results of the simulator studies for village entries. Countdown signs produced the best results across all three performance indicators, yielding a net 9% reduction in speed approaching the village. The speed profile of the countdown signs is also markedly different from other treatments, show- ing motorists slowing earlier. Rumble strips with a raised profile showed the second best performance (with a net 7% reduction in speed). This is notable because the location and duration of rumble strips are similar to other surface treat- ments, such as dragon’s teeth or combs with chevrons, which did not perform as well. The tactile feedback seems to pro- vide extra emphasis, which results in better performance. The vehicle-activated signs performed the worse, but the report noted that this device had to be located downstream of the village entry in order to comply with governing legislation, so it is not directly comparable to the other treatments. TABLE 17 SPEED REDUCTIONS IN SIMULATED TRANSITION ZONES Jamson et al. (2008) As part of a larger research project, Jamson et al. (2008) con- ducted a driver simulator study to estimate the effectiveness of a number of low-cost, speed-reducing treatments at vil- lage entries. The treatments included perceptual illusions as well as physical restrictions, and are listed here: • Vehicle-activated signs with SLOW DOWN • Countdown signs • Rumble strips with raised profile • Dragon’s teeth • Combs with chevrons • Combs (see Figure 12) • Build-outs • Trees FIGURE 12 Pavement marking combs [Source: Jamson et al. (2008)].

25 Donnell and Cruzado (2008) Donnell and Cruzado (2008) conducted research on the use of speed feedback signs to manage speeds at high-to-low speed transitions on rural roads in Pennsylvania. The study methodology was a before-after examination of operating speeds. At each of the 17 study sites, the feedback signs were placed for a 1-week period, as it is common practice for the DOT to use the limited number of feedback signs at a greater number of sites. At four sites, the signs were implemented for 2 weeks to determine if the longer deploy- ment had any effect on speeds. Enforcement was not a part of the overall treatment. The speed feedback sign was placed 500 ft (150 m) downstream of the speed threshold. Thirteen of the 17 sites were high-to-low speed transition zones, with a speed differential of either 15 or 20 mph (24 to 32 km/h). Speeds were measured before placement of the feed- back signs, during the time that the sign was in place, and after sign removal to determine if there were any residual effects on speed. Only free-flow speeds during daylight, in off-peak hours on weekdays, and on dry pavement were included in the analysis. Speeds were measured about 0.5 mile (0.8 km) upstream of the speed feedback sign (where the feedback sign could not be seen), at the feedback sign, and 500 ft (150 m) downstream of the feedback sign. Met- rics included mean speed, 85th percentile speed, speed variance, and percentage of vehicles exceeding the posted speed limit. The results showed that on average, mean speed reduc- tions of approximately 6 mph (10 km/h) were achieved at the speed feedback sign and downstream of the sign for the 13 sites where there was a high-to-low speed transi- tion (from the before to the during period). This effect was present only while the sign was in place, and mean speeds rebounded to before levels in the week after sign removal. Implementing the speed feedback sign for 2 weeks permit- ted the mean speed reductions to be sustained for the entire deployment period, but in these instances mean speeds also rebounded to pre-deployment levels shortly after the signs were removed. Abate et al. (2009) Abate et al. (2009) conducted an evaluation of two speed transition zones (one on each approach) to a village in Sal- erno, Italy. The measures employed included transverse rumble strips, dragon’s teeth markings, an optical narrowing created by pavement markings at the edge of the road, and an overhead information sign (or flag portal) that is cantilevered from the right side of the road (see Figure 15). A photograph of the flag portal is shown in Figure 16. FIGURE 13 Speed reductions at transition zones [Source: Jamson et al. (2008)]. Based on the performance of the countdown signs, the researchers advanced this measure to a second level of inves- tigation that aimed to determine if repeated exposure to this treatment would lessen the observed effects. To this end, sim- ulator participants were exposed to the countdown-treated approach four times, with results shown in Figure 14. The shape of the speed profile for treated drives is dis- tinctively different from the baseline. The change in speed over the impact zone and the speed at the village entry for all drives were statistically significantly different than the base- line speed. This result suggests that the countdown signs are an effective speed-reducing measure, and that they may have a lasting effect. FIGURE 14 Effects of countdown signs on speed [Source: Jamson et al. (2008)].

26 The 85th percentile speed on the approaches to the test village was about 55 mph (90 km/h). The speed limit in the built-up area of the village is 30 mph (50 km/h). Two sample gateways were advanced for testing: • Alternative 1—Transverse rumble strips, transverse optical bars, peripheral transverse bars, a roadside fence converging toward the carriageway, a transverse strip of colored road surface with markings resembling brick pavers, and a sign gantry. • Alternative 2—The same measures as Alternative 1, but with the “brick pavers” replaced with a mountable central island that creates a horizontal deflection. The deflection is about 123 ft (37.5 m) long, and creates a lateral shift of 8 ft (2.5 m). Figure 17 shows images of the two alternatives. FIGURE 17 Simulated transition gateways [Source: Lamberti et al. (2009)]. Alternative 1 Alternative 2 Mean speeds were measured in the rural area preceding the gateway (labeled the “control”), at the gantry of the gateway, and in the village midpoint. Table 18 shows the results of the FIGURE 15 Gateway treatment in Salerno, Italy [Source: Abate et al. (2009)]. FIGURE 16 Italian flag portal [Source: Abate et al. (2009)]. Speed was measured at several stations before, in, and after the transition zone. Although it is difficult to extract from the document the exact effect of the gateway treatment, it is clear that the authors reported a successful reduction in operating speed. Lamberti et al. (2009) The purpose of driver simulator research conducted by Lam- berti et al. (2009) was to investigate drivers’ speed choice on rural highways that cross small urban communities in situa- tions with and without gateways and village traffic calming, and to examine the effects of low-cost versus high-cost gateways. Operating speed on transition zones that did not have a gateway was measured in the field; operating speed in transition zones with gateways were estimated from driving simulator runs.

27 Township of Clarington, Ontario, Canada. A single transi- tion zone was shown as an approach to an intersection (the start of the urban area), with transition zone improvements of dragon’s teeth, edge and centerline markings, and a row of regularly spaced trees planted on one side of the road (the other side of the road is a wooded area). The speed limit in the urban area is 30 mph (50 km/h); the speed limit in the rural area is 50 mph (80 km/h). Both the mean and the 85th percentile speeds were used to assess treatment effectiveness, measured at the rural/urban threshold and 787 ft (240 m) downstream of the threshold. Table 19 shows the results: at the threshold there was almost a 5 mph (8 km/h) reduction or a 10% reduction in the 85th per- centile speed for both directions combined, and about a 3 mph (5 km/h) or 7% reduction in the mean speed. Even with the transition zone treatment in place, the mean and 85th percen- tile speeds were still 6 mph (10 km/h) and 12 mph (20 km/h) higher than the posted speed limit of 30 mph (50 km/h). The changes in the mean and 85th percentile speeds 787 ft (240 m) downstream of the threshold were less impressive at about 0.6 mph and 2 mph (1 km/h and 3 km/h ), respectively. However, the “before” mean and 85th percentile speeds at this down- stream location were significantly lower than the same mea- sures at the threshold, indicating that the built environment is having a calming influence on travel speeds. Cruzado and Donnell (2009) Cruzado and Donnell (2009) explored the effect of roadway, roadside, and traffic control elements on operating speeds in transition zones on rural, two-lane highways in Pennsylvania. Twenty sites were used in the analysis. All sites were free of major intersections, had less than 10% heavy vehicles, and had visible pavement markings on smooth road surfaces. Addition- experiment for the existing condition (ALT0), the Alternative 1 gateway (ALT1), and the Alternative 2 gateway (ALT2). TABLE 18 EFFECTS OF GATEWAYS ON OPERATING SPEEDS Station Direction Mean Speed, mph (km/h) Alt0 Alt1 Alt2 Control South 49.6 (79.4) 47.1 (75.3) 45.3 (72.4) Gateway South 48.1 (77.0) 38.1 (61.0) 37.6 (60.1) Village Midpoint South 47.1 (75.4) 40.5 (64.8) 38.0 (60.8) Control North 51.3 (82.1) 50.6 (80.9) 47.3 (75.6) Gateway North 45.0 (72.0) 38.1 (61.0) 38.5 (61.6) Village Midpoint North 47.2 (75.5) 41.5 (66.4) 38.9 (62.2) [Source: Lamberti et al. (2009)]. TABLE 19 SPEED DATA FROM BURKETON, ONTARIO Before After Change N* Mean 85th% N* Mean 85th% Mean 85th% At rural/urban threshold Southbound 45 66.9 77.1 42 61.5 69.1 –5.4 –8.0 Northbound 57 68.6 77.2 52 64.3 69.7 –4.3 –7.5 Combined 102 67.8 77.1 94 63.1 69.4 –4.7 –7.7 240 meters downstream of threshold Southbound 29 58.6 63.9 41 56.6 60.0 –2.0 –3.9 Northbound 44 55.1 61.8 54 55.2 59.9 0.1 –1.9 Combined 73 56.5 63.0 95 55.8 59.9 –0.7 –3.1 [Source: Chartier (2009)]. *N = Number of vehicles, 85th% = 85th percentile. All speeds reported in km/h. The results were promising—the gateways produced speed reductions of 6.9 to 10.6 mph (11 to 17 km/h) at the gateways. The additional cost of constructing a mountable island to cre- ate the horizontal deflection (Alternative 2) did not appear to be worth the effort, as the mean speeds were essentially the same as those produced by the lower cost Alternative 1 gate- way. However, these were the results of a simulator study, and field studies and observations find that drivers tend to flat- ten the horizontal deflections by traversing painted islands, resulting in a less effective gateway than a raised median. Chartier (2009) In a presentation concerning rural to urban transition zones, Chartier (2009) included a case study concerning a transi- tion zone improvement in the Hamlet of Burketon of the

28 These results are interpreted as follows: • The presence of a school/children warning sign had the largest influence on speed transition (the difference between the speed measured at two ends of the transi- tion zone)—a 7.6 mph (12.1 km/h) reduction in speed. • A horizontal curve that also has a warning sign resulted in a 4.9 mph (7.8 km/h) reduction in speed across the transition zone. • The remaining features of the transition zone that influ- enced speed in decreasing order of importance were the posted speed limits, an intersection ahead warning sign, a curve without warning signs, paved shoulder width, the transition zone length, and the number of driveways. • Lane width, lateral clearance, and the presence of curb and gutter were not statistically significant, and there- fore are not influential on speed transitioning. The researchers also noted that the speed reduction is compromised by a CURVE AHEAD warning sign. Russell and Godavarthy (2010) Russell and Godavarthy (2010) conducted an evaluation of four different speed management measures on rural roads in Kansas: colored pavement, a solar speed display, a mobile speed trailer, and optical speed bars. The objec- tive of the study was to test measures that mitigate speeds on the through approaches at stop-controlled intersections and at other road segments. The colored pavement (see Fig- ure 18) and the solar displays were implemented at rural to urban transitions; the remainder of the speed management measures were used at approaches to rural intersections or horizontal curves. Mean and 85th percentile speeds were the measure of effectiveness. All speeds were measured manually using a handheld radar gun, and are shown in Table 21. The colored pavement treatment did not definitively reduce operating speeds, but did show promise at one of the three sites. The solar speed displays and mobile speed trailer showed promise, but the researchers stated that these are short-term results from a limited number of installations and more extensive testing should be undertaken to confirm the speed-reducing capabilities of these devices. The optical speed bars, which were used on the approaches to horizontal curves, exhibited conflicting results and could not be deter- mined to be effective measures. ally, each site had a REDUCED SPEED AHEAD sign as per the 2003 MUTCD. The speed reductions were all either 20 mph or 15 mph (30 km/h or 25 km/h), except for one site that was a 10 mph (15 km/h) speed reduction. The rural speed limits were 40 mph, 45 mph, and 55 mph (65 km/h, 70 km/h, and 90 km/h). Speeds were recorded in the rural area, at the start of the transition zone (defined as the REDUCED SPEED AHEAD sign), and at the end of the transition zone (defined as the start of the reduced limit). Only free flow speeds were used in the analysis. Transition zone features that were collected for analy- sis were lane width, paved shoulder width, stabilized shoulder width, total paved roadway width, lateral offset to obstruc- tions, presence of curb and gutter, access density, horizontal and vertical alignment data, posted speed limits in the urban and rural areas, and the number and type of warning signs. The analysis employed both ordinary least squares linear regression and multilevel models, both of which provided simi- lar results for most of the independent variables. Table 20 shows the results from the multilevel model, which the researchers considered to be a better representation of the data. TABLE 20 FACTORS INFLUENCING SPEED REDUCTIONS IN TRANSITION ZONES

29 FIGURE 18 Colored pavement [Source: Russell and Godavarthy (2010)]. TABLE 21 RESULTS OF KANSAS STUDY ON RURAL SPEED MANAGEMENT Treatment Location Description 85th Percentile Speed (mph) Mean Speed (mph) Before After Before After Colored Pavement 65 mph → 55 mph → 45 mph on a two-lane road 54.8 55.3 51.1 50.7 65 mph → 55 mph → 30 mph on a two-lane road 57.4 47.7* 48.8 42.4* 51.8 52.9 46.6 47.8 Solar Speed Display 70 mph → 55 mph on a four- lane road with a painted median 56.8 56.4 53.2 51.5* 65 mph → 55 mph on a two- lane road 58.5 56.3* 54.9 51.8* 63.2 54.1* 57.3 50.2* Mobile Speed Trailer Approach to an intersection on a four-lane road with a 70 mph 73.2 71.1* 69.8 67.3* 69.1 58.8* 62.4 55.3* Speed Bars 55 mph → 45 mph advisory speed at a hori- zontal curve 45.2 45.2 42.2 41.3 47.6 46.3 43.9 40.8* 46.8 45.9 43.1 41.0* 65 mph → 30 mph advisory speed at a hori- zontal curve 52.5 59.7 47.8 52.9 55.3 56.1 47.8 49.2 *Significantly different from the Before measurements at a 95% level of confidence. Crash Studies Tziotsis (1992) The purpose of research by Tziotsis (1992) was to investigate the magnitude and characteristics of crashes that occur on “feeder roads” to provincial cities (i.e., roads on the immedi- ate outskirts). These facilities are generally partially devel- oped sections of road that connect the typical urban road to a rural road (i.e., a transition zone). The transition zones examined were on average 1.3 miles (2.2 km) in length and were defined by the speed limit [i.e., 45/50/55 mph (75/80/90 km/h) in transition zones, and 60/65mph (100/110 km/h) for the rural road zones]. Five years of crash data were used in the analysis of both aggregate crash statistics from several transition zones and investigations of the safety performance of individual transitions. The aggregate analysis confirmed that transition zones experienced crash rates that are markedly higher than those experienced on rural roads. Transition zones experienced an annual crash rate of 72 casualties per 100 million vehicle- miles (45 casualties per 100 million vehicle-kilometers), whereas rural zones produced a crash rate of 43 casualties per 100 million vehicle-miles (27 crashes per 100 million vehicle-kilometers). Furthermore, as a proportion of total crashes, rear-end and dusk/dawn crashes peaked within the transition zones. This report did not consider the direction of movement for crash-involved vehicles and therefore did not compare crash rates of vehicles leaving the built-up area with those of vehicles entering the built-up area. The researchers concluded that the increased crash rate in the transition zone was related to inadequate road design and an increase in roadside development that occurs in the transition zone. Specific transition zone design deficiencies were insufficient curve delineation, lack of turning lanes and acceleration/deceleration lanes, inadequate shoulders, and poles located in hazardous locations. Herrstedt et al. (1993) Herrstedt et al. (1993) provided data from a wide variety of traffic calming measures that have been put into practice in Denmark, France, and Germany. The examples span a wide variety of locations where these measures may be used effec- tively (e.g., urban centers, approaches to villages). Of interest to this synthesis are locations where measures were imple- mented on a highway that runs “through a town.” Table 2 presents examples of some treatment at the high-to-low speed transition where a highway traverses a town and the text men- tions a “gateway” or “portal.”

30 The traffic calming designs implemented varied signifi- cantly and were classified as follows: A. No measures in the village, but use of gateway sign- ing associated with significant markings/colored sur- face/minor narrowing, and in some cases physical measures at the gateway. B. Measures within the village, involving mainly road markings, colored surfaces and traffic islands, some with gateway features. C. Significant physical measures within the village, involving horizontal and vertical deflections, usually in conjunction with gateways. Overall, 1,400 casualty crashes were analyzed and cat- egorized as slight injury crashes, or fatal and serious injury crashes. The analysis included all crashes on the main road (including intersection crashes) between the gateways (or speed limit reductions for the village if no gateway was pres- ent). Table 22 shows the aggregated before-after frequency. TABLE 22 EFFECTS OF VILLAGE TRAFFIC CALMING IN THE U.K. Period No. of Crashes % Fatal + Serious Injury Years (average)Slight Injury Fatal + Serious Injury All Before 89.8 35.3 125.2 28.2 7.2 After 76.8 17.0 93.8 18.1 5.3 CRF 0.14 0.52 0.25 — — CRF = crash reduction factor The sites averaged one to three crashes per year pre-traffic calming, with a low of no crashes to a high of 15.6 crashes/year. Injury crashes were reduced at 34 of the 56 sites; only 10 of the 34 sites had statistically significant reductions. However, using the aggregated crash data, the reductions in all crash categories were statistically significant. Furthermore, using national crash trends as an improvised control group, Wheeler and Taylor con- cluded that the crash reduction factors for traffic-calming on main roads in villages are 0.20 to 0.25 for all injury crashes, and 0.33 to 0.50 for serious injury and fatal crashes. Wheeler and Taylor did not appear to explicitly control for RTTM effects, but the report mentions that “many vil- lages have more of a perceived problem than a real safety problem.” This indicates that the majority of the sites studied were not likely suffering from a high rate of crashes and the RTTM effect would be less pronounced. Additional analysis included examining the effects of traffic calming based on different types of calming, road volumes, and speed reduc- tions (see Table 23). The varied measures that were implemented showed a 44% and 36% reduction in injury and all crashes, respec- tively. However impressive this reduction in crash risk, it is of limited value in high-to-low speed transition zones. It is not possible to tease out the effects of the high-to-low speed transition treatment from the effects of the “in town” mea- sures that were also employed. Nonetheless, these examples highlight that transition treatments can be effective in reduc- ing crash risk as well as speeds. County Surveyors’ Society (1994b) The County Surveyors’ Society (1994b) of the U.K. collected and summarized data on 85 traffic calming installations in the United Kingdom, with 25 of the installations classified as rural road/area installations. The intentions of this effort were to provide practitioners with a snapshot of traffic calm- ing in the U.K., and to detail a number of case studies so that practitioners might better understand what plans and mea- sures work and do not work in different situations. Table 3 shows a summary of the crash risk effects of traf- fic calming on the 25 rural case studies. All of the traffic calming was implemented for one or more of the following objectives: reduce speed, reduce crash occurrence, and/or reduce through traffic. However, the measures implemented at each site were not implemented for the same reasons and are not of the same form (e.g., humps versus narrowings). Therefore, this analysis is necessarily general. Traffic calming on rural roads had a dramatic impact on crash occurrence—a 65% reduction in crash occurrence. Of the 20 sites that reported before-after crash and volume data, 18 experienced reductions in crash rates and two expe- rienced increases in crash rate. This result is impressive but should be treated with caution because the evaluation meth- odologies are naïve before-after studies of crash occurrence. There is no accounting for sundry effects or regression-to- the-mean (RTTM) artifacts, and at some sites the “after” periods were quite short. Nonetheless, a reduction in crash rate is expected, although it is more likely in the range of a 45% to 55% reduction. The report data do not indicate whether the rural road is a primary route or a local road. If the “before” traffic volume was used as a surrogate for road classification, for the eight sites with daily traffic volumes greater than 10,000 (which are assumed to be primary routes or arterial roads) the crash rate was reduced by 47%. Wheeler and Taylor (2000) Wheeler and Taylor (2000) conducted a retrospective study of crash occurrence broken down by severity for 56 traffic- calming designs in various villages across the U.K. The roads under study were all classified as major or main roads.

31 TABLE 24 EFFECTS OF TRAFFIC CALMING ON CRASH OCCURRENCE ON IRISH NATIONAL PRIMARY ROADS (1993 TO 1996) Group Crash Type No. of Crashes Average of Annual Number of Crash CMF Before After Before After Both Approaches (N =14) Fatal 11 1 0.11 0.01 0.13 Serious Injury 19 7 0.20 0.13 0.71 Minor Injury 49 20 0.54 0.35 0.64 All Casualty 79 28 0.84 0.50 0.59 One Approach (N = 7) Fatal 0 0 0.00 0.00 — Serious Injury 9 1 0.09 0.02 0.26 Minor Injury 9 6 0.09 0.10 1.07 All Casu- alty 18 7 0.18 0.12 0.67 CMF = crash mitigation factor. TABLE 23 CRASH REDUCTIONS FOR TRAFFIC CALMING ON MAIN ROADS IN THE UNITED KINGDOM Fatal + Serious Injury All Injury Before After CRF Before After CRF Traffic Calming Type A. Gateways Only 7.7 3.5 0.55 25.9 21.1 0.19 B. Calming in Village 15.6 10.3 0.34 48.0 46.0 0.04 C. Aggressive Calming 11.7 3.5 0.70 49.7 27.3 0.45 Traffic Volume <4,000 vpd 2.3 0.9 0.61 7.2 6.1 0.15 4,000 to 7,999 4.5 2.3 0.49 17.3 11.2 0.35 8,000 to 11,999 17.6 8.9 0.49 61.3 44.9 0.27 >12,000 vpd 9.3 5.3 0.43 32.5 32.1 0.01 Speed Reduction 0–2 mph (0–3 km/h) 5.8 2.7 0.53 19.7 17.8 0.10 3–4 mph (5–6.5 km/h) 7.5 5.5 0.27 26.6 22.8 0.14 5–6 mph (8–10 km/h) 2.1 1.5 0.29 8.4 5.7 0.32 ≥7 mph (11 km/h) 4.8 0.9 0.81 16.7 8.9 0.47 CRF = crash reduction factor, vpd = vehicles per day. The reduction in crashes at the “gateways only” sites is relatively impressive. The report notes that the gateways were typically more substantial than simply signing and minor markings; they often included minor narrowings, surface treatments, and so on. In the “calming in villages” group, the majority of the crash savings resulted from sites that had calming in the village, and gateways (as opposed to those without). It is clear that the “aggressive calming” group yielded the most significant crash reductions. Crowley and MacDermott (undated) Crowley and MacDermott (undated) conducted an evaluation of traffic calming projects implemented on national primary roads in Ireland between 1993 and 1996. The evaluation quan- tified the reductions in crashes resulting from traffic calming in villages and towns. Twenty-one different traffic-calming projects were evaluated. There is considerable variation in the traffic calming implemented at each site; the average cost of implementation for traffic calming on both approaches was €215,000 (2000 prices) with a range of €29,000 to €432,000. The 21 sites were divided into 14 sites that had traffic calming implemented on both approaches and 7 sites that had traffic calming implemented on only one approach. Table 24 shows the effects of the traffic calming on crash occurrence.

32 of the roundels were supplemented with flashing beacons. All of the VAS were located 66 ft to 164 ft (20 m to 50 m) downstream of the beginning of the speed limit change (i.e., within the village speed limit). The 5-year “before” and “after” crash analysis covered the period from 1990 to 2000. For the speed limit VAS, all crash data were obtained for the length of road of about 0.6 mile (1 km) from the start of the speed limit. This section would be without intersections, which could generate crashes not influenced by the sign. The casualty crash frequency was analyzed using Empirical Bayes techniques that account for RTTM and maturation effects. At the 19 sites with a 30 and 40 mph (50 and 65 km/h) urban speed limit (and avail- able crash data) the VAS produced a statistically significant reduction in casualty crash frequency of 34% (± 8%). Garder et al. (2002) Garder et al. (2002) conducted a study to evaluate the safety effect of traffic calming on arterial roads and to examine the acceptance of these measures. They found that there are a limited number of traffic-calmed arterials—even fewer that have been evaluated. The North American experience indi- cates that measures that might be considered traffic calming have mostly been implemented for mobility/capacity reasons (i.e., roundabouts, and four-lane to three-lane conversions). The effectiveness of arterial traffic calming has been “moder- ate,” although there is a clear reduction in pedestrian injuries. Public opinion surveys indicated that horizontal deflections are preferred to vertical deflections on arterial roads. Souleyrette et al. (2003) Souleyrette et al. (2003) analyzed the safety records of vari- ous types of on-street parking in smaller communities in the state of Iowa. Specifically, the researchers wanted to compare the safety performance of streets that had diagonal parking with streets that had other types of curbside park- ing. Several factors were examined to determine possible contributions to crash occurrence, including road width, clearance to parked vehicles, traffic volumes, community population, and length of parking area. None of these fac- tors, with the possible exception of population, displayed a clearly definable relationship to crash occurrence. The dif- ference in average non-intersection crash rates for diago- nal and parallel parking streets was almost negligible (see Table 26). In fact, those observed rates were less than sam- ple locations with no parking at all. The authors recognized that the research did not present a statistically sound sample of locations, but stated that the data gathered were quite sub- stantial and covered most areas of the state of Iowa. They concluded that there is no compelling reason for a blanket prohibition of angle parking along Iowa’s primary exten- sions in all urban areas, and that each community should be examined on a case-by-case basis. The evaluation demonstrated impressive results but did not control for exposure (i.e., traffic volumes) and other sun- dry effects. Moreover, the evaluation expressly noted that the primary criterion for selecting traffic calming projects is the number of crashes, and therefore RTTM effects may overestimate the effectiveness of the traffic calming. The RTTM effect is highlighted if the crash movement factors for the 14 locations with traffic-calming implemented on both approaches are calculated separately for the nine locations with annual casualty crash frequencies less than one, and the five locations with annual casualty crash fre- quencies greater than one (see Table 25). The effectiveness of traffic calming at the locations with the higher “before” crash frequency appears to be much higher, at least part of which is simply RTTM effects. TABLE 25 REGRESSION-TO-THE-MEAN EFFECTS ON THE IRISH DATA Group Crash type Average of Annual Number of Crash CMF Before After Frequency < 1.0 (N = 9) Fatal 0.05 0.01 0.28 Serious Injury 0.07 0.06 0.82 Minor Injury 0.18 0.15 0.83 All Casualty 0.30 0.22 0.74 Frequency > 1.0 (N = 5) Fatal 0.06 0.00 — Serious Injury 0.13 0.08 0.65 Minor Injury 0.36 0.20 0.55 All Casualty 0.54 0.28 0.52 CMF = crash mitigation factor. Winnett and Wheeler (2002) Winnett and Wheeler (2002) conducted an evaluation of the effectiveness of vehicle-activated signs (VAS) in reducing speeds and crashes in the United Kingdom. One type of VAS evaluated was a dynamic speed limit sign (a “speed roun- del”) for speed limit changes employed mainly at village sites on rural, undivided roads (see Figure 7). Sites were selected for VAS implementation if they had either a recent history of crashes in which inappropriate speed was a contributory factor, or a record of excessive speed for the conditions was believed to be a potential prob- lem. Sites selected for evaluation also required suitable sight lines to the VAS by the approaching driver, and traditional traffic control devices (such as fixed signs and markings) that were in compliance with the applicable standards. The VAS implemented was a speed roundel, the U.K. equivalent to the SPEED LIMIT SIGN (R2-1), that was blanked out and activated by motorists exceeding the speed limit. Some

33 rumble strips at seven pilot locations. A rumblewave surface is shown in Figure 9. FIGURE 19 Effect of environmentally adapting through roads on crash occurrence [Source: Agustsson (2005)]. Rumblewave surfaces have been piloted at seven loca- tions, including a high-to-low speed transition, as shown in Table 10. Table 27 shows the effects of rumblewaves on the rates of personal injury crashes. The reduction in casualty crashes is significant, averaging a 55% reduction (range from 24% to 100%). However, in three of the six locations that were evaluated for crash risk, one of the problems being addressed was a high incidence of crashes; this raises the specter of RTTM bias and suggests that the 55% reduction is an overly optimistic estimate. Forbes (2006) Forbes (2006) conducted a retrospective, observational before-after study of the safety effects of traffic manage- ment in rural settlements located on rural arterial roads in Ontario, Canada. The research documented the types of measures implemented and their efficacy related to crash risk at 12 treatment sites selected by convenience. The crash analysis employed an Empirical Bayes analysis using crash frequency categorized by severity as the primary measure of effectiveness. Table 12 shows the sites used in the analysis and the data available at each site. The analy- sis indicated that the various traffic management measures were successful in reducing crash occurrence by 22% (28% for casualty crashes) at the five sites where crash data were available. Overall, these results indicated that traffic manage- ment in rural settlements in Ontario was effective in improv- ing safety. Nonetheless, the results should be viewed with some caution as the sample sizes were small, the treatments applied at each site varied considerably, and there may have been co-interventions that also contributed to the improve- ments (i.e., increased awareness and enforcement that were unaccounted for in the analysis). TABLE 26 CRASH RATES FOR CURBSIDE PARKING Parking (one side/ other side) No. of Segments Average AADT Average Crash Rate (100 MVM) All Non- intersection Diagonal/Diagonal 72 2,100 1,620 400 Parallel/Parallel 26 2,350 910 420 Diagonal/None* 4 1,070 2,710 860 Parallel/None* 3 1,260 1,540 0 Diagonal/Parallel 19 2,300 1,750 320 Parallel/Parallel with Diagonal in the centre of the street* 3 3,510 1,450 250 None/None* 14 5,040 1,870 630 *Small sample size—use with caution. AADT = average annual daily traffic, MVM = million vehicle miles. Agustsson (2005) Agustsson (2005) reported on the effectiveness of “environ- mentally friendly through-roads,” which are streets where traffic is managed by using different forms of speed-reducing measures. Twenty-one of these newly developed roads were implemented to reduce speed, increase safety, and improve road design. The measures used include gates, roadside res- ervations, medians, roundabouts, raised areas, changes in road surface, road markings, signing, lighting, road closures, rumble strips, and bicycle lanes. The average length of the through-road is about 0.6 mile or 1 km (it is not clear if this is the total length of the road or the length of the road that was treated, although it is expected that it is the latter). Ten years of crash data (5 years in the before period and 5 years in the after period) are used in a before-after study with a control group. The control group consists of national and country roads that traverse towns with populations of less than 5,000. Crash occurrence was reduced for all crash severities; however, none of the differences were significant at a 95% level of confidence (see Figure 19). A more detailed analysis by Agustsson revealed that although the percentage of multivehicle crashes was reduced the percentage of single- motor-vehicle crashes increased (both significant differences at a 95% level of confidence). The likely explanation for this result is that the new road schemes present more obstacles, such as islands and bollards, for a motorist to strike. Department for Transport (2005) The U.K. Department for Transport (2005) tested rumble- wave surfaces, which are a quieter alternative to transverse

34 the before or after periods at treated sites. Property dam- age only crashes decline in some instances and increase in others. The only discernable trend is that property damage only crashes seem to increase in the year after simplification, indicating a behavior adjustment period for road users. Quimby and Castle also reported that the removal of lon- gitudinal pavement markings can be effective in lowering speeds and crash rates in certain rural road circumstances. A centerline removal plan for 30 mph (50 km/h) zones in Wiltshire, England, resulted in a 35% decrease in crashes and a 5% reduction in travel speeds. However, the authors also report that another study involving seven sites in Eng- land using a before-after with comparison group methodol- ogy showed that removing directional dividing lines did not produce any statistically significant change in crashes. Knapp and Rosales (2007) An increasingly popular form of traffic management that may be useful for rural settlements located on multilane roads is “road diets,” or lane reductions. Typically, “road diets” involve converting a four-lane road to a two-lane road, with a center turn lane and bicycle lanes on either side of the road. Knapp and Rosales (2007) aggregate the results from a number of statistically robust road diet studies, analyz- ing the effects on crash risk. The treatments were typically four-lane to three-lane conversions with varied settings, locations, and methodological approaches. Table 28 sum- marizes the details of the safety effects presented in four of the statistically robust studies. Except for the Huang et al. (2005) study, road diets appeared to produce a 20% to 40% reduction in crash risk. The articles do not provide settings for these studies, so the impressive results are not necessarily directly applicable to transition zones. None- theless, road diets are a measure that might be implemented at the downstream end of a transition zone and carried through the developed area to reinforce the lower speed limits. TABLE 27 EFFECTS OF RUMBLEWAVE SURFACES ON INJURY CRASHES Site* Before (36 months) After Reduction (%)No. of Casualty Crashes Annual Frequency Months in-service No. of Casualty Crashes Annual Frequency 2 13 4.3 33 3 1.1 –75 3 2 0.7 24 0 0 –100 4 31 10.3 23 8 4.2 –60 5 13 4.3 22 6 3.3 –24 6 5 1.3 22 1 0.5 –67 7 13 4.3 24 5 2.5 –42 Average 4.3 1.9 –55 *Site 1 crash data were not reported. Quimby and Castle (2006) As part of the Quimby and Castle (2006) investigation into the effects of simplified streetscapes on crash risk, three projects relevant to this synthesis were identified. Opeinde in the Netherlands is a large area simplification plan that included removal of pavement markings, and different curb and road surfacing which denote a public space that is shared by different modes of transport rather than one where motor- ized traffic has priority. The plan includes marking the entry points to the town with a large tubular steel arch that serves as a gateway (see Figure 20). FIGURE 20 Opeinde Gateway [Source: Quimby and Castle (2006)]. The 5-year before crash record included one fatal crash, seven injury crashes, and 24 property damage only crashes. The 3-year after crash record for this simplified streetscape involved one injury crash and five property damage only crashes. When combined with the crash data from a small number of other Dutch simplified streetscape projects, conclusions are difficult to draw. With the exception of the previously cited treatment, there have been no serious crashes in either

35 TABLE 28 STUDY RESULTS FOR ROAD DIETS Researcher Statistical Methods No. of Sites Change in Crash Risk Measure Reduction Huang et al. (2005) Before-after with yoked comparison 12 converted sites, 25 comparison sites All crashes 6% Stout and Souley- rette (2006) Before- after 14 All crashes 21% Before-af- ter with yoked comparison 14 All crashes 38% Gates et al. (2007) Empirical Bayes 7 All crashes 44% Pawlovich et al. (2007) Full Bayes 15 Crashes per mile Crash rate 25% 19% Andersson et al. (2008) Andersson et al. (2008) analyzed the safety performance of town gates in transition zones between rural and urban areas of Denmark. The area of influence for each gate was deter- mined to be 656 ft (200 m) on either side of the gate [i.e., a 1,312 ft (400 m) section of road]. A total of 251 town gates were included in the analysis, broken down into the follow- ing three categories: • Gates consisting of physical measures only (102 sites) • Gates consisting of visual measures only (40 sites) • Gates consisting of a combination of physical and visual measures (109 sites). Examples of these gateways are shown in Figure 21. Phys- ical measure gates generally consisted of a central traffic island with deflections for both directions of travel, bicycle facilities, and illumination. At just under 80% of the physical measure gates, upstream warning of the gate was provided. The visual measures gates were typically characterized by an urban zone sign placed on a special background. More than 70% of these gates were fully or partially illuminated, with 28% having special illumination for the urban zone sign. The physical and visual measures gates were charac- terized by either a speed hump or a central traffic island. Approximately 70% had an urban zone sign on a special background, and 40% had special illumination. Almost half of the physical and visual measures gates had warning signs and almost all of them were illuminated. Safety performance was estimated using 3 to 5 years of before-after crash data and a control group consisting of county and state roads in urban and rural settings, excluding motorways and highways. At 31% of the 251 sites there were no recorded crashes in the area of influence in both the before and after periods. The following crash trends were reported: • A significant increase of 34% in the number of prop- erty damage only crashes • No significant change in the number of personal injury crashes • A 100% increase in the number of single motor vehicle crashes • A 29% decrease in crossing crashes with road-users • A significant increase of 28% in urban area crashes, and a minor 6% increase in rural areas crashes (not statistically significant). FIGURE 21 Danish gateways [Source: Andersson et al. (2008)]. Of the three categories of gates, the physical and visual measures gates offered the best safety performance. The physical measures gates showed a 43% and 68% increase in personal injury and property damage only crashes, respec- tively. Faring somewhat better were the visual measures gates, which produced no change in personal injury crashes and a statistically insignificant decrease of 29% in prop- erty damage only crashes. Finally, the combined physical and visual measures gates yielded a 28% decrease in injury crashes, and a 36% increase in property damage only crashes (neither result was statistically significant). The difference between the posted speed limits in the rural and the urban areas seemed to influence safety performance, with gates on roads where the difference in posted speed limits is less than 20 mph (30 km/h) being more effective

36 DOCUMENTED PRACTICES AND GUIDANCE Australia In an early report concerning better balancing the needs of traffic movement, and the needs of village residents and businesses, Armstrong et al. (1992) advanced the concept of “entry portals” to raise driver awareness of changes in the road environment that require different driver behavior. A portal marks the beginning of an area where a different (usually lower) speed profile applies. Where the difference in the desired speed on the open road and the the village street is great, and in order to avoid any abrupt changes in speed, it is suggested that two portals be introduced. The upstream portal conditions the driver to a speed reduction and the downstream portal forces the driver to reduce speed to the level required in the village before entering the village. No evaluations or case studies are provided. Similarly, there is no specific design guidance. The most relevant piece of documentation on recommended practices for rural high- to-low speed transitions was the New Zealand Land Trans- port Safety Authority’s “Guidelines for Urban/Rural Speed Thresholds” (Land Transport Safety Authority 2002). These guidelines outline the principles to be used in the application and design of engineering treatments at urban/rural thresh- olds to promote consistency and good design practice. With respect to warrants for engineering measures at these locations, it is noted that thresholds are a potential technique only on roads that have a difference in the war- ranted speed limits of 12 mph (20 km/h) or more and when one or more of the following conditions are present: • Vehicle speeds on the approach to the settlement or through the urban areas are inappropriately high • The injury crash rates are higher than average or need to be reduced • Vulnerable road users such as pedestrians and cyclists feature in the crash analysis. In general, the guidelines touch on all of the expected factors in gateway development: placement, roadway nar- rowing, lighting, conspicuity, accommodation of cyclists or pedestrians, surface treatments, vertical deflections, land- scaping, and traffic control devices. The New Zealand guidelines recommend the dimensions shown in Figures 23–25 for effective rural/urban thresholds. Vertical deflections such as speed humps and raised cross- walks are not recommended for urban/rural thresholds. than gates at transitions of greater than 20 mph (30 km/h). Based on somewhat insufficient data, the researchers con- cluded that at gates that are only physical measures, speed humps are more effective than traffic islands, but the same is not true if the gate is also outfitted with visual measures. Veneziano et al. (2009) The state of California undertook a Gateway Monument Demonstration Program and constructed seven gateways for five communities from 2005 to 2008, inclusive. The gate- ways were freestanding structures or signage at the roadside that communicated the name of a city, county, or township to road users (see Figure 22 for an example gateway). FIGURE 22 An example gateway in California [Source: Veneziano et al. (2009)]. Crash data assembled for analysis came from at least 0.1 mile (0.2 km) upstream and downstream of the gateway monument, adjusted according to location-specific features that warrant inclusion of additional road segments. Three years of before data and 3 years of after data (if available) were used in an Empirical Bayes crash analysis. Table 29 summarizes the site data. An examination of crash number and type by the research- ers at each of the gateway sites indicated that no deterioration in safety was observed at any gateway sites. On a collective basis, the Empirical Bayes analysis showed a reduction in the total number of crashes of 2.2% to 32.0%, depending on the base safety performance function used in the analysis. The researchers concluded that these results indicate that gateways are not detrimental to safety, as opposed to being a safety benefit.

37 FIGURE 23 Minimum widths for urban/rural thresholds [Source: Land Transport Safety Authority (2002)]. TABLE 29 SUMMARY OF CALIFORNIA GATEWAY DATA Location Length (m) No. of Lanes Lane Width (ft) Shoulder Width* (ft) Before After AADT** Years of Data Total Crashes AADT** Years of Data Total Crashes Willow Creek 0.7 2 12 11 3,000 Aug 03–Jul 06 2 3,300 Aug 06–Apr 08 1 Paso Robles Rte. 46 0.6 2 12 8 17,000 Aug 98–Jul 01 0 20,000 Aug 01–Jul 04 4 Nevada County 0.4 4 12 10 27,500 Sep 03–Aug 06 5 34,000 Sep 06–Apr 08 1 Tehachapi 0.4 4 12 8 10,100 Nov 02–Oct 05 3 10,900 Nov 05–Apr 08 1 Paso Robles US 101 0.05 1 12 9 9,600 Aug 98–Jul 01 3 12,900 Aug 01–Jul 04 6 Before After Minor AADT Major AADT Years of Data Total Crashes Minor AADT Major AADT Years of Data Total Crashes Rocklin WB 2,770 25,000 May 02–Apr 05 12 3,400 29,500 May 05–Apr 08 7 Rocklin EB 11,050 24,000 May 02–Apr 05 11 11,350 28,200 May 05–Apr 08 15 [Source: Veneziano et al. (2009)]. *Average width for two sides of the roadway. **AADT (average annual daily traffic) was obtained for the middle year of each before and after period. Europe Herrstedt et al. (1993) developed a catologue of ideas on Danish traffic calming that also provides some insights and guidance on speed reductions in transition zones. To start, the authors noted that the selection of speed reduction mea- sures in any situation depends first and foremost on the target/ desired speed, and the road classification (i.e., roadway func- tion). For desired speeds of 30 mph (50 km/h) or higher, eight treatments are identified in the catalogue (see Table 30). There is no specific reference to which, if any, of these treatments might be acceptable at a rural-urban threshold. European countries have long recognized the need for transition zones on the approaches to villages as a compo- nent of an overall speed management strategy (European Transport Safety Council 1995). The European approach to these zones is founded on two principles: • Measures in transition zone must be complemented by measures along the through route within the urban area; and

38 TABLE 30 SPEED REDUCTION MEASURES FOR DESIRED SPEEDS OF 30 MPH (50 KM/H) OR HIGHER Treatment Road Class Desired Speed AADT Traffic Road Local Road ≥40 mph (60 km/h) 30 mph (50 km/h) >3,000 ≤3,000 Pre- warnings X X X X X X Gates X X X X X X 2-lane raised areas X X X X X 2-lane humps X X X X X Staggerings X X X X X X Staggerings with raised area X X X X X 2-lane nar- rowing from road center X X X X X 2-lane nar- rowing from roadside X X X X X [Adapted from Herrstedt et al. (1993)]. • The transition zone measures should achieve a cumula- tive effect culminating at a feature called the gateway to the town or village. With respect to the second principle, the European guid- ance is to influence the driver’s perception of appropriate speed by altering the physical relationship between the width of the road and the height of the nearby vertical elements such as trees and buildings. Research has shown that speeds are lower where the height of vertical elements is greater than the width of the road. In the transition zone, speed can be lowered to more acceptable levels by progressively intro- ducing road narrowing and vertical elements at the road- side. The transition zone should be terminated at a gateway, which should coincide with the threshold to the urban area. It is recommended that the gateway be the most prominent visual element in the transition zone, and visible over at least the stopping sight distance for the 85th percentile of the approach speed. The Irish National Roads Authority (2005) have devel- oped a set of traffic calming guidelines for towns and villages located on national roads that includes a specific section on transition zones. The Irish guidelines for transition zones rely heavily on the concept of “optical width”: the relation- ship of the horizontal and vertical elements of the road and FIGURE 24 Standard transition measure in New Zealand [Source: Land Transport Safety Authority (2002)]. FIGURE 25 Standard transition measure with raised median in New Zealand [Source: Land Transport Safety Authority (2002)].

39 • Prohibition of overtaking in the transition zone, using signs, solid centerlines, and gateway islands • Eliminating or reducing the hard shoulder, using cross- hatching inside the edge line to increase the visual effect • Narrowing the carriageway • Provision of rumble strips or rumble areas if speeds are not sufficiently reduced by other measures • Signs with a vertical emphasis • Use of appropriate softscape elements such as trees, shrubs, and grass boulevard treatment, which change in composition and degree of formality along the tran- sition zone into the town • Provision of cyclist and pedestrian facilities • Use of the town sign in conjunction with the area speed limit sign in the design of the gateway. The Irish guidelines also recommend the use of a gate- way at the downstream end of the transition zone to mark a change in the character of the surrounding area from rural to urban. Gateway design features include the following: • The gateway should be conspicuous, the most promi- nent element in the transition zone, and located at the downstream end of the transition zone. • The gateway should be visible over the stopping dis- tance for the 85th percentile approach speed. • The gateway should not interfere with sightlines at intersections, etc. • The gateway location should be cognizant of likely future developments. • When the gateway has been located in the field, the existing speed zones should be reviewed and changed, if necessary, so that the location of the 30 mph (50 km/h) or 35 mph (60 km/h) speed limit sign corre- sponds with the gateway. • Illumination, where provided, should extend at least two poles beyond the gateway. • Curbs on gateway islands and build-outs should be painted (yellow and black). • Direct lighting of gateway signs at gateways without a center island is optional, but has been found to be very effective, particularly on long approaches. • The road surface may be colored or textured for the length of the gateway. • Hard shoulders should, in general, be replaced with parking bays within the gateway. • A ¾ inch (2 mm) high narrow rib may be overlaid on crosshatching lines. For roadsides in transition zones, the Irish guidelines identify landscaping as an important element and promote individual treatment according to the landscape character of the area. The main roadside/landscaping design elements may include the following: the roadside. Urban and rural cross sections generally have vastly different ratios of horizontal dimension (offered by the road bed and the roadside clear zone) to the height of the vertical elements located at the roadside. Figure 26 provides an example of the different optical widths for urban and rural areas. The guidance for effective transition zones is that the optical width should be progressively reduced throughout the length of the transition zone to achieve the dominance of the vertical elements culminating in a gateway. These guidelines also advocate for a gradual change from a rural to an urban character in the transition zone. To this end, it is noted that the urban and rural environments typi- cally have the characteristics shown in Table 31. Design fea- tures that may be included in high-to-low speed transition zones include the following: FIGURE 26 Optical width [Source: Irish National Roads Authority (2005)]. TABLE 31 URBAN AND RURAL STREET CHARACTERISTICS Feature Rural Road Urban Street Boulevard grass Not mowed Mown grass Roadside vegetation Native species Evergreen ground cover Trees in the road allowance Irregular spac- ing and clumping Single or double rows of regularly spaced trees Sidewalks Absent Present

40 – Trees should be planted at 6 to 12 ft (2 to 4 m) on- center within each clump. – No tree whose girth would be expected to exceed 6 inches (150 mm) should be located any closer than 15 ft (4.5 m) from the road edge. • A single row of full standard trees may be provided at 60 ft (20 m) spacing along the grass boulevard or within the hedgerow in settings that are already urban in character. Figures 27 through 33 show examples of typical transi- tion zone landscape designs. • The grass boulevard should be maintained to a high standard over the length of the transition zone to signal a degree of formality. • Hedges, when provided, should be 5 to 6 ft (1.5 to 2.0 m) high and composed of a mix of indigenous/natural- ized shrubs (70%) and deciduous ornamental shrubs (30%) at the start of the zone, changing to an even split between deciduous ornamental shrubs and evergreen shrubs toward the end, so as to provide a higher ame- nity value in the vicinity of the built-up area. • Full standard trees should be planted in “clumps” at the back of the transition zone signs where a suitable backdrop does not exist. – Each clump should consist of three to five native or naturalized trees that integrate well into the existing landscape. FIGURE 27 Transition zone for 40 to 50 ft (12 to 16 m) right-of- way without a path [Source: National Roads Authority (2005)].

41 FIGURE 29 Transition zone for 50 to 63 ft (16 to 19 m) right- of-way [Source: National Roads Authority (2005)]. Finally, the Irish guidelines allow for the provision of rumble strips and tactile surfaces where adequate speed reductions are not being achieved in the transition zone. Two treatments are available: • A “rumble area” overlaid on to the surface with a length of 400 ft (120 m) and the last patch terminating 164 ft (50 m) from the gateway sign; and • The rumble strip installation, which consists of bars of thermoplastic material over a length of about 656 ft FIGURE 28 Transition zone for 40 to 50 ft (12 to 16 m) right-of- way with a path [Source: National Roads Authority (2005)]. The landscaping at the gateway (see Figure 34) should reinforce the vertical character of the sign and narrow the driver’s cone of vision. To achieve this: • Provide evergreen shrubs, less than 5 ft (1.5 m) high, to anchor down the sign. • Plant an upright standard tree within the shrub planting and behind the sign. A number of similar trees with a final height of 26 to 40 ft (8 to 12 m) should be planted at regular intervals inside the gateway. • Embankments may be mass planted with ground cover shrubs and a hedgerow planted along the boundary fence at the top of the embankment.

42 • More complex environments tend to produce lower operating speeds owing to increased cognitive load and perceived risk. • Natural traffic calming such as winding roads and “humpback” bridges can be very effective and more acceptable to drivers. FIGURE 31 Transition zone with an on-road cycling path [Source: National Roads Authority (2005)]. • Emphasizing changes in environment can increase awareness and/or reduce speeds. • Enclosing a distant view and/or breaking up linearity can reduce speeds. • Creating uncertainty can reduce speeds. • Combinations of measures are more effective than individ- ual measures but are most costly and visually intrusive. • Roadside activity can reduce speeds. (200 m) and installed so that it corresponds with the length of longitudinal pavement markings in the transi- tion zone. FIGURE 30 Transition zone for 63 to 69 ft (19 to 21 m) right- of-way [Source: National Roads Authority (2005)]. Kennedy (2005) reported the following broad principles that form the basis for successful psychological traffic calm- ing, which includes gateways and treatments usually imple- mented at high-to-low speed transitions:

43 • The core zone is the area of greater development and activity, which requires slower travel speeds for safety reasons; and • The transition zone, which lies between the approach zone and core area, is where drivers are expected to achieve the necessary speed reduction. FIGURE 33 Standard transition zone landscaping in a built-up, semirural area [Source: National Roads Authority (2005)]. FIGURE 34 Typical gateway landscaping [Source: National Roads Authority (2005)]. Typical devices and techniques that would be used in the approach zone are gateways, pavement markings, and rumble strips—features that highlight the change in the road Kennedy also stated that as well as being effective in managing speeds, measures need to be visually appeal- ing, particularly in historic areas and rural environments. She suggests using local building materials for gateways, and developing plans that are consistent with the colors and character of the area. Kennedy’s work reinforces the axiom that there is no single, unique, and widely accepted measure, and that each situation must be dealt with individually in a holistic manner. FIGURE 32 Standard transition zone landscaping in an open rural area [Source: National Roads Authority (2005)]. As part of a program to develop cost- and safety-ef- ficient designs for rural roads for developing countries Kirk et al. (undated) from the U.K. Transport Research Laboratory offer advice on managing speed on the approaches to roadside and ribbon development along major roads. The cost- and safety-efficient material rec- ognizes that there are three distinct zones in the rural to urban transition: • The approach zone is used to warn drivers that they are about to enter a section of road that has a higher level of development and the need to adapt driving behavior;

44 terline and the edge line markings to halt somewhere in the transition zone to communicate the downstream change in road function. The U.K. Department of Transport (2007) has distilled traffic calming research and experience into a single publi- cation that includes some design suggestions for gateways and entry treatments. Specifically, the guidelines state the following: • A gateway should be visible over at least the stopping sight distance for the 85th percentile approach speed so as not to surprise the driver. • The gateway should be visually linked to the start of the village. • Gateways should be as conspicuous as possible while remaining visually pleasing. • Gateways are only marginally enhanced by pavement markings, because markings are not visible from sig- nificant distances. • Surface treatments and road narrowings at gateways should be at least 16 ft (5 m) but no longer than 33 ft (10 m). • Physical narrowings must take into consideration heavy vehicles, agricultural vehicles, and other larger commercial vehicles. If physical narrowings cannot be achieved because of expected vehicle types, then pave- ment markings and different surface materials can be used to visually narrow the road while providing over- run areas. • Roadside features should be set back sufficiently to avoid vehicles coming in contact with these elements. Careful consideration must also be given to the conse- quences of impacting any roadside element. North America The Chesapeake Country Scenic Byway Alliance (2001) looked into the issue of rural to urban transitions as part of a corridor management plan for a National Scenic Byway. This planning document advocates that consolidating community entrance signs and reinforcing their visibility with attractive landscaping is the most direct way to convey to drivers that they are transitioning from a rural road to a settled place. The entrance/gateway signing should be large enough to be noticeable, and distinguishable from proximate commercial signs. The Alliance emphasized the need for traffic calming in and on the approaches to settled areas, and specifically mentioned the following techniques: • Making the road look narrower, through modest physi- cal changes in paving and landscaping • Encouraging roadside businesses to use landscaping rather than pavement near the roadside so as to con- solidate entrances, and mark entries • Using decorative planting at entries and around the base of welcome signs environment but do not physically slow drivers. The physi- cal changes in road geometry occur in the transition zone. After the warning in the approach zone, horizontal deflec- tions and changes in the road cross section are implemented in the transition zone to physically slow drivers down before entering the core area. Although not specific to transition zones, the Netherlands and some Scandinavian countries are employing the concept of recognizable road design (RRD) to elicit driver behavior that is more consistent with the road and its setting (SWOV 2007). RRD starts with the same basic North American premise that roads have two primary functions: mobility and access. This principle is carried forward into the road design, suggesting that each road category (designated by function) has its own design and speed limit characteristics. Further, the design and speed limit for every road in a specific cat- egory needs to be homogeneous to achieve a road design that is recognizable and elicit proper driving behavior. The initial attempts at creating a RRD system use the fol- lowing characteristics to distinguish between road categories: 1. Road surface 2. Median treatment 3. Type of edge line markings 4. (Anti) flow marking, or diagonal stripes that partly cover the lane from the edge line and/or the center- line marking. Stripes in the driving direction are a narrow-illusion marking (/ \). 5. Color and shape of curb marker posts 6. Setting characteristics such as buildings, parking spaces, and exit roads 7. Presence of on-street bicycle lanes. Of these characteristics, it is thought that the median treatment and the type of edge line markings are the essen- tial elements for RRD. It is noted that in the Netherlands, a broken or dashed edge line is used to denote a road category, whereas broken edge lines are not used in North America. The RRD principles are applicable to transition zones in that transition zones can use characteristics of the “access roads” to communicate to drivers that they are approaching a built-up area from a rural area (or a “distributor road”). In practical terms, under the Dutch guidelines the high-speed rural area (80 km/h) would have a median or a marked centerline and a broken/dashed edge line. The lower speed urban area would have neither a marked centerline nor an edge line. Therefore, it would seem appropriate for the cen-

45 In Virginia, the Thomas Jefferson Planning District Commission (2004) has developed the “Design Manual for Small Towns: Transportation and Land Use Strategies for Preserving Small Town Character,” similar to the Puget Sound document. It provides similar platitudes concerning speed management and engineering measures available, but lacks any details concerning effectiveness or specific war- rants for use. In a novel effort to lower operating speeds near schools in Needham, Massachusetts, the municipality erected non-standard traffic signs designed by middle school stu- dents (Kocian 2008). These signs resemble posters that might be found on refrigerator doors, rather than on the roadside (see Figure 35), and are intended to solicit an emotional/compassionate response to slow down. Despite the obvious shortcomings concerning the legibility of the font and the increased response time required for a non- standard sign, the notion of tapping into a motorist’s emo- tional or empathetic side to achieve reductions in speed may hold some promise for future research in rural/urban transition zones. FIGURE 35 Experimental empathetic traffic sign [Source: Kocian (2008)]. In a presentation on rural to urban transition zones, Chartier (2009) provided advice to practitioners in the form of principles and design guidelines. The overarching prin- ciple that is advocated is the concept of “optical width”: the relationship of the horizontal and vertical elements of the road and the roadside. This is similar to the advice provided by the Irish. The optical width is seen as a powerful visual cue for approaching motorists in selecting an appropriate travel speed. Lowering the optical width in the transition zone is an effective speed management measure. This may be achieved by reducing the horizontal elements (e.g., lane narrowings), increasing the vertical dimension (e.g., plant- ing appropriately sized trees closer to the pavement edge), or some combination of both. • Planting street trees continuously along the approach to a community to reinforce the transition from a rural to a settled area. As part of a larger study on connections between rural town centers in Washington State, Puget Sound Regional Council (2004) developed a toolkit of context-sensitive solutions to offer some guidance to roadway designers in providing state routes that serve their mobility function and also are an effective main street for a rural community. The Options and Innovations Toolbox presents planning and design tools, including many tools that were considered new applications that were untested. The toolbox is specifically oriented to rural corridors and their town centers. The tool- box suggests that managing speed in town centers may be assisted by considering the need for speed reductions in the planning stages of a road’s life-cycle including: • Consideration of the full corridor, not just individual segments (e.g., the transition zone or the town center). The slower speeds that are desirable in a rural town center may be more easily achieved if the overall travel time in the corridor is considered and design features are more appropriate for the setting. For example, synchronized traffic signals may be used in the town center to promote travel at a slower, consistent speed, while the design features in the rural areas should sup- port higher travel speeds between communities. • Access management that discourages the placement of accesses in the transition zone where drivers are already preoccupied with speed and path choices. • Appropriate selection of road classification and design speeds for the corridor as it passes through the com- munity. Most of what is permitted with respect to lane widths, lateral clearances, clear zones, and the like are in part determined by the design speed selected. The toolbox mentions land use planning that places appropriate businesses or uses at the edge of town, landscap- ing, and urban design guidelines as speed management con- siderations in the planning phase. The toolbox also touches on specific elements of roadway design that could be consid- ered in transition areas: • Effective transition area design requires a sequence of two or more elements to safely transition speed gradually. For example, a transition might start with a landscaped median, followed by replacing the shoulders with a curb-gutter-and- sidewalk street edge. Additionally, view-framing street trees, colored shoulders, and a gateway may be placed. • Specifically mentioned physical improvements suit- able for transition areas are colored shoulders, medi- ans, landscaping, gateways, and roundabouts or special intersections.

46 TABLE 32 SUMMARY OF LITERATURE REVIEW ON THE EFFECTIVENESS OF RURAL TRAFFIC MANAGEMENT ON SPEED Researcher No. of Study Sites Measures/Treatment Method of Study Results Van Houten and Van Houten (1987) 1 “BEGIN SLOWING HERE” sign 86 meters upstream of lower speed limit Before-after 18% to 26% reduction in the per- centage of motorists traveling over 35 mph (60 km/h) Herrstedt et al. (1993) 8 A variety of treatments on roads that run through a town, including gate- ways and measures in the town Before-after 11% reduction in mean speed; 15% reduction in motorists traveling over 35 mph (60 km/h) Pyne et al. (1995) 0—driving simulator Gateway consisting of chicane, count- down speed limit signs, and transverse markings in the village Before-after 4.2 mph reduction in mean speed, and 7.2 mph reduction in the 85th percentile speed Barker and Helliar- Symons (1997) 12 Speed roundels on the pavement surface Before-after 3 mph reduction in mean speed for villages with a 40 mph speed limit; no reduction in mean speed for vil- lages with a 30 mph speed limit 5 Countdown speed limit signs Before-after No significant reduction in mean speed County Surveyors’ Society (1994a) 11 Traffic-calming on the approach to the village Before-after 4.8 to 16 km/h reduction in 85th percentile speed 9 Traffic-calming in the village 4.8 km/h reduction in 85th percen- tile speed 4 Traffic-calming on the approach to and in the village 14.4 to 20.8 km/h reduction in 85th percentile speed County Surveyors’ Society (1994b) 23 Variety of measures Before-after 8.7 km/h speed reduction (or 21% reduction in speed) 8 Variety of measures on roads with a daily traffic volume greater than 10,000 7.5 km/h speed reduction (or 16% reduction in speed) Berger and Linauer (1998) 5 Raised medians islands that provide narrowing and deflection to approach- ing traffic Before-after 0 to 38% reduction in mean speed; 2 to 42% reduction in 85th percen- tile speed Farmer et al. (1998) 6 Speed feedback signs Before-after 4.3 mph reduction in mean speed at the end of the transition zone DETR (undated) 9 Traffic calming in villages on major roads (≥8,000 vpd) Before-after 15% to 19% reduction in mean speed; 16% to 19% reduction in 85th percentile speed Alley (2000) 0—driving simulator Various gateways Before-after Winnett and Wheeler (2002) 36 Vehicle actuated speed signs Before-after Up to an 80% change in percentage of vehicles exceeding the speed limit Hildebrand et al. (2004) 6 treatment 7 control Transitional speed zones Cross-sectional No significant impact on mean speed, percentage exceeding the speed limit, or speed variance Agustsson (2005) 21 Environmentally friendly through roads Before-after 17% reduction in mean speed, and reduction in the percentage exceed- ing the speed limit from 75% to 36% Department for Transport (2005) 7 Rumblewave surface Before-after Reductions in mean and 85th per- centile speeds from 1 km/h to 6 km/h Forbes (2006) 9 Various gateway treatments and vil- lage traffic calming Before-after 6 km/h reduction in the 85th per- centile speed Table continued on p.47

47 Table continued from p.46 TABLE 32 SUMMARY OF LITERATURE REVIEW ON THE EFFECTIVENESS OF RURAL TRAFFIC MANAGEMENT ON SPEED Researcher No. of Study Sites Measures/Treatment Method of Study Results Sandberg et al. (undated) 4 treatment and 1 control Speed feedback signs Before-after with control 6.9 mph reduction in the 85th percentile speed over 12 months Arnold and Lantz (2007) 2 Optical speed bars Before-after 3 to 9.5 mph reduction in 85th percen- tile speed over 90 days Hallmark et al. (2008) 1 Transverse pavement markings Before-after Up to a 2 mph reduction in 85th percen- tile speed 2 Transverse pavement markings with speed feedback signs 3 mph to 7 mph reduction in 85th per- centile speed 1 Lane narrowings using painted median and edge markings Mixed results on 85th percentile speed 1 Converging chevrons and 25 mph pavement marking Up to a 4 mph decrease in 85th percen- tile speed 1 Lane narrowing using edge mark- ings and 25 mph pavement markings Mixed results on 85th percentile speed 1 Speed table 4 to 5 mph decrease in 85th percentile speed 1 Lane narrowing with a median of tubular markers Up to a 3 mph decrease in 85th percen- tile speed 1 Speed feedback sign 7 mph decrease in 85th percentile speed 1 SLOW pavement legend Mixed results on 85th percentile speed 1 35 mph pavement legend with a red background Up to a 9 mph decrease in 85th percen- tile speed Dixon et al. (2008) 0—driving simulator Layered landscape Before-after with control 4.6 mph and 1.2 mph reductions in mean and 85th percentile speeds Gateway with narrowing 5.5 mph and 3.0 mph reductions in mean and 85th percentile speeds Median treatment only 3.4 mph and 0.1 mph reductions in mean and 85th percentile speeds Median with gateway 10.2 mph and 5.6 mph reductions in mean and 85th percentile speeds Medians in series with no pedes- trian crosswalks 10.7 mph reductions in mean and 85th percentile speeds Medians in series with pedestrian crosswalks 10.0 mph and 5.6 mph reductions in mean and 85th percentile speeds Jamson et al. (2008) 0—driving simulator Donnell and Cru- zado (2008) 13 Speed feedback sign located 500 feet downstream of threshold Before-after 6 mph drop in mean speeds that lasts while the speed feedback sign is in place Lamberti et al. (2009) 0—driving simulator Transverse bars, sign gantry Before-after 11 to 17 km/h reduction in the mean speeds at the gateway Chartier (2009) 1 Dragons teeth, edge lines, cen- terline, and roadside trees Before-after 10% reduction in 85th percentile speed at the threshold Russell and Goda- varthy (2010) 3 Colored pavement Before-after No significant change to a 13% to 17% reduction in mean and 85th percentile speeds 3 Solar speed display Before-after 1% to 14% reduction in 85th percentile speed; 3 to 12% reduction in mean speed 2 Mobile speed trailer Before-after 3% to 15% reduction in 85th percentile speed; 4 to 11% reduction in mean speed 3 Optical speed bars Before-after Mixed results showing speed increases and decreases

48 • Introduce cycling and pedestrian facilities • Incorporate town entry sign with area speed limit sign in design of gateway • Provide rumble strips and possibly roundabout if speeds not sufficiently reduced by other measures. The presentation also mentions the following design guidelines for gateways: • Make the gateway conspicuous and the most promi- nent element in the transition zone • Locate the gateway at the end of the transition zone Chartier provides specific design guidelines for transition zones: • Prohibit passing, using signs, solid center lines, and gateway islands • Phase out paved shoulders • Use transverse pavement markings such as crosshatch- ing, blocks, and dragon’s teeth • Narrow lane widths using edge lines • Use road side signs to increase the vertical dimension • Use soft landscape elements such as trees, shrubs, and grass boulevard treatments, which change in composi- tion and degree of formality along the transition TABLE 33 SUMMARY OF LITERATURE REVIEW ON THE EFFECTIVENESS OF RURAL TRAFFIC MANAGEMENT ON CRASH RISK Researcher No. of Study Sites Measures/Treatment Method of Study Results Herrstedt et al. (1993) 8 A variety of treatments on roads that run through a town, including gateways and measures in the town Before-after 44% and 36% reduction in casualty and all crashes, respectively County Surveyors’ Society (1994b) 20 Variety of measures Before-after 65% reduction in crash rate 8 Variety of measures on roads with a daily traffic volume greater than 10,000 47% reduction in crash rate Wheeler and Taylor (2000) 56 Variety of traffic-calming mea- sures on major roads Before-after with control group 20% to 25% reduction in casualty crashes, and 33 to 50% reduction in serious injury and fatal crashes Crowley and MacDer- mott (undated) 14 Variety of traffic-calming mea- sures on both approaches of pri- mary roads Before-after 41% reduction in casualty crashes 7 Variety of traffic-calming mea- sures on one approach of primary roads 33% reduction in casualty crashes Winnett and Wheeler (2002) 19 Vehicle-actuated speed signs on roads with 30 and 40 mph speed limits Empirical Bayes 34% reduction in casualty crashes Souleyrette et al. (2003) 141 segments Diagonal parking Cross-sectional No substantial difference between non-intersection crash rates Timesonline (2004) Unknown Psychological traffic calming (removal of signs and markings) Unknown 14% reduction in crashes Agustsson (2005) 21 Environmentally friendly through roads Before-after with control group Non-significant reduction in crashes (19% reduction in all crashes, 30% reduction in casualty crashes) Department for Trans- port (2005) 7 Rumblewave surface Before-after 24% to 100% reduction in casualty crashes, averaging 55% Forbes (2006) 5 Various rural traffic calming measures Empirical Bayes 22% reduction in all crashes, 28% reduction in casualty crashes Knapp and Rosales (2007) 62 Road diets (usually 4 to 3 lane conversions) Before-after with yoked, compari- sons, Bayes methods 19% to 44% reduction in all crashes Veneziano et al. (2009) 7 Gateway monuments Empirical Bayes 2.2% to 32% reduction in crashes

49 • Ensure forward visibility over stopping distance for 85th percentile approach speed • Do not obscure intersection sightlines • Consider likely future developments when locating • Place a reduced speed limit sign at the gateway location • Extend roadway lighting, where provided, at least two poles upstream of the gateway • Consider painting curbs on gateway islands and build- outs • Consider lighting gateway signs, particularly on long approaches • Consider coloring or texturing the roadway surface for the length of the gateway • Provide a minimum width of 16.7 ft (5.1 m) between signs at a gateway to accommodate large commercial vehicles • Make gateway signs and lighting poles in center islands demountable and frangible • Avoid sign clutter. Global The World Bank (2005) produced a manual concerning safe road design that includes a chapter on linear villages, with some limited detail concerning speed transition zones. The authors advocate that the location and layout of rural/urban speed transition zones are of critical importance and should be determined from the perspective of the road user. Determine whether there is a built-up area: • The distance from the buildings to the centerline of the road is a maximum of 3 times the height of the adjoin- ing buildings, with a maximum of 82 ft (25 m). • The length of the built-up area is at least 1,312 ft (400 m). • The building density (building frontage related to road length) for buildings on one side of the road is ≥50% and for buildings on both sides is ≥30%. Determine the location of the border: • The border should be where the setting changes (tak- ing into account the potential for short-term changes in development). • The border should be supported with new environmen- tal characteristics. • The planned location of the border is visible at the actual approach speeds. By following this guidance, it is suggested that optimum conditions are created in terms of clarity, recognition, and acceptance of the lower speed limit by road users. Such a rede- sign of the public space is the only way to ensure compliance with speed limits at the border. Examples of effective mea- sures mentioned are roundabouts, center islands, and bends and plateaus that are suitable for 30 mph (50 km/h) travel. SUMMARY OF LITERATURE REVIEW The majority of the effectiveness studies concerning transi- tion zone treatments and rural/village traffic calming have been conducted in Europe. The general conclusion is that engineering measures are effective at reducing speeds and crashes. One study in particular also revealed that public acceptance of rural settlement traffic calming is high. The effects of transition zone treatments on operating speed are generally small and are not sustained downstream of the urban/rural threshold without additional downstream measures. The reported crash reduction factors have been quite significant, although methodological shortcomings with some of the studies likely overestimate effectiveness. Nonetheless, the results are impressive enough that even if the sundry factors were accounted for, a sizeable crash reduction is a likely outcome. A summary of the study results on speed and crashes are shown in Tables 32 and 33, respectively. In addition to the evaluation studies, some general trends and advice may be garnered from the reviewed studies and design guidelines: • More extensive and aggressive measures tend to pro- duce greater reductions in speed and crash occurrence than less extensive and passive measures. • There needs to be a distinct relationship between a settlement speed limit and a change in the roadway character. • No one particular measure is appropriate for all situ- ations. Each settlement must be assessed and treated based on its own characteristics and merits. • To maintain a speed reduction downstream of the transition zone, it is necessary to provide additional measures through the village. Otherwise, speeds may rebound to previous levels as soon as 820 ft (250 m) from the start of the lower speed zone. Some jurisdictions in Europe are experimenting with shared spaces, which include removing traffic control and physical separations between road users. This approach to speed management is a paradigm shift in thinking, whereby guidance and direction to the motorist is removed from the street, and drivers are required to exercise additional caution and take more responsibility for their own driving behav- ior. At present, this approach does not provide any specific guidance concerning speed transition areas and whether this approach is suitable in these critical zones. At any rate, the shared space concept is in fairly limited deployment and the results should be considered unreliable at present.

Next: CHAPTER THREE Survey Results »
Speed Reduction Techniques for Rural High-to-Low Speed Transitions Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Synthesis 412: Speed Reduction Techniques for Rural High-to-Low Speed Transitions explores techniques for lowering traffic speeds in rural transition zones. Transition zones are those portions of high-speed roads that have lower posted speed limits as the roadway approaches a settlement.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!