Field Evaluation of Reflected Noise from a Single Noise Barrier (2018)

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.

97 Comparison Finding 2: Differences in Broadband Levels Opposite the Barrier at Different Microphone Heights Broadband levels averaged slightly higher opposite the I-70 sound-absorbing barrier compared to the No-Barrier site for both microphone heights, similar to the sound-reflecting barriers. For the sound-absorbing barrier at I-70, the community microphones showed slightly higher levels opposite the Bar- rier compared to the No-Barrier site. This result is similar to that seen with the sound-reflecting barriers at I-24 and MD-5. Comparison Finding 2 discusses the situation in which the community microphones were positioned at the same dis- tances from the road but at different heights above the road (I-24, MD-5, and I-70 locations). Comparison Finding 3 dis- cusses the situation in which the community microphones were positioned at different distances from the road. To review, for the I-24 sound-reflecting barrier, Figure 61 (in Chapter 5) showed differences in the unweighted and A-weighted levels for BarCom03 and NoBarCom05 at the lower height. For most of the running 5-minute Leq periods, the unweighted and A-weighted BarCom03 levels are higher than the NoBarCom05 levels by a range of 0 dB to 1 dB, with differences of as much as 1.5 dB. Figure 96 shows the differences for the I-24 upper micro- phones at BarCom04 and NoBarCom06. For most of the run- ning 5-minute Leq periods, the BarCom04 unweighted levels are higher than those at NoBarCom06 by a range of 0 dB to as much as 1.1 dB, with some instances lower than at NoBarCom06. The A-weighted levels at BarCom04 mostly range from 0.4 dB lower to 1 dB higher than the levels at NoBarCom06. At the MD-5 sound-reflecting barrier, Figure 63 (in Chap- ter 5) showed differences in the unweighted and A-weighted levels for BarCom03 and NoBarCom05 (lower microphones). At the lower microphones opposite the barrier, the daytime unweighted running Leq (5 min.) at BarCom03 ranged from 1 dB lower to 2 dB higher than those at NoBarCom05. Day- time A-weighted levels at BarCom03 generally ranged from 0.5 dB to 2.4 dB higher than those at NoBarCom05. Eve- ning unweighted levels at the two microphones generally were within –2 dB to 1.5 dB of each other. The BarCom03 A-weighted levels ranged mostly from 0 dB to 1.5 dB higher than the NoBarCom05 levels. Figure 97 shows the differences in the unweighted and A-weighted levels for the upper microphones at MD-5 (BarCom04 and NoBarCom06). For most of the running Leq (5 min.) periods during both the afternoon and evening, the BarCom04 levels were higher than the NoBarCom06 levels. The unweighted Leq (5 min.) generally ranged from 0.5 dB lower to 1.5 dB higher at BarCom04 during the day and –1 dB lower to 2 dB higher at BarCom04 during the evening. The A-weighted levels ranged from 0.5 dB lower to 1 dB higher at BarCom04 than at NoBarCom06 during both daytime and nighttime. For the I-70 sound-absorbing location, Figure 98 shows the differences in unweighted and A-weighted levels for the lower microphone pair of BarCom03 and NoBarCom05. For nearly all the running 5-minute Leq periods, the BarCom03 -3 -2 -1 0 1 2 3 14 :3 5 14 :4 5 14 :5 5 15 :0 5 15 :1 5 15 :2 5 15 :3 5 15 :4 5 15 :5 5 16 :0 5 16 :1 5 16 :2 5 16 :3 5 16 :4 5 16 :5 5 17 :0 5 17 :1 5 17 :2 5 17 :3 5 17 :4 5 17 :5 5 18 :0 5 18 :1 5 18 :2 5 D iff er en ce in L ev el , d B Time dBA dBZ Figure 95. I-70—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarRef01 minus NoBarRef02.

-3 -2 -1 0 1 2 3 13 :1 3 13 :2 2 13 :3 1 13 :4 0 13 :4 9 13 :5 8 14 :0 7 14 :1 6 14 :2 5 14 :3 4 14 :4 3 14 :5 2 15 :0 1 15 :1 0 15 :1 9 15 :2 8 15 :3 7 15 :4 6 15 :5 5 16 :0 4 16 :1 3 16 :2 2 16 :3 1 16 :4 0 16 :4 9 16 :5 8 17 :0 7 D iff er en ce in L ev el , d B Time dBA dBZ Figure 96. I-24—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarCom04 minus NoBarCom06. -4 -3 -2 -1 0 1 2 3 4 12 :0 0 12 :1 7 12 :3 4 12 :5 1 13 :0 8 13 :2 5 13 :4 2 13 :5 9 14 :1 6 14 :3 3 14 :5 0 15 :0 7 15 :2 4 15 :4 1 15 :5 8 19 :5 1 20 :0 8 20 :2 5 20 :4 2 20 :5 9 21 :1 6 21 :3 3 21 :5 0 22 :0 7 22 :2 4 22 :4 1 22 :5 8 23 :1 5 23 :3 2 D iff er en ce in L ev el , d B Time dBA dBZ Figure 97. MD-5—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarCom04 minus NoBarCom06. -3 -2 -1 0 1 2 3 14 :3 5 14 :4 5 14 :5 5 15 :0 5 15 :1 5 15 :2 5 15 :3 5 15 :4 5 15 :5 5 16 :0 5 16 :1 5 16 :2 5 16 :3 5 16 :4 5 16 :5 5 17 :0 5 17 :1 5 17 :2 5 17 :3 5 17 :4 5 17 :5 5 18 :0 5 18 :1 5 18 :2 5 D iff er en ce in L ev el , d B Time dBA dBZ Figure 98. I-70—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarCom03 minus NoBarCom05.

99 unweighted sound pressure levels range from 0.3 dB to 1.7 dB higher than those at NoBarCom05, averaging approximately 0.7 dB higher. The A-weighted sound levels at BarCom03 range from 0 dB to 1.2 dB higher than those at NoBarCom05, averaging approximately 0.7 dB higher. Figure 99 shows the differences in unweighted and A-weighted levels for the upper microphone pair of Bar- Com04 and NoBarCom06 at the I-70 location. For nearly all the running 5-minute Leq periods, the BarCom04 unweighted sound pressure levels range from 0.5 dB to 2.0 dB higher than those at NoBarCom06, averaging approximately 1.1 dB higher. The A-weighted sound levels at BarCom04 range mostly from 0.5 dB to 1.5 dB higher than those at NoBarCom06, averaging approximately 1 dB higher. Comparison Finding 3: Differences in Broadband Levels Opposite the Barrier at Different Microphone Distances Broadband levels averaged slightly higher opposite the I-75 sound-absorbing barrier compared to the No-Barrier site for both microphone distances. The reflection effect is less than at the sound-reflecting barriers at the farther distances. At the two sound-reflecting locations (I-90 and SR-71) and at the sound-absorbing location (I-75), the community microphones showed slightly higher levels opposite the bar- rier compared to those at the No-Barrier site. At the microphones closest to the road for the sound- reflecting SR-71 barrier, the Barrier/No-Barrier differences are small because the microphones are close to the road. At the sound-reflecting I-90 barrier, the levels at the closer Bar- rier microphone are consistently higher than the levels at the closer No-Barrier microphone. For the sound-absorbing I-75 barrier, the differences at the closer microphones are more variable, sometimes showing higher levels at the Barrier site and sometimes showing higher levels at the No-Barrier site. At the more distant microphones, the Barrier levels generally were higher than the No-Barrier levels for the sound-reflecting barriers and, to a lesser extent, slightly higher for the sound- absorbing barrier. Details follow by location. For the sound-reflecting I-90 barrier, Figure 62 previously showed differences in the unweighted and A-weighted lev- els for the microphones closer to the road (BarCom03 and NoBarCom05). For all the running 5-minute Leq periods, the BarCom03 unweighted sound pressure levels are on the order of 0 dB to 1.5 dB higher than the NoBarCom05 levels. For the A-weighted sound levels, the BarCom03 levels are on the order of 0.4 dB to 1.3 dB higher than the NoBarCom05 levels. Figure 100 shows the differences for I-90 at the more distant BarCom04 and NoBarCom06 microphones. For most of the running 5-minute periods, the unweighted and A-weighted BarCom04 Leq levels are higher than those at NoBarCom06. The unweighted levels at BarCom04 range from 0.7 dB lower than NoBarCom06 to 1.5 dB higher. The A-weighted levels at BarCom04 range from 0.2 dB to 1 dB higher. As time passed over the 4-hour period at the I-90 location, the Leq (5 min.) (not shown) dropped slowly, on the order of 1 dB to 2 dB. During this same period, the differences in measured sound levels between the Barrier and No-Barrier microphone pairs increased by 0.5 dB. For the SR-71 sound-reflecting barrier, Figure 57 (in Chap- ter 5) showed little difference in Leq between the BarCom03 -3 -2 -1 0 1 2 3 14 :3 5 14 :4 5 14 :5 5 15 :0 5 15 :1 5 15 :2 5 15 :3 5 15 :4 5 15 :5 5 16 :0 5 16 :1 5 16 :2 5 16 :3 5 16 :4 5 16 :5 5 17 :0 5 17 :1 5 17 :2 5 17 :3 5 17 :4 5 17 :5 5 18 :0 5 18 :1 5 18 :2 5 D iff er en ce in L ev el , d B Time dBA dBZ Figure 99. I-70—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarCom04 minus NoBarCom06.

100 levels at BarCom04 range mostly from 0 dB to 4 dB higher and the A-weighted levels at BarCom04 ranged from 1 dB to nearly 4 dB higher. For both unweighted and A-weighted cases, the average difference is 2.1 dB higher at BarCom04 than at NoBarCom06. For the I-75 sound-absorbing barrier, Figure 101 shows the differences in the unweighted and A-weighted levels for the microphones closer to the road (BarCom03 and NoBarCom05). At BarCom03, the unweighted levels range from 0.9 dB lower to 1.5 dB higher than at NoBarCom05, averaging approximately 0.3 dB higher. The A-weighted levels at BarCom03 range from 0.9 dB lower to 1.4 dB higher than at NoBarCom05, averaging approximately 0.5 dB higher. -3 -2 -1 0 1 2 3 13 :0 0 13 :1 0 13 :2 0 13 :3 0 13 :4 0 13 :5 0 14 :0 0 14 :1 0 14 :2 0 14 :3 0 14 :4 0 14 :5 0 15 :0 0 15 :1 0 15 :2 0 15 :3 0 15 :4 0 15 :5 0 16 :0 0 16 :1 0 16 :2 0 16 :3 0 16 :4 0 16 :5 0 17 :0 0 17 :1 0 17 :2 0 D iff er en ce in L ev el , d B Time dBA dBZ Figure 100. I-90—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarCom04 minus NoBarCom06. -3 -2 -1 0 1 2 3 17 :1 2 17 :2 2 17 :3 2 17 :4 2 17 :5 2 18 :0 2 18 :1 2 18 :2 2 18 :3 2 18 :4 2 18 :5 2 19 :0 2 19 :1 2 19 :2 2 19 :3 2 19 :4 2 19 :5 2 20 :0 2 20 :1 2 20 :2 2 20 :3 2 20 :4 2 20 :5 2 21 :0 2 D iff er en ce in L ev el , d B Time dBA dBZ Figure 101. I-75—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarCom03 minus NoBarCom05. and NoBarCom05 microphones. These microphones were positioned just off the roadway shoulder and close to the traf- fic on the lanes opposite the barrier. The unweighted levels at BarCom03 ranged from 0.9 dB lower to 1 dB higher than the unweighted levels at NoBarCom05. The A-weighted levels at BarCom03 ranged from 0.7 dB lower to 0.5 dB higher than the A-weighted levels at NoBarCom05. Figure 73, also in Chapter 5, illustrated the differences for the considerably more distant BarCom04 and NoBarCom06 microphones at the SR-71 location. For virtually all the run- ning 5-minute Leq periods, the unweighted and A-weighted BarCom04 levels were higher than the NoBarCom06 levels. Compared to the levels at NoBarCom06, the unweighted

101 Figure 102 shows the differences for the more distant BarCom04 and NoBarCom06 microphone positions at I-75. At BarCom04, the unweighted levels range from 0.5 dB lower to 1.9 dB higher than the levels at NoBarCom06, averag- ing approximately 0.6 dB higher. The A-weighted levels at BarCom04 range mostly from 0.8 dB lower to 1.4 dB higher than those at NoBarCom06, averaging approximately 0.5 dB higher. Comparison Finding 4: One-Third Octave Band Differences at Different Microphone Heights One-third octave band levels for equivalent Leq (5 min.) periods were slightly higher opposite the I-70 sound-absorbing barrier compared to the No-Barrier site. The effect was less at the higher microphone. The differences are similar to the sound- reflecting barriers. A change in ground effects could be the cause of the larger mid-frequency difference at the lower microphone. The data analysis for the FHWA Method compared one-third octave band levels at Barrier and equivalent No- Barrier sites. As described in the second sound-reflecting barrier finding, the differences in Barrier and No-Barrier levels are frequency specific and vary by location and site. For sound-reflecting barriers, clear examples exist of enhanced levels opposite the Barrier compared to the cor- responding No-Barrier position. The slightly higher levels seen across the roadway from the sound-reflecting barriers compared to the No-Barrier site also were seen for the sound-absorbing barriers. Comparison Finding 4 discusses the I-24, MD-5, and I-70 locations at which the community microphones were posi- tioned at the same distances from the road but at different heights above the road. Comparison Finding 5 discusses the I-90, SR-71, and I-75 locations at which the community micro- phones were positioned at different distances from the road. At I-24 and MD-5 (two sound-reflecting locations) and at I-70 (a sound-absorbing location), the community micro- phones showed slightly higher levels opposite the barrier compared to the No-Barrier site. The upper graph of Figure 103 shows the differences in sound levels recorded at the reference microphones (BarRef01 and NoBarRef02) at I-24 for the Upwind Lapse groups. In general, the BarRef01 sound levels are approximately 0.9 dB to 1.3 dB higher than the NoBarRef02 levels across the entire frequency spectrum. Higher levels at BarRef01 are expected because the microphone is positioned between the barrier and the road. At 25 Hz, the difference is 2 dB; at 200 Hz and 250 Hz, it is approximately 0.5 dB. The middle graph of Figure 103 shows the differences in measured sound levels between the lower microphones (BarCom03 and NoBarCom05), both of which were posi- tioned 5 ft. above the roadway elevation. In general, the BarCom03 levels are equal to or slightly greater than the NoBarCom05 levels over most of the frequency range through 4 kHz. The increase is less than 1 dB from 31.5 Hz to 250 Hz, and on the order of 1 dB to 2 dB in the frequency bands from 315 Hz to 1 kHz. Above 4 kHz, the levels at NoBarCom05 are higher than the levels at BarCom03, which was attributed to insect noise in the vegetation behind the NoBarCom05 microphone that was not present near the BarCom03 site. -3 -2 -1 0 1 2 3 17 :1 2 17 :2 2 17 :3 2 17 :4 2 17 :5 2 18 :0 2 18 :1 2 18 :2 2 18 :3 2 18 :4 2 18 :5 2 19 :0 2 19 :1 2 19 :2 2 19 :3 2 19 :4 2 19 :5 2 20 :0 2 20 :1 2 20 :2 2 20 :3 2 20 :4 2 20 :5 2 21 :0 2 D iff er en ce in L ev el , d B Time dBA dBZ Figure 102. I-75—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarCom04 minus NoBarCom06.

102 -4 -3 -2 -1 0 1 2 3 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All ULG Groups -4 -3 -2 -1 0 1 2 3 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All ULG Groups -4 -3 -2 -1 0 1 2 3 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All ULG Groups Figure 103. I-24—averages of the differences in sound pressure levels (dB) for all Upwind Lapse groups, Leq (5 min.) é one standard deviation, BarRef01 minus NoBarRef02 (top), BarCom03 minus NoBarCom05 (middle), and BarCom04 and NoBarCom06 (bottom).

103 The lower graph of Figure 103 compares the levels at upper microphones BarCom04 and NoBarCom06, both of which were positioned 15 ft. above the roadway surface. The results show that the sound levels at BarCom04 are equal to or slightly higher than the levels at NoBarCom06 in the frequency bands from 20 Hz through 1.25 kHz. At 31.5 Hz to 63 Hz, the lev- els at BarCom04 are approximately 1 dB higher than those at NoBarCom06. At 1.6 kHz and above, the NoBarCom06 levels are higher than the BarCom04 levels, ranging from a frac- tion of 1 dB at 1.6 kHz to 2.5 dB in the 6.3 kHz band. The higher levels at NoBarCom06 in the high-frequency bands were attributed to insect noise in the vegetation behind this microphone. For the MD-5 sound-reflecting barrier, Figure 104 shows the results for the Downwind Neutral meteorological class. As a point of reference, the upper graph shows that, in gen- eral, the BarRef01 levels at this location vary little compared to the NoBarRef02 levels from 500 Hz through 6.3 kHz. Below 500 Hz, the BarRef01 levels generally are less than 1 dB higher than those at NoBarRef02. Little difference was expected, because at MD-5 the BarRef01 microphone was positioned atop the barrier (unlike I-24 and I-70, where the BarRef01 microphones were positioned in front of the barrier). The middle graph in Figure 104 compares the levels at the lower-height microphones (BarCom03 and NoBarCom05). Through 125 Hz, any differences are less than 0.5 dB. Between 160 Hz and 500 Hz, the BarCom03 levels are higher, ranging from 1 dB higher (at 160 Hz and 500 Hz) and peaking at 6 dB higher (at 315 Hz). From 630 Hz through 6.3 kHz, the BarCom03 levels exceed those of NoBarCom05 by a range of 0.5 dB to 1.5 dB (at 2 kHz). The barrier reflection effect is prominent in the low-frequency range (250–500 Hz), as was also seen in the results for the same microphones at the I-90 location. In both cases, a possible explanation is that direct and reflected sound take different propagation paths. The direct sound at both the Barrier and No-Barrier sites likely experiences ground effects/wave interference that cause a dip in sound level in that frequency range. Because the reflected sound at the barrier site experiences a differ- ent propagation path, it also experiences different ground effects and wave interference with ground reflections, such that a dip in the 250–500 Hz range would be nonexistent or diminished. This effect was not evident at the I-24 sound- reflecting site, which could be related to the presence of the median barrier on I-24 and the greater setback of the barrier from the road compared to MD-5. The lower graph in Figure 104 compares the levels at the higher BarCom04 and NoBarCom06 microphone posi- tions. The low-frequency levels are higher for BarCom04, but only by a maximum of 2 dB at 160 Hz. Across the rest of the frequency spectrum, the BarCom04 level is slightly higher than the NoBarCom06 level, with the greatest dif- ference, 1 dB, occurring at 400 Hz and 6.3 kHz. At 315 Hz the difference is 0 dB, compared to 6 dB at the lower-height microphones. A possible explanation is that the propaga- tion effects caused by ground reflections are minimized at the higher microphones because of the greater height above the ground. For the sound-absorbing I-70 barrier, the upper graph in Figure 105 shows that, in general, the BarRef01 levels are less than 1 dB higher than the NoBarRef02 levels across most of the frequency spectrum from 80 Hz through 3.15 kHz. Below 63 Hz, the BarRef01 levels are from 1 dB to 3 dB higher than the NoBarRef02 levels, and above 3.15 kHz, the BarRef01 lev- els are 1.2 dB to 2 dB higher than the NoBarRef02 levels. The field notes offer no explanation for the larger differences in the low- and high-frequency bands. The middle graph in Figure 105 shows the differences in measured sound levels between BarCom03 and NoBarCom05, both of which were positioned 5 ft. above the ground and 11 ft. above the roadway elevation. Over most of the frequency spectrum between 80 Hz and 5.0 kHz, the BarCom03 levels are 0 dB to 1 dB higher than the NoBarCom05 levels. The exceptions are at 315 Hz and 400 Hz, where the BarCom03 levels are approximately 3 dB higher than at NoBarCom05 and at 200 Hz and 250 Hz, where the BarCom03 levels are lower than the NoBarCom05 levels by 0.5 dB to 1 dB. The site conditions were similar at the BarCom03 and NoBarCom05 microphone locations, providing no clear explanation for the effects in the 200 Hz to 400 Hz bands or the effects at the low and high ends of the spectrum. One possible explanation, as also described for MD-5, is a change in the ground effects and propagation paths that go with moving up to the higher microphone. Like MD-5, the I-70 location had no median barrier The lower graph in Figure 105 compares the levels at BarCom04 and NoBarCom06, which were positioned 10 ft. above the ground and 16 ft. above the roadway elevation. Over most of the frequency spectrum between 50 Hz and 4.0 kHz, the BarCom04 levels are 0 dB to 1.1 dB higher than the NoBarCom06 levels. The exceptions are from 160 Hz to 315 Hz, where the BarCom04 levels range from 1.5 dB to 2.4 dB higher and from 3.15 Hz to 6.3 kHz, where the BarCom04 levels are lower than the NoBarCom06 levels by approximately 0.7 dB. Above 4 kHz, the BarCom04 levels are 1.5 dB to 5.0 dB higher than at NoBarCom06. No specific reason is apparent for the substantial differences at the highest frequencies; however, in context, the differences are unimportant, because at both sites the sound pressure levels at these frequencies are much lower than they are at the middle frequencies.

104 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All DNG Groups -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All DNG Groups -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All DNG Groups Figure 104. MD-5—averages of the differences in sound pressure levels (dB) for all Downwind Neutral groups, Leq (5 min.) é one standard deviation, BarRef01 minus NoBarRef02 (top), BarCom03 minus NoBarCom05 (middle), and BarCom04 and NoBarCom06 (bottom).

105 -4 -3 -2 -1 0 1 2 3 4 5 6 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CLG Groups -4 -3 -2 -1 0 1 2 3 4 5 6 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CLG Groups -4 -3 -2 -1 0 1 2 3 4 5 6 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CLG Groups Figure 105. I-70—averages of the differences in sound pressure levels (dB) for all Calm Lapse groups, Leq (5 min.) é one standard deviation, BarRef01 minus NoBarRef02 (top), BarCom03 minus NoBarCom05 (middle), and BarCom04 and NoBarCom06 (bottom).

106 Comparison Finding 5: One-Third Octave Band Differences at Different Microphone Distances One-third octave band levels are slightly higher opposite the I-75 sound-absorbing barrier compared to the No-Barrier site for the microphones at different distances from I-75. The differences at the larger distance were less than the differences at the sound-reflecting barriers. At I-90, SR-71, and I-75, community microphone pairs were located at different distances from the roadway (and different heights above the roadway). In general, the I-90 and SR-71 sound-reflecting barrier results showed greater differences between Barrier/No-Barrier levels at the community micro- phones than did the I-75 sound-absorbing barrier results. For the I-90 sound-reflecting barrier, Figure 106 shows the averages of the differences in the Barrier and No-Barrier microphone levels for all the Calm Neutral groups. In the upper graph, in general, the BarRef01 levels are 0 dB to 1 dB higher than the NoBarRef02 levels at 400 Hz and below. Above 400 Hz and through 3.15 kHz, the BarRef01 levels are 0.5 dB to 1 dB higher than the NoBarRef02 levels; however, above 4 kHz, the No-Barrier levels are higher, likely due to localized insect noise. The middle graph in Figure 106 compares the levels at the lower-height microphones (BarCom03 and NoBarCom05), both of which were located 69 ft. from the center of the near travel lane of I-90 and positioned 10.4 ft. above the road- way surface. The levels in the frequency bands from 20 Hz through 80 Hz are 0.5 dB to 1 dB higher at BarCom03 than at NoBarCom05. For frequency bands 1 kHz and higher, the BarCom03 levels are approximately 1 dB to 2 dB higher than those at NoBarCom05. The most noticeable differences are in the 250 Hz to 500 Hz bands, in which the levels range from 2.5 dB to 5 dB higher, peaking at 400 Hz. As noted in the discus- sion of the MD-5 results, a possible explanation for the barrier effect being prominent at I-90 for BarCom03 in the low- frequency range (250 Hz to 500 Hz) is that the direct and reflected sounds take different propagation paths with different ground effects and wave interference with ground reflections. The lower graph in Figure 106 compares the levels at the higher BarCom04 and NoBarCom06 microphones, both of which were located 93 ft. from the center of the near travel lane of I-90 and positioned 17 ft. above the roadway surface. The levels in the frequency bands from 31.5 Hz through 250 Hz are 0.5 dB to 1 dB higher at BarCom04 than at NoBarCom06. From 1.25 kHz to 3.15 kHz, the levels at BarCom04 are approximately 0.5 dB higher than those at NoBarCom06. The most noticeable differences occur in the 315 Hz to 630 Hz frequency bands, where the levels range from 1.5 dB to 3 dB higher, peaking at 400 Hz. Again, as at MD-5, the smaller differences at these mid-frequencies com- pared to the lower (and, in this case, closer) microphones could be attributable to differences in the propagation path to the more elevated microphone. For the SR-71 sound-reflecting barrier, Figure 107 shows the averages of the differences in the Barrier and No-Barrier microphone levels for all the Downwind Neutral groups. For reference, the upper graph shows that the BarRef01 levels are higher than the NoBarRef02 levels across almost the entire spectrum, except for 8 kHz and 10 kHz. Higher levels were expected at the BarRef01 microphone given its location in front of the barrier. The BarRef01 levels are higher by 3 dB at 31.5 Hz, 2.5 dB at 125 Hz, and 1.5 dB at 2.5 kHz. In the range from 400 Hz to 1.25 kHz, the differences are less than 0.5 dB. The middle graph in Figure 107 compares the levels at the BarCom03 and NoBarCom05 microphones, which were located close to SR71 on the opposite side of the road from the barrier. The levels at BarCom03 are generally the same or higher than those at NoBarCom05. In the frequency bands from 20 Hz through 125 Hz, the levels are 0.5 dB to 1.5 dB higher at BarCom03 than at NoBarCom05. From 160 Hz through 1.6 kHz, the levels differ only by 0.5 dB or less. From 2 kHz through 5 kHz, the levels at BarCom03 are 0.5 dB higher than those at NoBarCom05. At and above 6.3 kHz, the NoBarCom05 levels are less than 1.5 dB higher than those at BarCom03. The lower graph in Figure 107 compares the levels at the distant BarCom04 and NoBarCom06 positions for SR-71. In the frequency bands from 20 Hz through 80 Hz, the levels at BarCom04 are 2 dB to 4 dB higher compared to NoBarCom06. From 315 Hz through 8 kHz, the BarCom04 levels are 1.5 dB to 3 dB higher than those at NoBarCom06. From 100 Hz through 250 Hz, however, the NoBarCom04 levels range from 0 dB to 3 dB lower than the BarComp06 levels, with the difference peaking at 200 Hz. For the I-75 sound-absorbing barrier, Figure 108 shows the results for an average of all the Calm Inversion groups. For reference, the upper graph shows the BarRef01 lev- els to be approximately 0.6 dB to 2.7 dB higher than the NoBarRef02 levels across the entire spectrum, except at 20 Hz, where the difference is only 0.2 dB. From 80 Hz to 400 Hz, the difference is 2 dB or more; from 500 Hz to 1,600 Hz, the difference is 1 dB or less. These bands are the frequen- cies for which the sound-absorbing material would be expected to be most effective. At frequencies above 2 kHz, the difference is over 1.5 dB. The middle graph in Figure 108 shows the differences in measured sound levels between BarCom03 and NoBarCom05, both of which were positioned 10 ft. above the I-75 roadway elevation. Through 1.25 kHz, the BarCom03 levels are within ± 1 dB of the NoBarCom05 levels (being slightly lower than

107 -6 -4 -2 0 2 4 6 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CNG Groups -6 -4 -2 0 2 4 6 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CNG Groups -6 -4 -2 0 2 4 6 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CNG Groups Figure 106. I-90—averages of the differences in sound pressure levels (dB) for all Calm Neutral groups, Leq (5 min.) é one standard deviation, BarRef01 minus NoBarRef02 (top), BarCom03 minus NoBarCom05 (middle), and BarCom04 and NoBarCom06 (bottom).

108 -8 -6 -4 -2 0 2 4 6 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All DNG Groups -8 -6 -4 -2 0 2 4 6 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All DNG Groups -8 -6 -4 -2 0 2 4 6 8 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All DNG Groups Figure 107. SR-71—averages of the differences in sound pressure levels (dB) for all Downwind Neutral groups, Leq (5 min.) é one standard deviation, BarRef01 minus NoBarRef02 (top), BarCom03 minus NoBarCom05 (middle), and BarCom04 and NoBarCom06 (bottom).

109 -4 -3 -2 -1 0 1 2 3 4 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CIG Groups -4 -3 -2 -1 0 1 2 3 4 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CIG Groups -4 -3 -2 -1 0 1 2 3 4 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CIG Groups Figure 108. I-75—averages of the differences in sound pressure levels (dB) for all Calm Inversion groups, Leq (5 min.) é one standard deviation, BarRef01 minus NoBarRef02 (top), BarCom03 minus NoBarCom05 (middle), and BarCom04 minus NoBarCom06 (bottom).

110 NoBarCom05 at 125 Hz and 160 Hz). At 1.6 kHz and 2 kHz, the BarCom03 levels are 1.3 dB and 1.0 dB higher than the NoBarCom05 levels, whereas above 3.15 kHz the NoBar- Com05 levels are slightly higher than the BarCom03 levels. In any case, at these higher frequencies the differences are much smaller than at the middle frequencies. The lower graph in Figure 108 compares the levels at BarCom04 and NoBarCom06, both of which were posi- tioned 20 ft. above the I-75 roadway surface and farther away than BarCom03 and NoBarCom05. Through 2.0 kHz, the BarCom04 levels are from 0 dB to 1.1 dB higher than the levels at NoBarCom06. From 2.5 kHz to 5.0 kHz, the BarCom04 lev- els are 0.5 dB to 1.1 dB higher than the levels at NoBarCom06; however, above 6.3 kHz, the BarCom04 levels are 0.8 dB to 1.0 dB lower than the levels at NoBarCom06. No insect noise was present at the NoBarCom06 location, so the cause behind the higher levels at some of the higher frequencies is not clear. Comparison Finding 6: Broadband L90 and L99 Statistical Descriptors Broadband L90 and L99 statistical descriptors show that—unlike at the sound-reflecting barriers—the background noise levels at both microphone heights across from the I-70 sound-absorbing barrier are not elevated above the Leq. In addition to examining the differences in measured sound levels based on equivalent pairs of running 5-minute Leq data, the research team investigated differences in the Ln statistical descriptors for the overall A-weighted sound level data (without segregation into equivalent periods). The focus was on the possible change in the background level in the presence of the noise barrier. At the I-24 (sound-reflecting barrier), MD-5 (sound- reflecting barrier), and I-70 (sound-absorbing barrier) loca- tions, the community microphones were located at the same distances from the road but positioned at different heights above the road. Evidence of elevated background levels was strong for the I-24 reference microphone position in front of the barrier (as was seen in Figure 48), but background levels opposite the I-24 barrier were found to be elevated to a lesser degree. Background levels opposite the MD-5 barrier also were elevated for the daytime sampling, whereas the night- time background at the MD-5 No-Barrier site was dominated by frog noise. Mixed evidence of elevated background noise was found opposite the sound-absorbing I-70 barrier. Figure 109 presents the differences between BarCom03 and NoBarCom05 (the lower microphones across from the I-24 barrier), computed as BarCom03 minus NoBarCom05. The L90 results generally appear higher at the barrier site, whereas the L99 results are mixed. Not much difference is expected, given the dominance of the direct sound from the close-by traffic and the possible shielding of far-lane noise and barrier reflections by the median barrier. Figure 110 presents the differences between BarCom04 and NoBarCom06 (the upper microphones across from the I-24 barrier), computed as BarCom04 minus NoBarCom06. More evidence exists of elevated background noise levels at BarCom04 compared to BarCom03, likely because the ele- vated position of the microphone means less shielding of the reflected noise by the median parapet. For the MD-5 sound-reflecting barrier, Figure 111 pre- sents the differences for BarCom03 and NoBarCom05 (the -5 -4 -3 -2 -1 0 1 2 3 4 13 :1 3 13 :2 2 13 :3 1 13 :4 0 13 :4 9 13 :5 8 14 :0 7 14 :1 6 14 :2 5 14 :3 4 14 :4 3 14 :5 2 15 :0 1 15 :1 0 15 :1 9 15 :2 8 15 :3 7 15 :4 6 15 :5 5 16 :0 4 16 :1 3 16 :2 2 16 :3 1 16 :4 0 16 :4 9 16 :5 8 17 :0 7 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 109. I-24—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom03 minus NoBarCom05.

111 -4 -3 -2 -1 0 1 2 3 4 5 13 :1 3 13 :2 2 13 :3 1 13 :4 0 13 :4 9 13 :5 8 14 :0 7 14 :1 6 14 :2 5 14 :3 4 14 :4 3 14 :5 2 15 :0 1 15 :1 0 15 :1 9 15 :2 8 15 :3 7 15 :4 6 15 :5 5 16 :0 4 16 :1 3 16 :2 2 16 :3 1 16 :4 0 16 :4 9 16 :5 8 17 :0 7 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 110. I-24—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom04 minus NoBarCom06. -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 12 :0 0 12 :1 7 12 :3 4 12 :5 1 13 :0 8 13 :2 5 13 :4 2 13 :5 9 14 :1 6 14 :3 3 14 :5 0 15 :0 7 15 :2 4 15 :4 1 15 :5 8 19 :5 1 20 :0 8 20 :2 5 20 :4 2 20 :5 9 21 :1 6 21 :3 3 21 :5 0 22 :0 7 22 :2 4 22 :4 1 22 :5 8 23 :1 5 23 :3 2 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 111. MD-5—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom03 minus NoBarCom05. lower microphones across from the barrier). Evidence of the elevated background level is present at BarCom03 dur- ing the daytime hours (left side of graph). Although the Leq (5 min.) average about 1 dB higher than at NoBarCom05, the L90 (5 min.) at BarCom03 range from 2 dB lower to 8 dB higher than at NoBarCom05, averaging approximately 3 dB higher. Almost all the daytime L90 (5 min.) are higher at BarCom03 than at NoBarCom05, which is evidence of an increase in the background sound level due to reflected sound off the barrier. The daytime L99 (5 min.) differences are more variable, but still, on average, the L99 are higher at BarCom03 than at NoBarCom05. In the evening (right side of graph), the clear trend is for the L90 (5 min.) and L99 (5 min.) at NoBarCom05 to grow louder relative to BarCom03 as time progressed. This trend is caused by the increased level and constancy of frog and insect noise at the No-Barrier site. Figure 112 presents the differences between BarCom04 and NoBarCom06 (the upper microphones across from the MD-5 barrier). During the daytime, mixed evidence exists

112 of the elevated background level at BarCom04 compared to NoBarCom06. For the first part of the afternoon measure- ments, the NoBarCom06 background level appears to be a bit higher than the level at BarCom04. For the second part of the afternoon measurements, the BarCom04 background level is higher. In the evening, the elevated background level at NoBarCom06 is again due to frog and insect noise in the No-Barrier area. For the sound-absorbing I-70 barrier, Figure 113 presents the reference microphone differences. The results show that while the A-weighted Leq (5 min.) data average about 0.5 dB higher at BarRef01 than at NoBarRef02, the L90 at BarRef01 range from 0 dB to 2 dB higher than at NoBarRef02 until approximately 17:00. The level at BarRef01 then fluctuates from 1 dB lower to 2 dB higher than that at NoBarRef02 through the end of data collection at approximately 18:30. -20 -15 -10 -5 0 5 10 12 :0 0 12 :1 7 12 :3 4 12 :5 1 13 :0 8 13 :2 5 13 :4 2 13 :5 9 14 :1 6 14 :3 3 14 :5 0 15 :0 7 15 :2 4 15 :4 1 15 :5 8 19 :5 1 20 :0 8 20 :2 5 20 :4 2 20 :5 9 21 :1 6 21 :3 3 21 :5 0 22 :0 7 22 :2 4 22 :4 1 22 :5 8 23 :1 5 23 :3 2 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 112. MD-5—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom04 minus NoBarCom06. -4 -3 -2 -1 0 1 2 3 4 5 6 14 :3 5 14 :4 4 14 :5 3 15 :0 2 15 :1 1 15 :2 0 15 :2 9 15 :3 8 15 :4 7 15 :5 6 16 :0 5 16 :1 4 16 :2 3 16 :3 2 16 :4 1 16 :5 0 16 :5 9 17 :0 8 17 :1 7 17 :2 6 17 :3 5 17 :4 4 17 :5 3 18 :0 2 18 :1 1 18 :2 0 18 :2 9 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 113. I-70—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarRef01 minus NoBarRef02.

113 -6 -4 -2 0 2 4 6 8 14 :3 5 14 :4 5 14 :5 5 15 :0 5 15 :1 5 15 :2 5 15 :3 5 15 :4 5 15 :5 5 16 :0 5 16 :1 5 16 :2 5 16 :3 5 16 :4 5 16 :5 5 17 :0 5 17 :1 5 17 :2 5 17 :3 5 17 :4 5 17 :5 5 18 :0 5 18 :1 5 18 :2 5 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 114. I-70—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom03 minus NoBarCom05. Less of a trend exists for the L99, with the BarRef01 value ranging above and below that for NoBarRef02 over the entire period. Thus, even though the first 2 hours of the L90 data suggest an increase in the background sound levels in front of the barrier, the last 2 hours of L90 data and the L99 data do not support that conclusion. Figure 114 presents the differences between BarCom03 and NoBarCom05 (the lower microphones) across from the I-70 barrier. As with the reference microphones, on average, the Leq (5 min.) results are on the order of 0 dB to 1 dB higher at the barrier site. Likewise, during the first 2 hours of data collection, the L90 results are slightly higher at BarCom03 compared to NoBarCom05—but not higher than the Leq (5 min.)—and the differences fluctuate positively and negatively during the last 2 hours. In addition, there is no indication that the L99 data are higher at BarCom03, with a range in differences from –5 dB to nearly +6 dB. Figure 115 presents the differences between BarCom04 and NoBarCom06 (the upper microphones) across from the I-70 barrier. The results are like the other microphone -6 -4 -2 0 2 4 6 8 15 :2 0 15 :3 0 15 :4 0 15 :5 0 16 :0 0 16 :1 0 16 :2 0 16 :3 0 16 :4 0 16 :5 0 17 :0 0 17 :1 0 17 :2 0 17 :3 0 17 :4 0 17 :5 0 18 :0 0 18 :1 0 18 :2 0 18 :3 0 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 115. I-70—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom04 minus NoBarCom06.

114 pairings: slightly higher Leq (5 min.); slightly higher L90 during the first half of data collection (but not higher than the Leq), and no evidence of L99 being consistently higher at BarCom04. Comparison Finding 7: Broadband L90 and L99 Statistical Descriptors at Different Distances Across from the I-75 Sound-Absorbing Barrier Broadband L90 and L99 statistical descriptors show that—unlike at the sound-reflecting barriers—background levels at both microphone distances across from the I-75 sound-absorbing barrier are not elevated above Leq. The I-90 (sound-reflecting barrier), SR-71 (sound-reflecting barrier), and I-75 (sound-absorbing barrier) locations are dis- cussed in this section. Compelling evidence exists of elevated background levels for the I-90 and SR-71 community micro- phones. These differences could be attributed to the presence of reflected sound from passing vehicles reaching the microphone in addition to the direct sound, producing a sustained sound that keeps the background level from dropping off during gaps between vehicles. Convincing evidence exists of elevated back- ground sound levels and elevated Leq at the more distant SR-71 community microphone. For the I-75 sound-absorbing bar- rier, mixed evidence exists of increased background levels at the community microphones opposite the barrier. Figure 116 presents the differences in L90 (5 min.) and L99 (5 min.), along with Leq (5 min.) for the A-weighted sound levels at the microphone pair closer to the I-90 roadway (BarCom03 and NoBarCom05). The Leq (5 min.) average about 0.5 dB to 1 dB higher at BarCom3 than at NoBarCom05, whereas the L90 at BarCom03 range from 1 dB to 2 dB higher, and the L99 at BarCom03 range from 1 dB to 4 dB higher. Figure 117 presents differences for the more distant micro- phone pair at I-90 (BarCom04 and NoBarCom06). Strong evidence exists of elevated background levels at BarCom04 compared to NoBarCom06, with the differences similar to those described in the BarCom03 comparison to NoBarCom05. To review, for the sound-reflecting barrier at SR-71, Fig- ure 49 (in Chapter 5) showed that whereas the Leq (5 min.) average about 0 dB to 1 dB higher at BarRef01 than at NoBarRef02, the L90 (5 min.) and L99 (5 min.) at BarRef01 are much higher than at NoBarRef02—L90 by as much as 4 dB and L99 by as much as 7 dB. Figure 118 presents the differences at the community microphones close to the road on the opposite side of the SR-71 barrier, computed as BarCom03 minus NoBarCom05. Strong evidence exists of the elevated background level at BarCom03 compared to NoBarCom05 even though these two microphones are close to the edge of the shoulder across from the barrier. The Leq (5 min.) at BarCom03 range between about 0.5 dB higher to 1 dB lower than the Leq (5 min.) at NoBarCom05; how- ever, the L90 at BarCom03 average about 1 dB higher than at NoBarCom05, and the L99 at BarCom03 average approxi- mately 1.5 dB higher. Even though there are also times when the NoBarCom05 levels exceed the BarCom03 levels, on average, the trend is for the L90 and L99 to be higher at the Barrier microphone. As observed in the field by the researchers when listening to the sound across from the barrier, there was a change in the character of the sound at -4 -3 -2 -1 0 1 2 3 4 5 6 13 :0 0 13 :0 9 13 :1 8 13 :2 7 13 :3 6 13 :4 5 13 :5 4 14 :0 3 14 :1 2 14 :2 1 14 :3 0 14 :3 9 14 :4 8 14 :5 7 15 :0 6 15 :1 5 15 :2 4 15 :3 3 15 :4 2 15 :5 1 16 :0 0 16 :0 9 16 :1 8 16 :2 7 16 :3 6 16 :4 5 16 :5 4 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 116. I-90—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom03 minus NoBarCom05.

115 -4 -3 -2 -1 0 1 2 3 4 5 6 13 :0 0 13 :0 9 13 :1 8 13 :2 7 13 :3 6 13 :4 5 13 :5 4 14 :0 3 14 :1 2 14 :2 1 14 :3 0 14 :3 9 14 :4 8 14 :5 7 15 :0 6 15 :1 5 15 :2 4 15 :3 3 15 :4 2 15 :5 1 16 :0 0 16 :0 9 16 :1 8 16 :2 7 16 :3 6 16 :4 5 16 :5 4 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 117. I-90—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom04 minus NoBarCom06. the Barrier site compared to the No-Barrier site. Observers sensed the physical presence of a large object with the barrier in place, an echolocation phenomenon. Figure 119 presents the differences between the distant community microphones at SR-71, computed as BarCom04 minus NoBarCom06. This graph starts 45 minutes into the measurement period because a great deal of roofing nail gun noise was audible at NoBarCom06 during the first 45 min- utes. The results for this microphone pair differ from those for most of the microphone pairs at the other study locations because these microphones were located the farthest from the roadway. For the first 23 minutes of the period shown in Figure 119, the BarCom04 Leq (5 min.) range from 2.5 dB to 3.8 dB higher than the NoBarCom06 Leq (5 min.). During this time, the meteorological class was Calm Neutral and, as seen in the figure, the L90 (5 min.) and L99 (5 min.) differ- ences range from 2 dB to 5 dB higher at BarCom04 than at NoBarCom06. During the last 3 hours of measurement, the Leq (5 min.) differences become more variable, from 0.5 dB to 2.5 dB higher at BarCom04. The L90 (5 min.) differences also become more variable, being 0 dB to 3.5 dB higher at BarCom04. The L99 (5 min.) become even more variable, with the BarCom04 values ranging from 1 dB lower than those at NoBarCom06 to 5.4 dB higher. During this later period, the meteorological class was Downwind Neutral. On average, over the full measurement period, the BarCom04 Leq (5 min.), L90 (5 min.), and L99 (5 min.) are 1.7 dB, 2.0 dB, and 2.1 dB higher than the measured levels at NoBarCom06, -4 -3 -2 -1 0 1 2 3 4 5 6 9: 00 9: 10 9: 20 9: 30 9: 40 9: 50 10 :0 0 10 :1 0 10 :2 0 10 :3 0 10 :4 0 10 :5 0 11 :0 0 11 :1 0 11 :2 0 11 :3 0 11 :4 0 11 :5 0 12 :0 0 12 :1 0 12 :2 0 12 :3 0 12 :4 0 12 :5 0 13 :0 0 13 :1 0 13 :2 0 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 118. SR-71—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom03 minus NoBarCom05.

116 respectively. These results suggest that the overall traffic noise levels are higher at the SR-71 Barrier site, not just the back- ground levels. Because the traffic is 400 ft. away, the sound levels rise and fall less, overall, compared to the levels mea- sured at microphones close to the road. As a result, under the studied traffic flows, little chance occurs for lulls in the noise. For the sound-absorbing I-75 barrier, Figure 120 presents the reference microphone differences in the A-weighted L90 (5 min.), L99 (5 min.), and Leq (5 min.), computed as BarRef01 minus NoBarRef02. These results show that, even though the Leq (5 min.) are from 0.1 dB to 1.7 dB higher at BarRef01 than at NoBarRef02, almost all the L90 at BarRef01 are higher than those at NoBarRef02, with a range from –0.5 dB to 3.1 dB. (Some large differences occur during the period from approximately 20:20 to 20:48; however, the field notes offer no explanation of unusual activity during this time that could account for these exceptions.) The L99 differ- ences are more variable: less than one-quarter of the L99 are lower at BarRef01 than at NoBarRef02; the rest are higher at BarRef01 than at NoBarRef02, again with unexplained large -4 -3 -2 -1 0 1 2 3 4 5 6 9: 45 9: 55 10 :0 5 10 :1 5 10 :2 5 10 :3 5 10 :4 5 10 :5 5 11 :0 5 11 :1 5 11 :2 5 11 :3 5 11 :4 5 11 :5 5 12 :0 5 12 :1 5 12 :2 5 12 :3 5 12 :4 5 12 :5 5 13 :0 5 13 :1 5 13 :2 5 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 119. SR-71—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom04 minus NoBarCom06. -2 -1 0 1 2 3 4 5 6 7 8 17 :1 2 17 :2 0 17 :2 8 17 :3 6 17 :4 4 17 :5 2 18 :0 0 18 :0 8 18 :1 6 18 :2 4 18 :3 2 18 :4 0 18 :4 8 18 :5 6 19 :0 4 19 :1 2 19 :2 0 19 :2 8 19 :3 6 19 :4 4 19 :5 2 20 :0 0 20 :0 8 20 :1 6 20 :2 4 20 :3 2 20 :4 0 20 :4 8 20 :5 6 21 :0 4 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 120. I-75—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarRef01 minus NoBarRef02.

117 differences occurring from around 20:20 to 20:48. The gener- ally higher L90 at BarRef01 compared to NoBarRef02 suggest an increase in the background level in front of this sound- absorbing barrier. Such an increase could be due to the pres- ence of reflected sound reaching the microphone in addition to the direct sound from the passing vehicles; however, the times when the L99 are lower at BarRef01 than at NoBarRef02 do not support this theory. Figure 121 shows the differences at I-75 for the closer micro- phones across from the barrier (BarCom03 and NoBarCom05). The Leq (5 min.) generally are 0 dB to 1 dB higher at BarCom03 than at NoBarCom05; however, the L90 and L99 differences are much more variable. The A-weighted L90 ranges from 2 dB lower to 2.5 dB higher at BarCom03 compared to NoBarCom05, while the A-weighted L99 fall in a ±4 dB range from 0.5 dB to 1.5 dB higher at BarRef01 than at NoBarRef02. The results do not suggest a background level elevated beyond the increase seen in the Leq. Figure 122 presents the differences at I-75 for the more distant microphones across from the barrier (BarCom04 and NoBarCom06). The results for this microphone pair are like those for BarCom03 and NoBarCom05, showing little evidence of a consistent elevation of the background level at BarCom04 compared to NoBarCom06. -5 -4 -3 -2 -1 0 1 2 3 4 5 17 :1 2 17 :2 2 17 :3 2 17 :4 2 17 :5 2 18 :0 2 18 :1 2 18 :2 2 18 :3 2 18 :4 2 18 :5 2 19 :0 2 19 :1 2 19 :2 2 19 :3 2 19 :4 2 19 :5 2 20 :0 2 20 :1 2 20 :2 2 20 :3 2 20 :4 2 20 :5 2 21 :0 2 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 121. I-75—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom03 minus NoBarCom05. -5 -4 -3 -2 -1 0 1 2 3 4 5 17 :1 2 17 :2 2 17 :3 2 17 :4 2 17 :5 2 18 :0 2 18 :1 2 18 :2 2 18 :3 2 18 :4 2 18 :5 2 19 :0 2 19 :1 2 19 :2 2 19 :3 2 19 :4 2 19 :5 2 20 :0 2 20 :1 2 20 :2 2 20 :3 2 20 :4 2 20 :5 2 21 :0 2 So un d Le ve l D iff er en ce , d B Time L90 L99 Leq Figure 122. I-75—differences in A-weighted levels (dB), 5-minute L90, L99, and Leq, BarCom04 minus NoBarCom06.

118 Comparison Finding 8: One-Third Octave Band L90 and L99 Descriptors The one-third octave band L90 and L99 descriptors mostly support the conclusion that—unlike at the sound-reflecting barriers—the background noise is not elevated because of reflections opposite the sound-absorbing barriers. Comparison Finding 7 was based on analysis of the broad- band A-weighted sound levels and unweighted sound pres- sure levels. The analysis that led to Comparison Finding 8 examined the data in terms of one-third octave bands by use of color shading. To review, in the one-third-octave band spectrograms, the brown color indicates time periods during which the Barrier levels are higher than the No-Barrier levels and the blue color indicates that No-Barrier levels are higher. As in Chapters 5 and 6, each figure presents time running from top to bottom, with the total block representing approx- imately 4 hours. Each row represents the starting minute of a running 5-minute period. The one-third octave bands run in columns from left to right, with the first two columns on the left showing the differences for the broadband A-weighted sound levels and unweighted sound pressure levels. Each band’s data column contains the differences for seven sound pressure level Ln values (left to right L1, L5, L10, L33, L50, L90, L99), plus Leq. This section presents the results for the sound- reflecting barriers at I-24, I-90, SR-71, and MD-5, and the sound-absorbing barriers at I-75 and I-70. To review, Figure 50 (in Chapter 5) compared the refer- ence microphone levels for the I-24 sound-reflecting barrier. Figure 50 showed higher L90 (5 min.) and L99 (5 min.) values at BarRef01 than at NoBarRef02 in the frequency bands from 400 Hz through 2 kHz for most of the sample period and through 5 kHz for the first half of the sample period. These higher background levels are evidence of the sustaining of a vehicle’s pass-by noise due to the creation of an image source for each vehicle as the sound reflects off the barrier. Figure 123 presents the Ln differences for BarCom03 and NoBarCom05 at I-24, and Figure 124 presents the Figure 123. I-24—differences in Ln (5 min.) by one-third octave frequency bands, BarCom03 and NoBarCom05. Figure 124. I-24—differences in Ln (5 min.) by one-third octave frequency bands, BarCom04 and NoBarCom06.

119 Ln differences for BarCom04 and NoBarCom06 at I-24. Some evidence exists of slightly higher Ln values at BarCom03 versus NoBarCom05 in the 315 Hz to 800 Hz bands and at BarCom04 versus NoBarCom06 over much of the lower fre- quency range. Little evidence is seen of elevated background levels (which would be indicated by brown coloring on the right sides of the frequency band columns). The strong blue coloring in the 6.3 kHz and 8 kHz bands indicates elevated background levels at the No-Barrier microphones due to insect noise in the vegetation behind these microphones. For the I-90 sound-reflecting barrier, Figure 125 presents the Ln differences for BarCom03 and NoBarCom05. The brown color across the 250 Hz to 500 Hz columns indicates an increase in all the Ln descriptors, meaning the BarCom03 levels are higher than the NoBarCom05 levels. Vertical brown coloring on the right sides of the data columns in the frequency bands from 630 Hz to 3.15 kHz mean that the BarCom03 background levels are higher than the NoBarCom05 background levels, suggesting the reflected sound is sustaining the background level opposite the barrier. Figure 126 presents the Ln differences for BarCom04 and NoBarCom06 at I-90. This chart contains much less brown shading, showing that the BarCom04 upper microphone lev- els are not that much higher than the NoBarCom06 levels; however, some brown shading remains on the right sides of the mid-frequency columns, suggesting some elevation of the background at BarCom04. The blue color in the 20 Hz to 31.5 Hz bands and the bands at and above the 4 kHz band show the NoBarCom06 levels to be higher than at BarCom04 at these frequencies. For SR-71, Figure 52 (in Chapter 5) compared the BarRef01 and NoBarRef02 levels. The brown coloring on the right sides of the data columns showed higher L90(5 min.) and L99(5 min.) at BarRef01 in the frequency bands from 500 Hz through 4 kHz for nearly the entire sample period, and through 5 kHz for the first half of the sample period. Figure 125. I-90—differences in Ln (5 min.) by one-third octave frequency bands, BarCom03 and NoBarCom05. Figure 126. I-90—differences in Ln (5 min.) by one-third octave frequency bands, BarCom04 and NoBarCom06.

120 These higher values are evidence of a sustaining of a vehicle’s pass-by noise due to reflected sound, similar to the sound- reflecting I-24 Barrier location. Figure 127 presents spectral Ln differences for BarCom03 and NoBarCom05 at SR-71. These microphones were located just off the shoulder of the roadway. In the 1 kHz to 4 kHz bands, there appears to be a general trend for the BarCom03 L90 and L99 background levels to be higher than the NoBarCom05 levels, showing evidence of an elevated background in these bands even at the short distance from the nearby traffic. Figure 77 (in Chapter 5) presented the spectral Ln differ- ences for the distant BarCom04 and NoBarCom06 positions at SR-71. In contrast to the closer microphones at SR-71, differences occur for most of the Ln descriptors, not just for the L90 (5 min.) and L99 (5 min.) indicators of elevated back- ground. A pattern exists of higher broadband A-weighted levels at BarCom04 in the lower bands and the mid- to upper bands, with higher NoBarCom06 levels in the 100 Hz to 250 Hz bands and the highest frequency bands. For the MD-5 sound-reflecting barrier, Figure 128 presents the Ln differences for the lower microphones (BarCom03 and NoBarCom05). At MD-5, there were two measurement sessions (one in the afternoon and one at night), so two blocks of data appear the figure. The brown bands, which are centered on 250 Hz through 400 Hz, show increases in the BarCom03 levels relative to those for NoBarCom05 across most of the Ln descriptors in these bands, which is interpreted as evidence of overall increases in sound pres- sure levels due to reflections off the barrier. The daytime data also show higher levels at BarCom03 for L90 and L99 for the bands from 500 Hz through 3,150 Hz. The heavy blue shading in the night period’s high-frequency bands repre- sent the frog and insect noise in the No-Barrier area. Figure 129 presents the MD-5 Ln differences for the upper microphones (BarCom04 and NoBarCom06). The brown areas show small increases in the BarCom04 levels relative to those for NoBarCom06 across most of the Ln descriptors in the bands centered on 100 Hz through 160 Hz. The daytime data also show mostly higher L90 and L99 values at BarCom04 for the bands from 630 Hz through 2.5 kHz. The blue bands in the night’s high-frequency bands again represent the frog and insect noise in the No-Barrier area. The latter part of the evening sampling for L1 and Leq includes two horizontal lines of blue shading for L1 and Leq, which indicate two short- term loud events at NoBarCom06 that did not occur at the BarCom04 site. For the sound-absorbing I-75 barrier, Figure 130 compares the BarRef01 and NoBarRef02 Ln descriptors. The overall brownish tone in the figure indicates that the BarRef01 levels are slightly higher than the NoBarRef02 levels across most of the frequencies and all Ln descriptors. Above 1 kHz, the BarRef01 L99 (5 min.) are higher than the NoBarRef02 values, although this trend is not consistent over all the periods. In several higher frequency bands, occasional blue blocks for the L99 (5 min.) data occur at 3.15 kHz and above. An elevated L99 (5 min.) or L90 (5 min.) at BarRef01 would be an indication that background levels at this microphone were higher than the background levels at NoBarRef02. This difference was interpreted as evidence of a sustaining of a vehicle’s pass-by noise due to the creation of an image source for each vehicle as the sound reflects off the barrier. Figure 131 presents the Ln differences for BarCom03 and NoBarCom05 at I-75. The relatively neutral shading in the fig- ure, covering most of the descriptors in most of the frequency bands, indicates little to no difference in levels at the two micro- phones. The exception is the L99 (5 min.) data at 4 kHz and above. At these frequencies, the L99 (5 min.) data show a stronger blue tone, indicating that the background level at NoBarCom05 Figure 127. SR-71—differences in Ln (5 min.) by one-third octave frequency bands, BarCom03 and NoBarCom05.

121 Figure 128. MD-5—differences in Ln (5 min.) by one-third octave frequency bands, BarCom03 and NoBarCom05 in the afternoon (top) and evening (bottom). was generally higher than at BarCom03 at these frequency bands. The results show no evidence of elevated background across from the barrier, which indicates no sustaining of the possibly sound-reflected noise off the sound-absorbing barrier. Figure 132 presents the Ln differences for BarCom04 and NoBarCom06 at I-75. Below 2 kHz, there is no strong color indication of higher levels at BarCom04. In the 2.5 kHz to 5.0 kHz bands, the increased evidence of blue coloring is an indication of somewhat higher levels at NoBarCom06 than at BarCom04 across most of the Ln descriptors, with greater differences apparent for the L99 descriptor. The field team did not notice any localized noise sources that could explain the higher background levels indicated by the higher L99 values at NoBarCom06. Little evidence exists of increased background levels at BarCom04 across from the barrier. For the sound-absorbing barrier at I-70, Figure 133 compares the BarRef01 and NoBarRef02 levels. The overall brownish tone shown for the frequency bands from 20 Hz to 50 Hz and from 4 kHz to 10 kHz indicates that the BarRef01 levels are slightly higher than the NoBarRef02 levels across all Ln descriptors in those bands. In the higher frequency bands, the L99 (5 min.) columns show no clear trend, with some periods of higher background levels at BarRef01 and some periods of higher background levels at NoBarRef02. Across the rest of the spectrum from 80 Hz through 3.15 kHz, the figure indicates little difference in levels for most of the Ln descriptors at the two microphone positions, with some indi- cation of higher L99 (5 min.) at NoBarRef02 at 160 Hz, 200 Hz, and 3.15 kHz, and some indication of higher L99 (5 min.) at BarRef01 from 250 Hz through 630 Hz. Figure 134 presents the Ln differences for the lower micro- phones (BarCom03 and NoBarCom05) at I-70. In the 315 Hz and 400 Hz bands, all the Ln descriptors at BarCom03 are clearly higher than those at NoBarCom05. In contrast, the blue coloring indicates that the L90 (5 min.) and L99 (5 min.) at NoBarCom05 generally are higher than those at BarCom03 in the 160 Hz and 200 Hz bands, and at 250 Hz for the first half of the sample period. In many of the bands at and above 630 Hz, the L99 (5 min.) are slightly higher at NoBarCom05 than at BarCom03. Figure 135 presents the Ln differences for BarCom04 and NoBarCom06 at I-70. In the 200 Hz and 250 Hz frequency

122 Figure 129. MD-5—differences in Ln (5 min.) by one-third octave frequency bands, BarCom04 and NoBarCom06 in the afternoon (top) and evening (bottom). Figure 130. I-75—differences in Ln (5 min.) by one-third octave frequency bands, BarRef01 and NoBarRef02.

123 Figure 131. I-75—differences in Ln (5 min.) by one-third octave frequency bands, BarCom03 and NoBarCom05. Figure 132. I-75—differences in Ln (5 min.) by one-third octave frequency bands, BarCom04 and NoBarCom06. Figure 133. I-70—differences in Ln (5 min.) by one-third octave frequency bands, BarRef01 and NoBarRef02.

124 bands, the BarCom04 levels are higher than those at NoBar- Com06 for most of the descriptors. From 315 Hz through 2.5 kHz, the Ln descriptors are quite close to each other. At 8 kHz and 10 kHz, the levels for all the descriptors are substan- tially higher at BarCom04 than at NoBarCom06. The source of this high-frequency noise was not identified; however, based on the frequency, it was not related to traffic sources. Comparison Finding 9: Effects of Site Differences Finding suitable equivalent No-Barrier sites for studying both sound-absorbing and sound-reflecting barriers is difficult because of many factors, including reflections off other objects (such as houses or other buildings), differences in terrain, and localized noise sources, which affect results. A great deal of time went into site selection, including review of noise barrier data bases, discussions with state highway agency noise specialists, use of web-based mapping tools, preliminary site evaluations, and detailed advanced field reviews. Despite these efforts, field conditions are unpredictable, and it can be difficult to account for all factors that might make it difficult to collect the data needed to discern the generally small sound level differences between Barrier and equivalent No-Barrier sites. As an example, the SR-155 (Briley Parkway, in Nashville, Tennessee) sound-reflecting barrier location was ideal, from a traffic point of view, for conducting nighttime measure- ments to isolate individual vehicles. The road was elevated above the community behind a retaining wall, which required tall microphone masts to get above the roadway elevation. Differing layouts among the neighborhood houses limited the No-Barrier site possibilities. The only satisfactory No-Barrier site had trees, and insect and frog noise became a real problem after the sun set and as the night progressed. Also, a somewhat distant bridge joint generated enough noise to contaminate the No-Barrier levels as the evening progressed, traffic became light, and a temperature inversion enhanced the propagation of the bridge joint noise. Additionally, the Barrier site had a bit of roadway superelevation away from the community micro- phone side of the road, which led to some added shielding Figure 134. I-70—differences in Ln (5 min.) by one-third octave frequency bands, BarCom03 and NoBarCom05. Figure 135. I-70—differences in Ln (5 min.) by one-third octave frequency bands, BarCom04 and NoBarCom06.

125 -4 -3 -2 -1 0 1 2 3 0: 00 0: 10 0: 20 0: 30 0: 40 0: 50 1: 00 1: 10 1: 20 1: 30 1: 40 1: 50 2: 00 2: 10 2: 20 2: 30 2: 40 2: 50 3: 00 3: 10 3: 20 3: 30 3: 40 3: 50 D iff er en ce in le ve l, dB Time dBA dBZ Figure 136. I-270—differences in A-weighted and unweighted levels (dB), running Leq (5 min.), BarCom03 minus NoBarCom05. -4 -3 -2 -1 0 1 2 3 0: 00 0: 10 0: 20 0: 30 0: 40 0: 50 1: 00 1: 10 1: 20 1: 30 1: 40 1: 50 2: 00 2: 10 2: 20 2: 30 2: 40 2: 50 3: 00 3: 10 3: 20 3: 30 3: 40 3: 50 D iff er en ce in le ve l, dB Time dBA dBZ Figure 137. I-270—differences in A-weighted and unweighted levels (dB), running Leq, BarCom04 minus NoBarCom06. from the retaining wall parapet that caused the Barrier site Leq (5 min.) to be lower than the No-Barrier levels, apart from issues with insect, frog, and bridge joint noise. The I-270 location in Grove City, Ohio, provides another example of the difficulties in site selection when small dif- ferences in sound levels are being investigated. In this case, the issue appeared to be noise reflections off a house in the No-Barrier area. Illustrations from the I-270 data are pre- sented in Figures 136 and 137. I-270 is a six-lane freeway. The barrier is approximately 10 ft. high and is offset 12–18 ft. from the edge of the near- est travel lane, with an NRC of 0.85. NoBarCom06 was posi- tioned at the end of a row of one-story houses parallel to I-270, and the houses were approximately 40 ft. behind the microphone. The data from this location suggest that reflected noise off the houses in this row raised the NoBarCom06 levels, although no note of reflected noise was made during the field review as part of the location selection process or during the measurement periods. Mea- surements were conducted from approximately 00:00 (mid- night) to 04:00 to obtain data on individual vehicle pass-bys in addition to the 5-minute averages. Figure 136 shows the differences in unweighted and A-weighted levels for BarCom03 and NoBarCom05, which were located just off the shoulder of I-270. Figure 137 shows the differences in unweighted and A-weighted levels for BarCom04

126 and NoBarCom06. For most of the measurement periods, both the unweighted and A-weighted 5 minute Leq are lower at BarCom03 than at NoBarCom05, by ranges of approximately 0 dB to 1.5 dB and 0 dB to 1 dB, respectively. For some of the periods, the BarCom03 Leq (5-min.) are higher than the NoBarCom05 values by up to approximately 1 dB. For nearly all the periods, the unweighted and A-weighted BarCom04 Leq (5 min.) are lower than those at NoBarCom06, by a range of approximately 0 dB to 3 dB. Interestingly, the A-weighted differences look more pronounced in the first 15 minutes of sampling, ranging from –1.8 dB to –3.2 dB. One possible reason for the higher levels at NoBarCom06 is that traffic noise may have been reflecting off the single-family homes behind the NoBarCom06 microphone and increasing the level. Simplified image source modeling with the FHWA TNM 2.5 showed a 0.8 dB increase in the A-weighted sound level at NoBarCom06 due to reflections off the building. It seems unlikely, however, that reflections could cause the higher levels measured at NoBarCom05 compared to BarCom03. This is because the latter two microphones were close to the I-270 traffic, and NoBarCom05 was much farther from the houses than NoBarCom06. The terrain between the houses and NoBarCom05 might also provide some shielding of any reflected noise. The simplified TNM modeling showed no increase in the NoBarCom05 level due to building reflections. All the groupings of 5-minute periods that were judged equivalent for traffic parameters at the I-270 location fell into the single meteorological class of Calm Inversion. Figure 138 shows the differences in Leq (5 min.) between BarCom03 and NoBarCom05 (top) and BarCom04 and NoBarCom06 (bottom) for an average of all the Calm Inversion measurements. BarCom03 and NoBarCom05 were both located close to the road and positioned 5 ft. above the roadway elevation. The BarCom03 levels are lower than or equal to the NoBar- Com05 levels over the entire frequency range, except for a slightly higher level at 2 kHz. From 80 Hz through 2.5 kHz the difference is 1 dB or less except at two points: 315 Hz and 400 Hz, where the difference is approximately 1.5 dB, with BarCom03 being lower. The reason for NoBarCom05 being higher is not apparent because, as noted, simplified FHWA TNM 2.5 image source modeling showed no increase in the A-weighted Leq due to reflections off the houses in the No-Barrier area. -5 -4 -3 -2 -1 0 1 2 3 4 5 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1 k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CIG Groups -5 -4 -3 -2 -1 0 1 2 3 4 5 dB A d BZ 2 0 2 5 3 1. 5 4 0 5 0 6 3 8 0 1 00 1 25 1 60 2 00 2 50 3 15 4 00 5 00 6 30 8 00 1 k 1. 25 k 1. 6k 2k 2. 5k 3. 15 k 4k 5k 6. 3k 8k 10 k D iff er en ce in L ev el , d B 1/3 Octave Band Frequency, Hz All CIG Groups Figure 138. I-270—averages of the differences in levels (dB) for all Calm Inversion groups, Leq (5 min.) é one standard deviation, all microphones.

127 The lower graph in Figure 138 compares the levels at BarCom04 and NoBarCom06, both of which were positioned 15 ft. above the roadway surface. The BarCom04 levels are lower than the NoBarCom06 levels from 20 Hz to 100 Hz by a range of 0 dB to 3.5 dB; they also are lower from 500 Hz to 5 kHz by a range of 0.2 dB to 2.3 dB. From 125 Hz to 400 Hz, however, the BarCom04 levels are higher than the levels at NoBarCom06, by 0.5 dB at 125 Hz and 400 Hz to approxi- mately 3 dB from 160 Hz to 315 Hz. The pattern suggests that, at these frequencies, noise reflected off the sound-absorbing barrier is raising this lower frequency level at BarCom04. Above 5 kHz, the BarCom04 levels are higher than the NoBarCom06 levels. The overall unweighted sound pressure level is 1.5 dB higher at NoBarCom06 than at BarCom04 and the overall A-weighted sound level is 1.6 dB higher. The higher levels at NoBarCom06 in these lower frequency and certain higher-frequency bands could be attributable to reflection of noise off the houses, and the lower levels in the middle fre- quencies at NoBarCom06 compared to BarCom04 could be attributable to ground effects, despite the elevated position of the microphone. The simplified FHWA TNM 2.5 modeling showed a 0.8 dB increase in the Leq due to reflections off the closest houses. More details on the I-270 data are provided in Appendix B. The background sound pressure level analysis showed that even though the A-weighted Leq (5 min.) data at BarCom03 typi- cally ranged between 0 dB to 1 dB lower than the same data at NoBarCom05, substantial variation occurred in the L90 and L99 data, with many periods during which the BarCom03 levels were higher than the NoBarCom05 values. Likewise, for the more distant BarCom04 and NoBarCom06 microphones, there was a great deal of variation in the L90 and L99 data, with many periods during which the BarCom04 levels were higher than those at NoBarCom06, even though the A-weighted Leq (5 min.) data at BarCom04 typically ranged between 1 dB to 2.5 dB lower than the NoBarCom06 values. These L90 and L99 results at I-270 at least support the finding from the I-75 and I-70 loca- tions that the background level opposite a sound-absorbing bar- rier is not increased by sustained noise during vehicle pass-bys. Comparison Finding 10: Spectrograms Indicate that Sound-Absorbing Barrier Effects Are Subtle Compared to spectrograms for reflective barriers, where reflection effects are readily apparent, spectrograms for absorptive barriers reveal little indication of reflection effects. Spectrograms show the frequency content of sound as a function of time. This section qualitatively compares Phase 1 and Phase 2 results of the spectrogram analysis. As was shown in Chapter 5, for reflective barriers, reflections cause sound levels to increase over a broad range of frequencies and cause higher sound levels to be sustained for a longer period of time. This finding is evidenced by the spectrogram results at all loca- tions. The evidence is shown in the spectrogram as hot spots intensifying and being both wider and taller with a barrier present. The same effect occurs in the surrounding frequency bands, stepping through various colors of the spectrum. These results were extracted both from data on vehicle pass-by events and from blocks of data specified by time period (e.g., 5-minute data blocks). This section summarizes the results for sound-absorptive barriers as compared to the results for sound-reflective barriers. Examples of the vehicle pass-by events are shown first, followed by example 5-minute data blocks of highway traffic noise. When comparing the spectrogram data for the Barrier sites to the No-Barrier sites, the following observations can be made: • There appear to be some subtle differences between the sound-absorbing Barrier site spectrograms and the No- Barrier site spectrograms. These differences include a slight broadening of hot spots and possible intensification of the hot spots; however, the differences are not readily apparent and do not appear in all cases. • To reveal subtle differences in the sound-absorbing Barrier site spectrograms and to compare these to reflective bar- rier Barrier site spectrograms, the research team needed to develop a difference spectrogram method. This method is discussed in Comparison Finding 11. For the I-75 and I-70 locations, sound levels opposite the sound-absorbing noise barrier appear in some cases to be slightly higher than the sound levels at No-Barrier locations at key frequencies that contribute most to the overall noise level. Results for the I-270 location are not presented in this section because of the reflection of sound from homes on the community side of the highway. Figure 139 shows a vehicle pass-by spectrogram example from the I-75 sound-absorbing barrier location. This example presents results for the community-side far-microphone positions (BarCom04, labeled “Mic 4” in the figure and NoBarCom06, labeled “Mic 6” in the figure). These microphones were located 100 ft. from the road. The vehicle (a truck) was trav- eling northbound (on the barrier side of the road), arriving first at Mic 6 and then reaching Mic 4 about 20 seconds later. For the community-side far-microphone positions, it is difficult to see any differences. The darkest red area of the spectrogram appears to have a more distinguishable shape for the Barrier site. For the community-side near-microphone positions (detailed in Appendix B), results are similar, with the addi- tional observation that the light blue appears to expand to lower frequencies for the near-microphone position with no barrier. Figure 140 shows a second vehicle pass-by spectrogram example. This example is from the I-70 sound-absorbing barrier location. Results are shown for high-microphone

128 Figure 139. I-75—spectrograms at sound-absorbing barrier during heavy truck pass-by event at }18:33, BarCom04 (top), NoBarCom06 (bottom).

129 Figure 140. I-70—spectrograms at sound-absorbing barrier during heavy truck pass-by event at }17:45, BarCom04 (top), NoBarCom06 (bottom).

130 community positions (again labeled “Mic 4” and “Mic 6”). These microphones were located 75 ft. from the road. Fig- ure 140 shows a single heavy truck pass-by event on I-70 at approximately 17:45. The truck was traveling eastbound (on the opposite side of the roadway from the barrier), arriv- ing first at the No-Barrier site (Mic 6) and then reaching the Barrier site (Mic 4) about 32 seconds later. Together, Figure 139 and Figure 140 show the results from community-side microphones in both low and high positions. The low position is for the 100-ft. microphone at Site 1-75, and the high position is for the 75-ft. microphone at Site I-70. Both pairs of spectrograms show that the hot spots are slightly broader in frequency and may last slightly longer with the barrier present. In addition, the hottest spots (darkest red color) appear to be a darker red color with the barrier present, from approximately 500 to 2,000 Hz. The next two examples show longer periods of traffic noise. Figure 141 presents spectrograms for the I-75 sound- absorbing barrier location. These spectrograms represent a 5-minute data block from 17:55 to 18:00 for the community- side microphones (again labeled “Mic 4” and “Mic 6”) posi- tioned 100 ft. from the edge of the nearest travel lane. As in Figures 139 and 140, the dark red hotspots (here seen in the range of about 500 Hz to 1,600 Hz) appear to be more filled in and have a more distinct shape with the barrier present. For the community-side near-position microphones (detailed in Appendix B), it is difficult to see any differences in the shapes or intensity of the hottest spots (dark red); however, the light blue appears to expand to lower frequencies for the near- position microphone with no barrier. Figure 142 shows spectrograms for a 5-minute data block at the sound-absorbing barrier location at I-70. These high- position microphones were located 75 ft. from the road. At this location, for both the community-side positions and reference positions, it is difficult to see any differences in the shapes and shades of the hottest spots (dark red) in the spectrograms for the high microphones (Figure 142) and the low microphones (see Appendix B). Comparison Finding 11: Difference Spectrograms Indicate that Barrier Reflections Result in Comb Filtering, Which Changes the Sound Quality Difference spectrograms revealed harmonically related peaks caused by comb filtering, which can be attributed to the barrier and can be perceived as the sound being buzzy or raspy. Absorptive barriers may reduce the comb-filtering effect. This part of the comparison of the Phase 1 and Phase 2 results focuses on spectrogram differences. Key elements of the comparison are: • The research team developed a method to visualize spec- trogram differences for the same vehicle passing by a Barrier site and an equivalent No-Barrier site. • Spectrogram difference analysis revealed harmonically related peaks caused by comb filtering (in which direct and barrier-reflected sound waves combine, with constructive and destructive interference, resulting in comb filtering). • Comb-filtered sound can be perceived by human listeners as having a buzzy or raspy quality. • The results of this study suggest that absorptive barriers may reduce the comb-filtering effect, although this would need to be substantiated by conducting a narrow-band analysis. Spectrograms were generated for both sound-reflecting noise barriers and sound-absorbing noise barriers. For each barrier type, examination of isolated vehicle pass-by events helped the research team to compare spectral content between sound levels measured at the Barrier site and at the No-Barrier site. Some differences become apparent when examining the Barrier and No-Barrier spectrograms; however, subtle dif- ferences cannot always be distinguished. Because identifying subtle differences is essential to comparing sound-reflecting and sound-absorbing barriers, the research team developed a method to readily visualize subtle differences. The method yields “difference spectrograms,” which can be compared for sound-reflecting barriers and sound-absorbing barriers. The method is described in Chapter 2. Figures 143 and 144 present difference spectrograms based on results for the reflective barrier at MD-5. Figures 145 and 146 present difference spectrograms based on results for the absorptive barrier at I-75. Figures 143 and 144 include one clean vehicle pass-by event, and Figures 145 and 146 also include one clean vehicle pass-by event. The first figure for each site shows: (1) the overall A-weighted sound level time history for each event at each site (top graph); (2) the spectrograms at the Barrier and No-Barrier sites (middle two plots); and (3) the difference plot (bottom) for each event. The second figure for each site shows a slice in time that reveals the peaks in differences by one-third octave band frequency. Examination of the spectrogram difference plots reveals lines of hot spots across the spectrograms, particularly just before and after the maximum sound level of the event. The slice-in-time plots show the locations of the “hot lines,” revealing multiple peak frequencies with a special relation- ship: harmonic. For a harmonic relationship, the frequency must be a multiple of the base frequency; for example, for a base frequency of 25, harmonically related frequencies are 50, 75, 100, 125, and so forth, which are the base times 2, 3, 4, 5, and so forth, respectively. Table 20 lists the peak dif- ference frequencies for each event analyzed, with comments about the harmonic relationships. For the reflective barrier vehicle pass-by events, the peak difference frequencies are

131 Figure 141. I-75—5-minute spectrograms at sound-absorbing barrier, 17:55–18:00, BarCom04 (top), NoBarCom06 (bottom).

132 Figure 142. I-70—5-minute spectrograms at sound-absorbing barrier, 15:30–15:35, BarCom04 (top), NoBarCom06 (bottom).

133 Figure 143. MD-5—spectrogram difference plot for sound-reflecting barrier pickup truck pass-by event at }20:09, BarCom04 and NoBarCom06.

134 Figure 144. MD-5—differences in levels (dB) at sound-reflecting barrier, –0.5 seconds from maximum during pickup truck pass-by event at }20:09, BarCom04 and NoBarCom06. harmonically related with few exceptions. For the sound- absorbing barriers, the lines indicate many exceptions to a harmonic relationship to a single base frequency. (When multiple exceptions to the harmonic relationships occur, the exceptions are harmonically related to each other. The har- monic relationships among the exceptions could indicate the presence of reflection effects from another surface, such as the roadway pavement.) To help discern any trends when comparing the two barrier types, all the spectral slice-in-time plots for both the sound- reflecting and sound-absorbing barrier sites are presented in a single plot (Figure 147). The combined plot shows that reflective barrier peaks and dips are pronounced and show a strong harmonic relationship for lower frequencies (500 Hz and lower) as compared to sound-absorbing barrier peaks and dips. In addition, the sound-absorbing barrier lines are generally lower in amplitude (with smaller differences between Barrier site and No-Barrier site) than the sound- reflecting barrier lines for the lower frequency range. If narrow-band analysis were to reveal nothing more than what is currently seen with the one-third octave band data, one could assume that sound-absorbing barrier surfaces affect the reflected sound by reducing the comb-filtering effect. This reduction in effect cannot be substantiated, however, until narrow-band analysis is applied (which was beyond the scope of this study). Research into reasons why the barrier sites would have higher sound levels in these harmonically related frequencies revealed comb filtering as a cause. Comb filtering is an effect created by a direct-path sound wave combining with a reflected-path sound wave, where the reflected-path sound is delayed in time from the direct path. The combination results in harmonically related peaks in the received sound spectral content, where the two sound waves generate constructive interference. (For fre- quencies that arrive in phase, constructive interference increases the amplitude; for frequencies that arrive out of phase, destruc- tive interference decreases the amplitude.) Harmonically related peaks can result in perception of tonality. In audio engineering, the comb-filtering effect is described as sounding “metallic,” “boxy,” or “artificial,” and can make higher frequencies sound odd or harsh. Given that these descriptions are about recording studios, where reflecting walls are close by, the delay between direct and reflected waves is quite small (i.e., a few milliseconds [ms]). In psychoacoustics, comb filtering is discussed in terms of coloration of sound caused by a single reflection (Johansen 2006). The effect depends on the delay time of the reflected sound. Humans are particularly sensitive to coloration caused by delays of about 5 ms. Coloration can be regarded as a frequency domain effect (a change in timbre) for delay times up to approximately 25 ms. If the delay time exceeds 25 ms, the perception changes from coloration to a rough character effect, and the regular repetition is detected as a time domain effect. Further psychoacoustic evaluation describes “repetition pitch.” Signals consisting of a sound, together with a repetition of that sound after a delay time t, can evoke well-defined pitch sensations corresponding to 1/t (Bilsen and Ritsma 1970). For example, if a delay of 10 ms corresponds to a repetition pitch of 1/0.010 = 100 Hz. Longer delay times result in lower frequency repetition pitches. If the delay is too long, the effect would be perceived as an echo. The delay times associated with typical highway geom- etries could range from about 8 ms to 200 ms, depending on the number of travel lanes, median width, barrier placement, and placement of the vehicle (upstream or downstream). At the time of maximum sound level, the time delay is greater than when vehicles are upstream or downstream. For a single vehicle passing by, it is possible to perceive comb-filtering effects as the vehicle approaches, likely as more of an echo or fuller sound at the closest point of approach, and again as comb-filtering effects as the vehicle recedes. To help understand these effects, the audio file for a vehicle pass-by event at a No-Barrier site was delayed in time and added to the original event to simulate the barrier effect. The delay times ranged from 20 ms to 200 ms. For the 20-ms delay, the effect was an obvious raspiness or buzziness, and that effect decreased as the delay time increased. A test also was done with a sweeping time delay, starting at 20 ms upstream, increasing to 100 ms at the closest point of approach, and decreasing to 20 ms downstream. The result was as expected: raspiness or buzziness that decreased as the vehicle approached, a full sound at the closest point of approach, and increasing raspi- ness or buzziness as the vehicle receded.

135 Figure 145. I-75—spectrogram difference plot for sound-absorbing barrier heavy vehicle (truck) pass-by event at }18:33, BarCom04 and NoBarCom06.

136 Figure 146. I-75—differences in levels (dB) at sound- absorbing barrier, –0.25 seconds from maximum during heavy truck pass-by event at }18:33, BarCom04 and NoBarCom06. geometry. Although the audible range for humans typi- cally is described as being 20 Hz to 20,000 Hz, for very low frequencies (below 20 Hz, or infrasound), the sound may be slightly audible or perceived as vibrations in various parts of the body. It is possible that very low frequencies enhanced by the presence of a barrier may be contribut- ing to the objectionableness of the sound. However, the low frequencies are only in the 30 dBA to 40 dBA range, so tonal enhancements at these low levels may go unno- ticed. Understanding possible low-frequency effects would require further investigation. • Absorptive barriers may reduce the comb-filtering effect. Comparison Finding 12: The Barrier Reflections Screening Tool Can Be Used to Estimate Barrier Reflection Effects The Barrier Reflections Screening Tool provides a conservative estimate of the barrier-reflected effect and is appropriate for use in screening for potential adverse effects due to a reflective or absorptive noise barrier on the opposite side of a highway. The Barrier Reflections Screening Tool was developed to provide quick estimates for the expected increase in traffic noise due to reflections from a barrier on the opposite side of the road. This tool focuses on a single variable to estimate the barrier-reflected noise effect at receptors opposite a noise barrier: path-length difference (comparing the path-length for direct sound and for barrier-reflected sound). The tool uses this variable to conservatively estimate the increase in traffic noise based on the geometrical spreading of sound from a line source. The results can be used to determine areas where detailed evaluation of reflected noise and abatement options should be considered. Additional information is pro- vided in Appendix D. The Barriers Reflections Screening Tool is available for download using a link on the NCHRP Research Report 886 webpage. For validating the estimates generated by the tool, pre- dicted barrier effects were compared to measured effects for all measurement sites, including both reflective and absorptive barriers. Figure 148 uses black circles to show the estimated effect (from the screening tool). The figure compares the estimated effect to the range of measured barrier reflection effects, both unweighted (blue) and A-weighted (red). The distances listed for each site repre- sent the perpendicular distance from the traffic centerline to the receivers. For site I-270, a slightly modified version of the screening tool was used to estimate the effect of reflec- tions from the homes behind the microphone locations. These time-delayed audio files were then subjected to a spectrogram analysis. For the analysis, the sound levels between Barrier and No-Barrier cases were not adjusted; the only effect applied was the time offset. A difference plot for the 19:46 event at MD-5 was generated for a sweeping time delay, which simulates a real pass-by event (although the time delays do not exactly match those that would apply to the actual MD-5 event). The difference plot exposed lines of hot spots, as was seen for measured data. The spectrum at a slice in time found peak difference frequencies at 12.5 Hz, 20 Hz, 63 Hz, 100 Hz, 200 Hz, 500 Hz, 800 Hz, 1,250 Hz, and 12,500 Hz. With one exception (20 Hz), all the frequencies are harmonically related to 12.5 Hz. In summary, adding a sweeping time delay to a pass-by event and combining it with the original event simulates the barrier reflection effects, and this results in the comb-filtering effect. To summarize the perception of barrier reflections: • In addition to an increase in sound levels described in earlier findings, comb-filtering effects are adding tonal qualities to received sound, particularly in low to mid- frequencies. These additions would apply to both near and far distances from the road. This may change the sound quality by adding a raspiness or buzziness to the sound, particularly as a vehicle is approaching or receding. • The delay times associated with reflected sounds with the sites measured are about 8 to 200 ms, which may result in very low-frequency repetition pitches (5 Hz to 125 Hz). These frequencies and their harmonics create a comb-filtering effect that is dependent on each site’s

137 Site and Event Time Peak Difference Frequencies (Hz) Harmonic Relationships MD-5, 19:46 16, 20, 40, 80, 160, 400, 800, 1,600, 12,500 All except 16 are related to 20. Alternatively, all except 20 and 40 are related to 16. MD-5, 20:09 25, 50, 100, 200, 400, 1,000, 1,600, 6,300, 10,000 All are related to 25. I-90, 14:41 16, 25, 31.5, 50, 63, 125, 250, 500, 1,000, 2,000, 3,150, 5,000, 8,000 All except 16, 31.5, and 63 are related to 25. I-90, 16:17 16, 25, 50, 63, 100, 250, 500, 2,500 All except 16 and 63 are related to 25. SR-71, 10:44 25, 50, 80, 315, 400, 630, 800, 2,000, 2,500, 6,300, 8,000, 12,500 All except 80, 315, and 630 are related to 25. SR-71, 12:10 25, 50, 80, 400, 1,250, 2,500 All except 80 are related to 25. I-75, 18:18 16, 31.5, 50, 80, 200, 400, 630, 1,250, 4,000, 6,300, 12,500 All except 16, 31.5, 80 and 630 are related to 50. Alternately, all except 50 and 200 are related to 16. I-75, 18:33 20, 31.5, 63, 80, 160, 315, 400, 1,000, 2,000, 3,150, 6,300, 10,000 All except 20, 80, 160, 400, 1,000, 2,500, and 10,000 are related to 31.5. Alternately, all except 31.5, 63, 315, and 3,150 are related to 20. I-70, 17:44 12.5, 20, 25, 100, 160, 250, 315, 630, 1,000, 2,500, 4,000, 5,000, 6,300, 10,000 All except 12.5, 25, 250, 315, and 630 are related to 20. Alternately, all except 20, 315, and 630 are related to 12.5. I-70, 17:50 12.5, 20, 31.5, 50, 63, 100, 200, 630, 1,000, 2,500, 4,000, 6,300, 8,000, 12,500 All except 12.5, 31.5, 50, 63, and 630 are related to 20. Alternately, all except 20, 31.5, 63, and 630 are related to 12.5. Table 20. Peak difference frequencies for sound-reflecting and sound-absorbing barriers. Figure 147. Spectral difference plots for vehicle pass-by events for all sound-reflecting and sound-absorbing barrier sites. The effect was applied to the measured data to compare it to screening tool estimates. Key observations of the validation analysis results (addressed further in Appendix D) are as follows: • All the estimated values fall within or slightly above the top of the measured value ranges. This result indicates that the screening tool provides a conservative estimate of the barrier-reflected effect and is appropriate for use in screen- ing for potential adverse effects due to a noise barrier on the opposite side of a highway. • The barrier effect appears to be dominated by path lengths. Any trends related to barrier absorption likely are masked. It could be expected that the screening tool would over- estimate the barrier effect for absorptive barriers (sites I-75, I-70, and I-270), but because the absorptive barrier sites all had geometries with fairly large propagation dis- tances for the reflected path, that expectation was not real- ized in this study.

138 -2 -1 0 1 2 3 4 5 SR-71 (72 ft.) SR-71 (447 ft.) MD-5 (122 ft.) I-24 (137 ft.) Briley Pkwy (129 ft.) I-90 (114 ft.) I-90 (138 ft.) I-70 (141 ft.) I-75 (155 ft.) I-75 (105 ft.) I-270 (178 ft.) I-270 (97 ft.) Ch an ge in N ni se D ue t o Ba rr ie r (d B) Measurement Location Measured unweighted effect range (dB) Measured A-weighted effect range (dBA) Estimated effect (dB) Figure 148. Estimated barrier effect compared to A-weighted and unweighted results.

139 Applications, Conclusions, Recommendations, and Suggested Research Applications Applications of the work follow, broken into sections focused on the FHWA Method of data analysis, the spectro- grams, the psychoacoustics analysis, the layperson’s guide, and the Barrier Reflections Screening Tool. FHWA Method The most immediate application of these results is the understanding by traffic noise analysts and abatement prac- titioners that traffic sound levels and sound characteristics for receptors across from a proposed single reflective noise barrier can change after the installation of the barrier. The sound level increases can be seen in elevated broadband A-weighted sound levels and unweighted sound pressure levels with changes highlighted in certain frequency ranges. Also, the traffic noise background levels can be elevated, sus- taining the sound of individual vehicle passage, as discussed below for spectrograms. This understanding can lead to the appropriate specifica- tion of sound-absorbing surfaces on these single barriers, especially for highway widenings where the roadway already exists—such as in a Type I widening project or a Type II retrofit noise abatement project (both as defined in current state highway agency noise policies). The findings that the A-weighted sound levels and certain one-third octave band sound pressure levels showed some increases across from the sound-absorbing barriers are of interest, but need to be viewed with caution: two of the studied sound-absorbing barriers had NRC values of 0.80. The measured sound level increases might not be seen for barriers with higher NRCs. In many cases, the receptors across from a single barrier may have been found to be impacted by a proposed proj- ect in the project’s noise study. However, they did not qualify for abatement based on acoustical or physical feasibility rea- sons or because they did not meet a state highway agency’s noise abatement reasonableness criteria in its traffic noise policy. The specification of sound-absorbing single barriers can be a proactive step to not make the situation worse for such impacted receptors. The screening tool developed in this research can help in estimating the size of the increase in A-weighted sound levels due to the reflections. In other cases, the receptors may not be impacted to the same degree or in the same way; nonetheless, the results of this study suggest that such receptors can experience changes in the noise environment due to the introduction of the bar- rier on the opposite side of the road. Although it might be dif- ficult to justify making the barrier sound-absorbing in such a case, the practitioner at least has a better idea of the nature of the phenomenon. The layperson’s guide provides a tool for explaining the subject to the public and non-acoustical professionals. Spectrograms Spectrograms provide a color-coded visualization of fre- quency and temporal characteristics of highway traffic noise. Such visualizations can help to reveal differences when com- paring a site with and without a noise barrier, which may be useful when trying to explain the effect of barrier reflec- tions. Spectrograms can show a clear visual difference when a reflective barrier is present and only subtle visual differ- ences with an absorptive barrier present. A three-site com- parison with difference spectrograms could be made using a site without an opposing barrier, a site with a reflective barrier, and a site with an absorptive barrier. The difference spectrograms can help to explain the changes in sound qual- ity that occur with a barrier present and with an absorptive barrier present. The spectrograms and difference spectro- grams could help policy makers provide guidance on when it may be effective to use sound-absorbing barriers and help in showing the public what benefits an absorptive barrier could provide. C H A P T E R 7

140 Psychoacoustics Because the derived annoyance metrics were either uncor- related with site location or were contra-indicative for con- tinuous flow traffic, their direct use (as applied in this work) cannot be recommended for indicating increased annoy- ance due to single barriers in the presence of heavy traffic. However, these metrics are useful in cases where the overall roadway sound is dominated by individual vehicle signatures. This is because time-varying annoyance metrics are domi- nated by the effects of short-term loud events. Layperson’s Guide As part of this research, a four-panel foldout was cre- ated as a layperson’s guide (see Figure 149). This pamphlet, titled “Reflected Sound from Highway Noise Barriers,” is intended to provide state highway agencies a document they can provide to concerned citizens and other non-professionals that explains how noise is reflected from single barriers to the opposite side of the road, how the sound changes after installation of the barrier, and how absorptive barriers help to reduce impacts in some circumstances. The layperson’s guide is provided copyright free as a Microsoft Word document (.docx) file. Individual state high- way agencies are encouraged to add their own logos and modify the content as needed to suite their specific needs. The document can be downloaded from the NCHRP Research Report 886 webpage. Barrier Reflections Screening Tool The spreadsheet-based screening tool developed as part of this research is recommended for use in estimating the effect of barrier reflections for projects with communities oppo- site a noise barrier. The Barrier Reflections Screening Tool focuses on a single variable to estimate the barrier-reflected noise effect at receptor: path-length difference (comparing the path-length for direct sound and for barrier-reflected sound). The tool uses this variable to conservatively estimate the increase in traffic noise based on the geometrical spread- ing of sound from a line source. The results can be used to determine areas where detailed evaluation of reflected noise and abatement options should be considered. The Barrier Reflections Screening Tool can be downloaded from the NCHRP Research Report 886 webpage. Conclusions This research investigated potential changes in the sound characteristics opposite a single highway noise barrier due to the reflection of the traffic noise off that barrier, first for sound-reflecting barriers and then for sound-absorbing bar- riers. The results indicate that sound characteristics differ in these situations. The differences are a function of the dis- tance from the road and, to some extent, whether the high- way sound is continuous or dominated by individual vehicle pass-bys. The first type of change examined was an increase in the broadband A-weighted equivalent sound level and unweighted equivalent sound pressure level. In several cases, as a reference indicator of the potential for reflections, microphones were placed on the barrier side of the road between the barrier and the road. The results indicate that increased sound levels are seen for both sound-reflecting and sound-absorbing barriers. These increases are greater at this position than across the road at similar distances from the road. The increases in front of the sound-absorbing barriers need to be viewed in the context that two of the barriers were made of the same proprietary product with an NRC of 0.80 (80% absorptive). These results should not be generalized to apply to all sound-absorbing products, especially those with higher NRC ratings. Opposite the barrier near the road, evidence exists of slightly increased sound levels. The increases are greater for more distant receivers on the opposite side of the road. These increases are less apparent for the sound-absorbing barriers. The second type of change examined was seen in the fre- quency spectra of the received sounds. These changes appear to be greater at the lower microphones than at the higher micro- phones, possibly because of different ground effects on the direct and reflected sound propagating to the receivers. There appears to be an enhancement in the spectrum in the 1 kHz to Figure 149. Front panel, layperson’s guide (pamphlet).