Analysis of Spatial and Temporal Variation of Noise Level at Intersections of a Mid-Sized City in India Adarsh Yadav 1 Department of Civil Engineering, Indian Institute of Technology Roorkee Roorkee, Uttarakhand, India, 247667 Manoranjan Parida 2 Department of Civil Engineering, Indian Institute of Technology Roorkee Roorkee, Uttarakhand, India, 247667 Brind Kumar 3 Department of Civil Engineering, Indian Institute of Technology (BHU) Varanasi Varanasi, Uttar Pradesh, India, 221005

ABSTRACT This study aims to determine spatial and temporal noise variation and acoustical climate at intersections. Data monitoring comprises measurement of peak and off-peak hour variation at 19 intersections between February 2020 to September 2021 in Kanpur, India. Kanpur was proposed as a smart city in 2015 under National Smart Cities Mission. This work has been part of a National Level Project “IMPRINT” to document ambient noise levels for a Tier 2 city with highly heterogeneous traffic. Spatial variation of noise is larger (>15 dBA) than temporal variation (<5 dBA). Noise variations are primarily influenced by traffic flow, geometric characteristics, classified volume, and honking events. Frequent stoppage of vehicles for boarding and alighting of passengers also significantly affect noise levels. Noise risk zones are identified based on noise levels alongside intersections. Zones are classified as safe (<66 dBA), tolerable (66-71 dBA), low risk (71-76 dBA), moderate risk (76-81 dBA), high risk (81-86 dBA), and extremely high risk (>86 dBA) zones. Noise level crosses tolerable limit at most study locations, and some locations fall in high-risk zones. The study has highlighted the influence of different parameters on spatial and temporal noise variation at intersections and remedial action plans for traffic noise abatement. 1. INTRODUCTION

The population shift towards urban cities has increased rapidly in the past few decades. A recent United Nations report estimates that 55% of total world population are urban dwellers and projects that by 2050 the urban population will increase to 68% (United Nations, 2019). Urbanisation demands growth in the road transport sector (Maparu and Mazumder, 2021). Environmental pollution (i.e., air, noise etc.) is a byproduct of road transport and urbanisation. Noise pollution has emerged as a serious threat in the vicinity of road networks and ranked 2nd most environmental stressor and threat in urban cities. Traffic noise exposure may lead to various health-related issues such as stress, annoyance, headache, hearing damage, high blood pressure, and myocardial

1 ayadav4@ce.iitr.ac.in 2 m.parida@ce.iitr.ac.in 3 kumar_brind.civ@iitbhu.ac.in

infarction (Fyhri and Klæboe, 2009; Gilani and Mir, 2021; Yoshida et al., 1997). Central pollution control board (CPCB), New Delhi, prescribed day and night ambient sound levels for silence, residential, commercial, and industrial land uses are 50, 55, 65 and 75 dBA (Ministry of Environment and forest, 2000). The world health organisation (WHO) has advised that exposure to noise levels above 55 dBA is not acceptable (WHO, 2018). Road transport contributes more than 66% to urban noise pollution (Thakre et al., 2020). 73% of Brazilians in Curitiba city have reported traffic noise as the leading cause of annoyance (Calixto et al., 2003).

Traffic noise is emitted from different sub-sources of the vehicle, which is further grouped in power unit noise (engine, fan, exhaust, and transmission, etc.), aerodynamic (related to turbulent airflow around the vehicle), and tire/pavement interaction noise. Pavement influence was significant, and reasphalting improved the noise level from 3 to 8 dBA (Sánchez-Sánchez et al., 2018). Traffic noise is majorly influenced by traffic characteristics along with road characteristics, infrastructure characteristics and meteorological parameters (Ibili et al., 2021). Garg et al. (2017) measured the noise level at 35 locations in different land-use of 7 megacities from 2011-2014 under the National Ambient Noise Monitoring Network (NANMN). Traffic parameters were dominant parameters, and a higher noise level was reported commercial zone. Mishra et al. (2021) have considered the covid-19 lockdown effect on the noise level variation. An adequate reduction in the noise level is reported during the lockdown period. The commercial zone was identified as the noisiest location among the different land uses during and post covid lockdown periods. The daytime noise level is recorded as higher than the nighttime noise level. Mishra et al. (2010) measured the noise level alongside the bus rapid transit systems (BRTS) in Delhi. Ozer et al. (2014) monitored the noise level variation in a college campus. The study identifies road traffic and bus movement as significant contributors to traffic noise pollution. Dense and good plantation may reduce the noise level by 10-15 dBA alongside the roadway.

Traffic noise varies spatially and temporally even within the city due to variations in traffic parameters. Since traffic parameters and infrastructure constructs vary in the different land uses. Also, traffic parameters change with the change in time. Therefore, the current study aims to analyse the temporal and spatial variation of noise.

2. SITE SELECTION AND DATA COLLECTION

For the present study, sites were selected in Kanpur city of India. Kanpur is known as the industrial capital of Uttar Pradesh province and is listed as a smart city under the National Smart City Mission of the Government of India. 19 intersections were identified and further grouped into 4 categories such as commercial, residential, silent and industrial zones. The land wise details of these locations are presented with their location ID in the following table 1.

Table 1: Details of study locations

Location name Location ID Land use

Bada Chauraha I1 Commercial

Gaushala Chauraha I2 Residential

Shastri Chowk I3 Residential

Dabauli Chauraha I4 Residential

Tat Mill Chauraha I5 Commercial

Ghantaghar Chauraha I6 Commercial

Kidwai nagar Chauraha I7 Commercial

Maryampur Chauraha I8 Silence

Fazalganj Chauraha I9 Commercial

Kesha Colony I10 Residential

New Keshav Puram I11 Residential

Company Bagh Chauraha I12 Silence

Vijay Nagar Chauraha I13 Commercial

LML Intersection I14 Industrial

Narauna Chauraha I15 Commercial

Harjendra Nagar I16 Residential

Sachan Chauraha I17 Silence

Chhappan Bhog Chauraha I18 Residential

Deoki Chauraha I19 Commercial

The intersections are locally named churaha or chowk. These sites were located away from known noise sources to include traffic impact on the measured noise level. Road surfaces were smooth with the gradient lesser than 3%. Further, data collection was carried out in six different phases (from February 2020 to September 2021) to capture the seasonal, festival, weekend, and weekday variations. In each phase, continuous 3 hours of data were collected at each location to capture the peak and off-peak hour variation. A preliminary survey was carried out to finalise the study locations for performing the measurements. Based on the survey, intersection legs with greater importance and significant traffic volumes are identified. Simultaneously, separate measurement was carried out at both the exit and entrance arm of the selected leg. The exit arm is indicated as the arm through which a vehicle leaves the intersection, while the entrance arm is utilised to enter in road intersection. The literature has investigated that the intersection influence zone lies up to 150-300 m from the signal stop line (Abo-Qudais and Alhiary, 2004). In the present study, the instruments were installed within 100 m from the signal stop line to capture the influence of the intersection. Data monitoring includes measurement of sound level along with the traffic volume, speed, weather parameters, road geometry and other geometric distances. The equivalent noise level, traffic volume, speed, and weather parameters were measured using 01 dB sound level meters, video camera, radar gun, and Schelt technology weather station. Road geometry and geometric distances were measured using the measuring wheel or tape. To meet the objective of the study, hourly data are recoded at exit and entrance arms separately and presented in table 2.

Table2: Variables and Measured Parameters

Variables Measured Parameters

Sound level Hourly equivalent noise level (L eq ) Traffic Parameters Total traffic volume, Classified traffic volume, classified stream speed, average speed, number of honking events

Weather Parameters Atmospheric temperature, Pressures, Relative humidity, wind speed and direction

Road geometry and other distances Number of lanes, Carriageway width, Median width and height, distance of SLM from centerline of nearside and farside carriageway, distance of SLM from façade

3. Results and Analysis

The study has determined major influencing parameters and their association with the spatial and temporal variation of traffic noise levels at intersections. Further, noise risk zone is also identified based on the recorded noise levels. The result and analysis section has been classified into the following subsections: identification of influencing parameters, spatial-temporal variation, and noise risk zones.

3.1. Identification of Influencing parameters

Traffic noise is influenced by the various elements available on road traffic systems. Different parameters are measured at the study locations as discussed earlier in the site selection and data collection section. The equivalent noise level (L eq ), total traffic volume (Q), average speed (V), number of honks (H), distance of sound level from effective source (D g ), percentage of two- wheelers (%2W), percentage of car (%Car), percentage of e-rickshaw (%E-Rick), percentage of cycle (%Cycle), percentage of light commercial vehicle (%LCV), percentage of Vikram-rickshaw (%Vik), percentage of heavy vehicles (%HV), temperature (T), wind speed (W s ), relative humidity (RH) and atmospheric pressure (Atm) are utilised in the study for analysis. The distance of sound level from the effective source (D g ) is determined by taking the geometric mean of SLM distance from the centerline of near and far lanes. A statistical package for social science (SPSS v27) is used for carrying out statistical analysis in the current study. Traffic noise influencing parameters are identified based on a partial correlation test. The correlation of each parameter is calculated with equivalent noise levels ( 𝐿 𝑒𝑞 ) reported at exit and entrance arm. The correlation coefficient and descriptive statistics of different parameters are presented in table 3.

Table 3: Correlation matrices and Descriptive statistics

Coefficient of correlation Descriptive statistics L eq (ext) L eq (ent) Mean Standard Deviation L eq (ext) 1 0.85 73.5 3.4 L eq (ent) 0.85 1 73.7 3.5 Q ent 0.43 0.50 1928.21 946.76 Q ext 0.59 0.60 1995.25 929.06 V ent -0.25 -0.25 22.06 3.28 V ext -0.25 -0.23 23.20 3.46 H ent - 0.40 551.94 271.73 H ext 0.38 - 568.80 270.52 Dg ent - -0.43 12.79 3.81 Dg ext -0.37 - 13.28 4.27 %2W ent -0.09 -0.02 56.65 7.57 %2W ext -0.05 -0.02 57.01 7.12 %Car ent -0.21 -0.25 15.34 5.32 %Car ext -0.21 -0.20 15.10 4.85 % Vik ent 0.35 0.37 9.66 4.08 % Vik ext 0.25 0.27 9.45 4.20 %E-Rick ent 0.25 0.25 10.50 5.75 %E-Rick ext 0.16 0.20 10.73 5.75

%Cycle ent -0.01 -0.07 1.67 4.51 %Cycle ext 0.14 0.05 1.63 4.73 %HV ent -0.27 -0.30 5.10 4.08 %HV ext -0.32 -0.36 5.03 4.20 %LCV ent -0.22 -0.27 0.93 1.65 %LCV ext -0.23 -0.27 0.96 1.61 T -0.18 -0.21 28.97 5.54 Ws 0.06 0.08 1.44 1.88 RH 0.05 0.00 51.78 15.41 Atm 0.14 0.20 1001.00 13.28 As per table 3, the mean hourly equivalent noise level is almost equal both at the exit and entrance arm, which means the entrance and exit arm are almost equally acoustically polluted. In the total traffic volume, the percentage of two-wheelers contribution is dominant (more than 55%). While car, cycle and e-rickshaw contribution are significant and more than 15%, 10% and 9% respectively. The other vehicles, such as LCVs, Vikram-rickshaw and HVs contribution were not very significant in the traffic stream. Traffic speed is lower than 25 km/h and higher in the exit lane compared to the entrance lane. The mean value of temperature, wind speed, relative humidity and atmospheric pressure is 28.97 0 C, 1.44 m/s, 51.78% and 1001.00 mBar. Traffic noise is influenced by traffic, infrastructure, and atmospheric parameters. A partial correlation test was carried out to identify the correlation between measured variables and equivalent traffic noise. Correlation strength between two variables is determined by estimating the correlation coefficient (R). R-value greater than 0.5 indicates a higher correlation, and lesser than 0.3 represents a lower correlation. There is a good correlation between two variables if cofficient value is between 0.3 and 0.5. The maximum R-value can be 1, and the higher value indicates the high correlation of the variable with equivalent noise level. For example, if X and Y are highly correlated, Y indicates a higher dependency on the variable X and varies significantly with the variation of X. The R-value between L eq (ext) and L eq (ent) is 0.85, which indicates a high correlation. Traffic volume has a high correlation with the equivalent noise level, which means traffic noise is highly influenced by traffic volume. Speed has a low correlation with the noise level. Additionally, D g and L eq are significantly correlated, indicating that the position of sound level meters significantly affects the equivalent noise level. 2W, cars, e-rickshaw, and heavy vehicles contribute significantly to the traffic stream, but their correlation is low with the traffic noise level. Besides, the contribution of Vikram-rickshaw and heavy vehicles is low in the traffic stream, but their correlation with the traffic noise level is significant. The correlation between Vikram rickshaw at entrance and Heavy vehicles at exit arm is comparatively high. Yadav et al. (2021) have identified heavy vehicles and Vikram-rickshaw as the noisiest vehicle in the traffic stream due to higher engine noise. The study evaluated the lower correlation between the atmospheric parameters and equivalent noise level, which means the influence of atmospheric parameters is not significant on the noise level.

3.2. Spatial and Temporal Variation of Noise

The identified locations are classified into commercial, residential, silence, and industrial uses and further noise variation was analysed into different land use. The total equivalent noise level, Leq (T), is calculated at the roadway section to estimate the spatial and temporal variation. Leq (T) is calculated at each study location with the addition of Leq (ext) and Leq (ent) logarithmically as per Equation 1.

𝐿 𝑒𝑞 (𝑇) = 10log⁡(10 𝐿𝑒𝑞(𝑒𝑛𝑡)/10 + 10 𝐿𝑒𝑞(𝑒𝑥𝑡)/10 ) (1)

Traffic noise varies spatially within the city with a variety of different constructs. The survey reported Tat Mill, Bada Chauraha, Narauna Chauraha, Vijay Nagar, and Ghantaghar as the noisiest

location. Tat Mill and LML are reported as noisiest and quietest with the measured value of 85.2 and 69.5 dBA. The reason for this can be explained by the fact that traffic movement and honking were much less at LML Chauraha than at Tat Mill. The sound level meter (SLM) was placed at the edge of the pavement at Tat Mill Chauraha due to lack of space; otherwise, SLMs were placed nearer the building line at other selected locations. However, traffic flow was higher at Narauna Chauraha than at the Tat Mill, but the noise level was lesser at Narauna compared to the Tat Mill. Narrow road stretch, high congestion and greater honking lead to higher noise pollution at Tat Mill. Noise level crosses the Central Pollution Control Board (CPCB), Delhi, standards at almost every location except LML Chauraha. Due to industrial activities, the CPCB has prescribed a higher ambient noise limit in industrial areas. Although noise level crosses the prescribed limit by the World health organisation (WHO) at every location. As per figure 1, it can be analysed that the noise level is generally high at intersections located in the commercial area.

Figure 1: Spatial variation of traffic noise

Temporal variation is analysed by measuring the noise level in different months and times of day at the selected locations. Continuous three-hour measurement was carried out in different months of the year. Figure 3.2 indicates the temporal variation of noise levels at these locations. The commercial, residential, silence, and industrial locations are presented with red, yellow, blue, and violet bars. The black and green line represents the CPCB and WHO standards. The figure indicates that the monthly noise variation is up to 8 dBA, while daily noise variation rarely goes beyond 3 dBA. The maximum monthly variation is almost 8 dBA recorded at Narauna Chauraha. Fazalganj, Dabauli, LML, Sachan and Chhappan Bhog Chauraha have monthly noise variation in the range of 5-6 dBA. While monthly variation was lower than 5 dBA at other rest locations. Daily temporal variation is less than 2.5 dBA at most of the selected locations, noise level goes beyond 2.5 dBA only a few times at some locations. The maximum variation of almost 6 dBA was reported only once at LML Chauraha. The temporal variation was analysed in terms of the month and daily variation. The monthly noise variation was little significant due to changes in traffic volume, vehicular percentage in the traffic stream, and the number of honking events. While daily temporal variation was not very significant. Higher temporal noise variation was reported in the area with the lower traffic volume. Although variation pattern was not uniform at the intersections in the Kanpur city. Traffic characteristics do not change significantly over a period of time in a day. Therefore, daily variation was not significant. It was majorly influenced by the traffic characteristics, not by the change of time of day or month of the year.

Figure 2: Temporal Variation of Noise level

3.3. Noise Risk Zones

Noise risk zones are identified based on the measured ambient noise level. The zones are further categorising into 6 zones: Safe (<66 dBA), tolerable (66-71 dBA), low risk (71-76 dBA), and moderate risk (76-81 dBA), high risk (81-86 dBA), extremely risk (>86 dBA) (Banerjee et al., 2008). A total of 19 intersections were selected from the different land uses. Figure 2 indicates that the noise level mostly lies within the tolerable limit at LML Chauraha. Although, noise level crosses the tolerable limit sometimes at the LML intersection. Dabauli, Fazalganj, Harjendra Nagar, Kesha colony, New Keshav Puram, Sachan, Chhappan Bhog, and Deoki intersections are categorised in the low-risk zones (71-76 dBA) based on the recorded noise level. These locations are primarily under the category of low-risk zones, but sometimes noise level crosses the limit of low-risk zones at these locations except Kesha colony and New Keshav Puram. Noise level is mainly in the range of 76-81 dBA at Bada Chauraha, Gaushala, Shastri Chowk, Maryampur, Tat Mill Chauraha, Ghantaghar, Vijay Nagar and Company Bagh Chauraha, as presented in Figure 2. Therefore, these locations are identified as moderate risk zones based on the recoded noise level. Kidwai Nagar Chauraha has a noise level between 75.23 and 78.12 dBA; therefore, it is categorised in low and moderate risk areas.

Traffic noise is major annoying source of noise in midsized urban cities. In the past few decades, mid-sized cities are expanding, which demands rapid growth in the transportation sector. Therefore, traffic noise has grown a lot in the past few years. The intersection locations were selected from different parts of Kanpur city to understand ambient noise due to road traffic. Traffic noise levels at

these locations exceeded the safe zone limit. Most of the locations fall under low risk and moderate risk zones. While sometimes, the noise level exceeds these limits and reaches high-risk zones.

Conclusion

Traffic noise has become the most intrusive source of noise pollution in midsized urban cities with the growth of transportation sectors. Noise level fluctuates with variation in the influencing parameters. Traffic volume, honking and type of vehicle is identified as major source of traffic noise at the intersection. Spatial variation was significant, and a higher difference was reported almost 15 dBA. Temporal variation was analysed in terms of the time of the day and month of the year. It was identified that temporal variation was not significant in mid-sized cities. Monthly variation was up to 3 to 8 dBA, while daily variation rarely goes beyond 2.5 dBA. Traffic characteristics remain almost the same throughout the day in mid-sized cities like Kanpur. The traffic noise level has crossed safe standards prescribed by CPCB, New Delhi. While a higher difference is reported between the measured noise level and WHO safe limit at every location. The acoustic climate has crossed the safe limit and falls in the low and moderate risk zone. Sometimes, the traffic noise level increases a lot and reaches high-risk zones. Since Kanpur city was proposed in the list of smart cities in 2015 under the National Smart Cities Mission by the Government of India. Therefore noise mitigation and prevention measures must be incorporated to control traffic noise levels. Acknowledgement

This work was financially supported by the Ministry of Education and Ministry of Urban Development under the initiative of Impacting Research Innovation and Technology (IMPRINT, an initiative of the Government of India). The author is very grateful to the Indian Institute of Technology (IIT) Roorkee for providing the research facility. References

1. Abo-Qudais, S., Alhiary, A., 2004. Effect of distance from road intersection on developed

traffic noise levels. Can. J. Civ. Eng. 31, 533–538. https://doi.org/10.1139/L04-016 2. Banerjee, D., Chakraborty, S.K., Bhattacharyya, S., Gangopadhyay, A., 2008. Evaluation

and analysis of road traffic noise in asansol, West Bengal. J. Inst. Eng. Environ. Eng. Div. 89, 9–16. 3. Calixto, A., Diniz, F.B., Zannin, P.H.T., 2003. The statistical modeling of road traffic noise

in an urban setting. Cities 20, 23–29. https://doi.org/10.1016/S0264-2751(02)00093-8 4. Fyhri, A., Klæboe, R., 2009. Road traffic noise, sensitivity, annoyance and self-reported

health-A structural equation model exercise. Environ. Int. https://doi.org/10.1016/j.envint.2008.08.006 5. Garg, N., Sinha, A.K., Dahiya, M., Gandhi, V., Bhardwaj, R.M., Akolkar, A.B., 2017.

Evaluation and Analysis of Environmental Noise Pollution in Seven Major Cities of India. Arch. Acoust. 42, 175–188. https://doi.org/10.1515/aoa-2017-0020 6. Gilani, TA, Mir, MS, 2021. Association of road traffic noise exposure and prevalence of

coronary artery disease: A cross-sectional study in North India. Environ. Sci. Pollut. Res. 28, 53458–53477. https://doi.org/10.1007/s11356-021-14582-2 7. Ibili, F., Adanu, E.K., Adams, C.A., Andam-Akorful, S.A., Turay, S.S., Ajayi, SA, 2021.

Traffic noise models and noise guidelines: A review. Noise Vib. Worldw. https://doi.org/10.1177/09574565211052693 8. Maparu, T.S., Mazumder, T.N., 2021. Investigating causality between transport

infrastructure and urbanisation: A state-level study of India (1991–2011). Transp. Policy 113, 46–55. https://doi.org/10.1016/j.tranpol.2020.03.008 9. Ministry of Environment and forest, 2000. The Noise Pollution (Regulation and Control)

Rules, 2000 12311, 1088–1569.

10. Mishra, A., Das, S., Singh, D., Maurya, AK, 2021. Effect of COVID-19 lockdown on noise

pollution levels in an Indian city: a case study of Kanpur. Environ. Sci. Pollut. Res. 28, 46007–46019. https://doi.org/10.1007/s11356-021-13872-z 11. Mishra, R.K., Parida, M., Rangnekar, S., 2010. Evaluation and analysis of traffic noise

along bus rapid transit system corridor. Int. J. Environ. Sci. Technol. 7, 737–750. https://doi.org/10.1007/BF03326183 12. United Nations, Department of Economic and Social Affairs, Population Division (2019).

World Population Prospects 2019: Highlights (ST/ESA/SER.A/423). 13. Ozer, S., Zengin, M., Yilmaz, H., 2014. Determination of the Noise Pollution on University

(Education) Campuses: a Case study of Ataturk University. Ekoloji 23, 49–54. https://doi.org/10.5053/ekoloji.2014.906 14. Sánchez-Sánchez, R., Fortes-Garrido, J.C., Bolívar, J.P., 2018. Noise monitoring networks

as tools for smart city decision-making. Arch. Acoust. 43, 103–112. https://doi.org/10.24425/118085 15. Thakre, C., Laxmi, V., Vijay, R., Killedar, D.J., Kumar, R., 2020. Traffic noise prediction

model of an Indian road: an increased scenario of vehicles and honking. Environ. Sci. Pollut. Res. 27, 38311–38320. https://doi.org/10.1007/s11356-020-09923-6 16. WHO, 2018. WHO environmental noise guidelines for the European Region. World Heal.

Organ. 2589–2593. https://doi.org/https://apps.who.int/iris/handle/10665/279952 17. Yadav, A., Parida, M., Kumar, B., 2021. Determination of passenger car noise equivalent

for mid-sized urban cities in India. Proc. INTER-NOISE 2021 - 2021 Int. Congr. Expo. Noise Control Eng. https://doi.org/10.3397/IN-2021-1505 18. Yoshida, T., Osada, Y., Kawaguchi, T., Hoshiyama, Y., Yoshida, K., Yamamoto, K., 1997.

Effects of road traffic noise on inhabitants of Tokyo. J. Sound Vib. 205, 517–522. https://doi.org/10.1006/jsvi.1997.1020