Exploring spatial and temporal heterogeneity of environmental noise in Toronto
Environmental noise (originating mainly from traffic) has been implicated as a risk factor for adverse cardiovascular outcomes. Systematic reviews of the associations between noise from transportation-related sources (e.g., road, air and rail) and cardiovascular diseases concluded that individuals exposed to higher levels of noise are at increased risk for hypertension and ischemic heart disease, making exposure to traffic-related noise a potential public health problem. Therefore, it is important to understand the magnitude of exposure and to use this information to determine associated health risks.
As a pilot study, noise levels of 10 selected locations in Toronto were measured continuously for a full week to explore temporal variability. In addition, 30-minute samples were taken over 200 locations within the Greater Toronto Area to explore the spatial variability. The objective of this study is to explore and describe the heterogeneity of environmental noise using these samples and additional information collected by Public Health Ontario.
Thanks to Ray Copes and Hong Chen, Environmental and Occupational Health, Public Health Ontario for providing this data.
Please address queries about the data to Lennon Li (Lennon.firstname.lastname@example.org), Analytic Services Unit, Knowledge Services, Public Health Ontario.
Exposure to traffic-related noise is ubiquitous in modern society. Yet, environmental noise (originating mainly from traffic) has been implicated as a risk factor for adverse cardiovascular outcomes. A systematic review of the associations between noise from transport sources (e.g., road, air and rail) and cardiovascular diseases concluded that individuals exposed to higher levels of traffic-related noise are at increased risk for hypertension and ischemic heart disease. Acute exposure to noise is hypothesized to activate sympathetic and endocrine stress, and over a longer term it may result in permanent vascular changes such as increasing stress hormones and blood pressure, which may predispose individuals to arterial hypertension.
Thus, exposure to traffic-related noise has the potential to pose a tremendous public health burden. Indeed, in a recent report from World Health Organization it is estimated that 3% of myocardial infarction cases (or 1629 new cases) in Germany in 1999 were due to noise from road traffic. To reduce potential health impacts that may result from environmental noise in Ontario, it is important to understand the magnitude of exposure to the population and to use this information to determine associated health risks. Unfortunately, the measurement of population exposure to environmental noise is virtually nonexistent at present in Ontario.
Public Health Ontario (PHO) took its initiative and collected environmental noise data in two cycles in different seasons, see next section for detailed description. Our primary objective is to understand the spatial and temporal patterns in the data collected in Cycle 1. Our secondary objective is to see how the Cycle 2 data compares with the Cycle 1 data, refer to the objective section for details.
Ideally, measurements taken over time and at multiple locations that cover the entire area are needed. However, this needs to be balanced with time constraints, equipment safety and human resources. Two sampling strategies were used to balance these competing needs: a) convenience sampling on weeklong sites to explore temporal variability, a total of 10 sites were selected for continuous measurement for a full week. b) Lattice (regular grids) with close pairs sampling on short-term monitoring sites to explore spatial variability. A total of 70 equally spaced sites (3km apart) were selected to ensure of the coverage of Great Toronto Area, and another 130 random locations were generated for mobile sampling. To avoid systematic bias due to sampling date, Toronto is divided into 40 regular grids with approximately 5 sites per grid and sampling order of each grid is randomized. Additional 41 locations were selected within 200 meters from these 200 sites to capture spatial correlation in short distances. As a result, a total 241 locations were sampled within Greater Toronto Area with measurement of 30 minutes each during 9am-5pm on week days. Information on traffic count and land use etc. for each site was also collected.
A sound meter was used to measure the level of noise at each sampling site. The measurements were taken continuously (fractions of a second) for 30 minutes at mobile sampling site and a week at weeklong sampling site. The meter measures average sound pressure levels and then converts those into a log-scale decibel reading (dB). Because the measurements are continuous, the data need to be average onto different time units (e.g., seconds, minutes) for output and analysis, this average is called Leq (equivalent steady sound level of a noise energy-averaged over time, see page 3 of this document). Note that this average is not the arithmetic mean of the raw measurement data therefore it is not sensible to take the arithmetic means of the Leqs. The datasets provided here give 30 minutes average Leqs for both mobile and weeklong sampling sites using the following formula:
where Li = level reading i (dB) at a fraction of a second and n = total number of readings in 30 minutes. You could directly use the Leqs in the data as the outcome in the analysis.
As a pilot study, the primary objective is to understand noise variability in both space and time within the Greater Toronto Area using and weeklong samples (Weekly_Sample.csv) and mobile data collected in cycle 1 (Cycle1_Mobile.csv). Specifically,
- Describe weekly temporal variations of noise levels in Toronto; do Torontonians experience higher levels of noise at certain times of the day and/or certain days of the week?
- Describe the spatial variations; are there locations in Toronto experiencing higher noise levels? How do the site characteristics (traffic count, land use etc.) contribute to the observed variations?
An additional 313 sites were also collected by PHO at a different season of the year. Among which, 100 sites were re-samples from selected sites in cycle 1 (Cycle1_rep.csv) and 213 new locations (Cycle2_Mobile.csv).
Use this dataset to validate, calibrate and update your findings. Specifically,
- Are the spatial variations from Cycle 2 similar to what is observed in Cycle 1?
- If they are different, explore the reasons and sources, otherwise update the findings.
1. WeeklySample (CSV)
- SiteID: ID of the long term sampled site
- Time: Time of the measurement at half an hour
- LEQ: Average sound levels for every half an hour
2. Cycle1_Mobile, Cycle2_Mobile & Cycle2_rep (CSV)
- SiteID: ID of the mobile sites
- GridID: Grid the sites belongs to, Grid is only used for randomization purpose
- LEQ: Average sound levels of the 30 minutes
- Date: Date when the sample was taken
- Start.time: Starting time of the sound meter
- End.time: Ending time of the sound meter
- Lat: Latitude of the site
- Long: longitude of the site
- Total.traffic: total number of vehicles passing during the sampling period
- Distexp: Distance to the nearest express way
- Com100re: Area of commercial land use within 100 meters
- Ind100re: Area of industrial land use within 100 meters
- Open100re: Area of open space within 100 meters
- Rec100re: Area of recreational land use within 100 meters
- Res100re: Area of residential land use within 100 meters
- Wat100re: Area of water body within 100 meters
- Pop2001_500m: Population density within 100 meters
- Van Kempen E, Babisch W. The quantitative relationship between road traffic noise and hypertension: a meta-analysis. J. Hypertens. 2012 Jun;30(6):1075–86.
- Sørensen M, Andersen ZJ, Nordsborg RB, Jensen SS, Lillelund KG, Beelen R, et al. Road Traffic Noise and Incident Myocardial Infarction: A Prospective Cohort Study. PLoS ONE. 2012 Jun 20;7(6):e39283.
- Gan WQ, Davies HW, Koehoorn M, Brauer M. Association of long-term exposure to community noise and traffic-related air pollution with coronary heart disease mortality. Am. J. Epidemiol. 2012 May 1;175(9):898–906.
- Babisch W. Road traffic noise and cardiovascular risk. Noise Health. 2008 Mar;10(38):27–33
- Diggle, Peter J., Ribeiro Jr, Paulo Justiniano, Model-based Geostatistic, Springer Series in Statistics, 2007
- Royster, Berger, Royster. The Noise Manual. American Industrial Hygiene Association, Akron, Ohio, 2003.