Noise measurements as proxies for traffic parameters in monitoring networks

Abstract

SummaryDue to their typically high traffic intensity, urban ring roads are characterized by high noise levels and high concentrations of airborne pollutants. Hence, such locations are often priority measurement locations in monitoring networks. Unfortunately, the high purchase and operational cost of most airborne pollutant sensors severely limits the number of such sensors that can be deployed, leading to a (too) limited spatial resolution. The present research describes how low cost microphones could be used as proxies for traffic parameters. We consider a 7 day measurement campaign for an urban ring road in Antwerp, Belgium, where noise levels and traffic parameters were measured simultaneously. Noise indicators are calculated and are used to construct models to estimate traffic parameters. It was found that a proper choice of noise indicators allows for the accurate estimation of traffic intensities and means vehicle speeds, both for light and heavy vehicles. Furthermore, the usefulness of these estimated traffic indicators in a monitoring strategy is assessed. Carbon monoxide, hydrocarbon and nitrogen oxide emissions are calculated with the pollutant emission model Artemis. By comparing the Artemis outputs when using measured and estimated traffic parameters as input, the suitability of the constructed models is assessed. Estimations of emitted airborne pollutants were shown to be accurate, leading us to conclude that there are indeed significant opportunities to use noise measurements as proxies for traffic parameter measurements for the estimation of airborne pollutant emissions.PACS no. xx.xx.Nn, xx.xx.Nn

Observation of Vehicle Axles Through Pass-by Noise: A Strategy of Microphone Array Design

“…The speed estimation problem has been also extensively addressed, e.g., in [11]- [16]. A recent trend consists in considering the pass-by noise as a measure of the energy consumption; in 2011, Can et al successfully showed the correlation between emitted airborne pollutant and road traffic noise near a highway [17]. This is a brief overview of what information traffic noise can provide.…”

Correlating transportation noise frequencies with ultrafine particulate emissions by source: Implications for environmental health studies

“…A key constraint in the development and refinement of such models is the difficulty in capturing noise frequencies that are representative of traffic sources, rather than reliance on the conventional Aweighted SPL, dBA for model construction. In a recent Belgian study (Can, Dekoninck, et al, 2011) engine noise and its corresponding low frequencies (around 100 Hz) were found to predominate at both ends of the speed spectrum: -at low speeds and during accelerations, with rolling noise containing many medium frequencies (around 1 kHz), dominant for vehicles speeds exceeding 50-70 kph (op. cit.…”

“…Predictive models have been developed to estimate measured total particle number (TPN) adjusted for meteorological conditions for specific noise indicators (L2KHz -L125Hz = low to med frequency noise bands (Can, Rademaker, et al, 2011). These models have shown strong correlation (Spearman 0.62) between measured and predicted TPN based on wind speed and wind direction.…”

Identification of Road Pavement Types Using Bayesian Analysis and Neural Networks

“…Figure 2 depicts the ranking for the features (by backward searching method and based on the inter-intra criterion) originally introduced in the texture and noise correlation study. [15][16][17] The procedure of ranking the features is important to select the most valuable features in order to derease the computational load of the classification/identification process. These features are shown in Table 2.…”