|M.Sc Student||Shashank Judah|
|Subject||Utilization of Distributed Wireless Sensor Network for|
Assessing Exposure to Air Pollutants
|Department||Department of Civil and Environmental Engineering||Supervisors||Professor David Broday|
|Dr. Ilan Levy|
|Full Thesis text|
Estimating human exposure for traffic related air pollutants is based today on data measured at air quality monitoring (AQM) stations. This research is aimed at investigating the feasibility of using Wireless Distributed Sensor Network (WDSN) nodes to measure exposure to common traffic-related urban air pollutants. WDSN can provide measurement with higher spatiotemporal resolution compared to AQM stations with reasonable accuracy.
The preparatory phase of the research dealt with study of off-the-shelf (Libelium Waspmotes) and self-developed (EnvSens) WDSN nodes into which mini sensors (ECS and MOS), available in market and commonly used for industrial uses, were integrated. Next, calibration of the WDSN nodes in the laboratory for different pollutants was performed. Measurement error of nodes, effect of temperature and relative humidity (RH), and drift of the sensors’ readings were also studied in the laboratory. The measurement error was found in the range 5-10% whereas drift in the sensors’ readings after 3 months was 7%. Variations in temperature and RH were found to affect the WDSN nodes’ readings considerably.
WDSN nodes were further evaluated against AQM stations. Median values of WDSN NO2 and O3 varied by 15% and 45%, respectively, from the corresponding AQM data. Next, an experiment to study the spatial variation of pollutants’ concentrations by proximate units was performed. The nodes’ measurements revealed that measuring ambient NO2 and CO concentrations were feasible, and that the sensors’ sensitivity is sufficient for spatial analysis of atmospheric processes at the local scale. Following these results, a neighborhood scale measurement campaign was conducted, to study the variation of pollutants’ concentrations across a larger spatial scale as well as between mobile and stationary measurement modes. Results revealed that NO2 and CO concentrations were higher along the busy parts of the Atzmaut street compared to the nearby streets, whereas O3 concentrations were higher on the nearby streets and lower on the Atzmaut street. This reflects to some extent the traffic congestion conditions during the measurement period. Also, the effect of mobility on the nodes’ readings was found to be small.
In conclusion, this study suggests that there is a strong agreement among all nodes for all measured pollutant, and that their measurements are in most cases comparable to AQM reported data. The nodes are also capable of sensing the spatiotemporal variability of ambient concentrations of gaseous pollutants. Hence, they show potential to be used to assess exposure to traffic related gaseous pollutants.