The Urban Polution Monitoring Project


Current pollution monitoring systems use small numbers of fixed sensors that report multiple types of pollutant. We have built a number of mobile sensing systems that give a broader and denser picture of how pollution affects urban spaces and the people within them

 
 

In their everyday life, people in the UK are rarely concerned with the effects of urban pollution. However, high concentrations or extended periods of exposure can have serious health effects. Many local councils monitor urban pollution to determine, for example, the effects of local traffic congestion. However this is usually done from a small number of sparsely distributed, though highly accurate, pollution sensors. For the average person travelling through the city, such numbers are difficult to interpret because a single value might represent a sample several square kilometres away. However it is known that pollution levels can vary, sometimes dramatically, on a per street basis.

The pollutant we are studying is carbon monoxide (CO). Transport makes the greatest contribution to CO levels and CO affects urban areas more significantly than rural areas. Carbon monoxide levels vary greatly depending on local configuration of buildings. We have built a set of mobile sensors that can be carried by hand or placed on a bike rack. They incorporate GPS receivers, so we can get location information for each CO reading.

 

Third version of the mobile pollution-logging device
Third version of the mobile pollution-logging device

Because of the way pollution varies, we have visualised pollution data within 3D models of the city. These models are created from Ordnance Survey maps, aerial photograph and building heights calculated from LIDAR (LIght Detection And Ranging) data. The software has been written as a web service to allow 3D visualisations of the required area to be generated on demand. The result is a vrml model suitable for display by any browser that is vrml 2.0 compliant.

3D model of St Paul's area of central London
View of 3D model of St Paul's area of central London


The pollution data is plotted by dividing the area into a grid and colouring each square according to its mean value. By using the variance and mean of each grid square we can display areas that are likely to have high, low or variable levels of CO. In the following figure, the highest reading (bright red) is 4.8ppm, which is seen to drop to a minimum of 0.5ppm (blue area). The Air Quality Strategy for England, Scotland, Wales and Northern Ireland (http://www.defra.gov.uk/en vironment/consult/airquality/pdf/airstrat.pdf) recommends a standard that exposure to CO not exceed 10ppm (11.6mg/m3) running 8-hour mean.


Pollution level around Dorset Square, adjacent to Marylebone Road
View of the pollution level around Dorset Square, adjacent to Marylebone Road

To date, the sensors have been used successfully in three separate studies: in the Marylebone Road area of London while participating in the DAPPLE project, while cycling to work every day from Camden to UCL and also in the Clerkenwell area.

The CO sensors used while cycling The CO sensors used while walking
Cycling across Marylebone Road Walking along Marylebone Road
   
Marylebone Road area showing all the data collected over a two week period 2D viewing tool showing data from Clerkenwell Road
3D view of Marylebone Road 2D viewer showing data from Clerkenwell Road
Account

Request new password.


Search for People, Publications, or Pages