Mapping and visualizing the association of transportation and air quality in urban neighborhoods

Methods

Summary

A “measurement in a backpack” concept will be implemented in this project. Portable instruments and low cost sensors will be placed in a backpack and carried by students as they go to different places carrying out their daily activities. Noise will be measured by a portable unit with a data logger and USB (Extech). Black carbon (particulate matter surrogate) will be measured by a portable aethelometer (AE51). NO2 will be measured by a low cost sensor assembled on an Arduino board (with temperature, relative humidity) in a small case. A GPS unit will be used to record coordinates. Traffic conditions will be recorded with a small video camera. Air quality at different urban areas (residential or commercial) will be measured,  at different traffic conditions, such as rush hours and non-rush hours, weekdays and weekends, days and nights.  The measured data will be entered into Geographic Information System and visualized through maps, animations and interactive media, and posted on a designated website.

Challenges

A lot of data will be collected from different instruments. How to synchronize and process them can be a challenge. The time of the instruments will be synchronized prior to measurement. Students will be asked to provide a brief log of when and where: their locations, and potential air pollutant sources, to aid with data processing. Sensitive information, such as residential address data, will be screened out prior to data posting. GPS data poses a particular risk due to the ease of identifying where individuals live, work and what activities they participate in. To protect privacy, the data can be masked out by lowering its  precision, e.g. larger grids and longer time intervals. 

Pre Analysis Plan

The hypothesis to be tested: the use of portable and low cost sensors in a high sampling density can better reflect the spatiotemporal patterns of traffic generated pollutants, such as NO₂ and black carbon levels than using data from limited monitoring stations.  The team will evaluate data consistency, and address data variance (with the help of university's IT team) prior to sharing the data.