Low-cost technologies for high-quality air pollution monitoring

Outdoor air pollution causes 4.2 million premature deaths globally each year. Air pollution in most parts of the world is currently assessed using expensive ground-based sensors. This approach has significant limitations.

Regulatory monitoring stations are costly to operate and maintain, posing challenges for widespread implementation. As a result, these ground-based sites often have poor spatial coverage, and cities suffer from a scarcity of air quality information. Such shortcomings have led to substandard monitoring in most urban landscapes, where air pollution routinely exceeds safe levels for human health.

An alternative, low-cost sensors (LCS), are much cheaper and easier to install compared to government regulation-grade stations. So far, most research has focused on using LCS to measure pollution levels, few have utilised it to find information on the sources and relative magnitude of this pollution. This information is vital to plan and enact cost-effective control measures.

Another challenge is that LCS monitoring is typically short-term, inconsistent, and has shown to omit important sources of PM with smaller particle sizes. To assist in data quality and representativity, aerosol optical depth from satellite instruments can be used to measure PM concentration. However, the resolution of observations is courser, and the frequency of measurements are lower, meaning diurnal variations are difficult to quantify.

An alternative is the use of models such as WRF-Chem to simulate the transport, transformation, and emission of various air pollutants. But on its own, such models tend to struggle in capturing the complex atmospheric conditions and pollution sources associated with urban areas.

Research in integrating LCS with other sources of information (satellite observations and air quality models) is encouraging but in its early stages with very few studies. This project will have two main objectives, firstly to assess the capabilities of low-cost sensors for source apportionment within complex real-world environments, and secondly, to investigate how LCS can be integrated with other information channels to improve the spatiotemporal resolution and accuracy of air quality estimates.

The resulting output of this PhD project will be valuable for a more robust assessment of PM's health impacts, and informing policy action.

In terms of methodology, low-cost sensors will be used to quantify the origins of pollution at sites that are affected by emissions from different sources. The data collected from instruments will be analysed using statistical and machine learning techniques to assess the impact of specific pollution sources at different sites. To ensure the reliability of findings, results will be compared against regulatory-grade equipment.

Using novel analysis, the information collected from the low-cost sensors will be integrated with satellite and model data. This will be used to upgrade the quality of measured data and fill any gaps, or resolve any irregularities, both spatially and temporally. Ultimately, the objective of this co-dependency is to reduce the reliance on resource-heavy regulatory instruments to validate measurements.

There are field locations in both the UK, Africa, and East Africa through partnerships with industrial, governmental and UN organisations. The use of arrays of sensors across urban areas in combination with the developed techniques will allow for source triangulation of the location of emitters. These locations will be used to evaluate the effectiveness of low-cost techniques.