Identifying the key factors of indoor air quality in UK classrooms: A data-driven study

Indoor air quality in schools, and classrooms, is of critical importance for the health and well-being of pupils and staff. The recent COVID-19 pandemic highlighted the essential role that ventilation systems play in limiting the spread of airborne diseases indoors but the quality of the air we breathe indoors also impacts our health and wellbeing more widely.

This is particularly important in classrooms, where children, who are particularly vulnerable to pollution, spend a significant proportion of their time indoors and where a lack of appropriate ventilation can directly impact their learning and health.

Our research has shown that the levels of ventilation and pollution can vary greatly between schools, as well as throughout the year for a given school. This is thought to be due to factors such as the outdoor conditions, the school's building type and the school's ventilation system. However, when monitoring classrooms within a single school, significant differences in carbon dioxide concentrations have also been observed.

This points out that other factors, such as human behaviour, must also be consequential. In schools, the occupants can influence the air quality around them by, for example, carrying out different teaching activities or having different ventilation behaviours. This increases the complexity in understanding which factors are likely to have the most significant impact and makes the prioritisation of solutions to improve air quality difficult.

The introduction of affordable air quality sensors in schools presents an opportunity to address this challenge. These sensors allow the collection of indoor air quality data over long periods, at a potentially large scale and without overly disturbing the normal running of a classroom. In addition, these growing air quality monitoring data in classrooms can be linked with other datasets providing further contextual information on the schools' environment, including measurements of the outdoor air quality, nearby traffic data or details of the school's building stock.

This project will take advantage of this range of datasets to identify the key factors that impact the quality of the air inside classrooms. Although we know that the outdoor air, the building itself, the ventilation provision and the occupants can all impact indoor air, it is not apparent which of these matters the most and should be prioritised to improve the indoor air quality.

Both conventional statistical methods and machine learning techniques will be used to characterise which of these factors are critical in determining whether a given classroom is likely to have air quality issues, and the performance of the different approaches will be compared. This will allow us to identify the main blockers to good indoor air quality in classrooms as well as to determine what solutions or interventions should be prioritised.

This project will also allow us to estimate the size of the public health problem caused by inadequate air quality in schools given the current school building stock in the UK. This project presents an opportunity to use new data to further our understanding of indoor air quality in classrooms and identify how it can be improved to promote the health and learning outcomes of pupils and teachers.