AI and satellite data used to build British air pollution map

Researchers from London School of Hygiene & Tropical Medicine have led a study using AI and remote-sensing satellite technologies to produce the most detailed nationwide map of air pollution.

 The novel method offers an impressive level of detail, with daily measurements provided in a 1x1km grid across Britain. This could transform understanding of the health impacts of air pollution, by linking detailed exposure maps with health data.

Currently, scientists rely on ground-based monitors to measure air pollution. However, these monitors are sparsely located, mostly deployed in urban areas, and do not always take measurements continuously. This means that there are no nationwide air-pollution records accurate enough to be used by epidemiologists to evaluate health risks.

In this study, researchers combined readings from existing ground-based monitors and data from Earth-observation satellite instruments (which provide data on weather patterns, atmospheric aerosols, land use, and vegetation cover), and incorporated some data from other sources, such as population density, road density, and airport locations.

Using machine-learning algorithms, they combined these datasets to produce estimates of the ground-level concentration of fine particulate matter (PM2.5): one of the most dangerous air pollutants.

They divided Britain into a grid, and derived daily pollution series in the period 2008-2018. These results were cross-validated by comparing the estimates produced by the model to measurements taken from some ground-based monitors, and were found to be closely aligned.

“This research uses the power of artificial intelligence to advance environmental modelling and address public health challenges,” said Dr Rochelle Schneider, first author of the Remote Sensing study. “This impressive air pollution dataset represents PM2.5 records for 4,018 days in a spatial domain of 234,429 grid cells.”

“This provides a remarkable total of 950 million data points that comprehensively quantify the level of air pollution across the whole of Great Britain in an 11-year period.”

The results indicate that the south-east of England is the most polluted region, and hot spots were identified in urban and industrial areas. The results also show an overall decline in air pollution in Britain during the past decade.

“This study demonstrates how cutting-edge techniques based on artificial intelligence and satellite technologies can benefit public health research,” said Professor Antonio Gasparrini, senior author of the study. “The output reveals the shifting patterns of air pollution across Great Britain and in time with extraordinary detail.

“We now hope to use this information to better understand how pollution is affecting the nation’s health, so we can take steps to minimise the risk. The vast amount of data produced will provide a vital tool for public health researchers investigating the effects of air pollution.”

Next, the team will combine the data with local health records. This linked information will be used in epidemiological analyses to reveal a granular picture of the association between air pollution and health outcomes across the country.

The WHO estimates that there are approximately seven million deaths each year worldwide which can be attributed to air pollution. Air pollution is linked to an extremely wide range of health conditions, including lung disease, lung cancer, heart disease, strokes, and psychological conditions.