Air pollution – the combination of particulate matter and ozone – is a major environmental health problem and a leading risk factor for death globally. Nitrogen dioxide (NO2) constitutes a major component of particulate matter. As such, air pollution is often measured by mean concentration of NO2 particles per cubic meter of air volume (μg/m3).1
This air pollutant can cause stroke, heart disease, lung cancer, and respiratory diseases. For instance, it can lead to significant inflammation of the airways, which adversely affects the respiratory system and exacerbates asthma. Therefore, the health burden of air pollution can be measured by the number of asthma-related hospitalisations in an area.2
In 2017, air pollution caused 4.9 million deaths globally, making it the fourth highest risk factor of death, after high blood pressure, smoking, and high blood sugar. In other words, nearly 1-in-10 of all deaths around the world were directly related to air pollution.3
Combustion processes, from heating, power generation, and engines in vehicles and ships, are the primary source of air pollution. As such, health risks due to air pollution are particularly prevalent in low- and middle-income countries, because of industrialisation and the use of solid fuels for cooking. Geographical factors, such as dry conditions, sand and dust sources, exacerbate this problem.4
This income-based disparity in air pollution-related health problems is reflected locally within cities, as low-income areas tend to have more air pollution, as well as greater health risks associated with air pollution. One reason is that low-income neighbourhoods are often located near polluting disamenities like major roads and power plants.5, 6
The following maps visualise the data on air pollution, measured by NO2 μg/m3, and its associated health risks, measured by the number of asthma-related hospitalisations, in London and New York City, by borough/neighbourhood. It is critical to note that these maps are not intended to show causation; rather, they simply show the spatial distribution of the problem, from which correlation can be inferred.