Air pollution is now recognized as the world’s single largest environmental and human health threat. Indeed, a large number of environmental epidemiological studies have quantified the health impacts of population exposure to pollution. In previous studies, exposure estimates at the population level have not considered spatially- and temporally varying populations present in study regions. Therefore, in the first study of it is kind, we use measured population activity patterns representing several million people to evaluate population-weighted exposure to air pollution on a city-wide scale. Mobile and wireless devices yield information about where and when people are present, thus collective activity patterns were determined using counts of connections to the cellular network. Population-weighted exposure to PM_2.5 in New York City (NYC), herein termed “Active Population Exposure” was evaluated using population activity patterns and spatiotemporal PM_2.5 concentration levels, and compared to “Home Population Exposure”, which assumed a static population distribution as per Census data. Areas of relatively higher population-weighted exposures were concentrated in different districts within NYC in both scenarios. These were more centralized for the “Active Population Exposure” scenario. Population-weighted exposure computed in each district of NYC for the “Active” scenario were found to be statistically significantly (p < 0.05) different to the “Home” scenario for most districts. In investigating the temporal variability of the “Active” population-weighted exposures determined in districts, these were found to be significantly different (p < 0.05) during the daytime and the nighttime. Evaluating population exposure to air pollution using spatiotemporal population mobility patterns warrants consideration in future environmental epidemiological studies linking air quality and human health.

The coordination of humanitarian relief, e.g. in a natural disaster or a conflict situation, is often complicated by a scarcity of data to inform planning. Remote sensing imagery, from satellites or drones, can give important insights into conditions on the ground, including in areas which are difficult to access. Applications include situation awareness after natural disasters, structural damage assessment in conflict, monitoring human rights violations or population estimation in settlements. We review machine learning approaches for automating these problems, and discuss their potential and limitations. We also provide a case study of experiments using deep learning methods to count the numbers of structures in multiple refugee settlements in Africa and the Middle East. We find that while high levels of accuracy are possible, there is considerable variation in the characteristics of imagery collected from different sensors and regions. In this, as in the other applications discussed in the paper, critical inferences must be made from a relatively small amount of pixel data. We, therefore, consider that using machine learning systems as an augmentation of human analysts is a reasonable strategy to transition from current fully manual operational pipelines to ones which are both more efficient and have the necessary levels of quality control. This article is part of a discussion meeting issue ‘The growing ubiquity of algorithms in society: implications, impacts and innovations’.