Understanding Urban Movements through Big Data and Social Simulation

This research will fundamentally alter our understanding of daily urban movement patterns through a combination of 'big data' analysis and cutting-edge computer simulation. It will develop new methods to produce data that will help us to address key issues in crime and health.

A big data "revolution" is underway that has the potential to transform our understanding of daily urban dynamics and could have big impacts on the ways that scientists conduct social science research. Vast quantities of new data are being gathered about people in cities. New services are capturing information about peoples' daily actions from their use of social media, public transport systems and mobile telephones, to name a few. Data from these sources, although noisy, messy and biased are unprecedented in their scope, scale and resolution. This research will first develop new geographical methods that can make sense of these data and derive information about peoples' daily movements in space and time. It then proposes to develop a computer simulation of city-wide daily urban movements that will be calibrated automatically from streams of crowd-sourced data.

This research is important because previous projects that have attempted to model detailed urban movements have been hampered by a lack of high-resolution data and by methods that have difficulty in modelling the complex individual-level interactions of people that ultimately characterise cities. Large-volume sources, such as censuses, capture attributes and characteristics of the population, rather than their attitudes and behaviours. On the other hand, detailed surveys that attempt to capture this behavioural information are naturally limited by their size and scope. In contrast, new 'big' public data streams are voluminous and contain information about a user's location as well as a textual or multimedia component that often describes their behaviour or actions. The new simulation model will make use of these data to create a much more accurate picture of urban dynamics than we have had before now.

This new picture will have the capacity to alter our understanding of key social phenomena that depend on where people are at different times of day, rather than simply where they live. It will use the simulation outputs to generate new estimates of where people are and apply these estimates to two empirical areas:

1. Crime. The research will re-analyse crime rates based on estimates of where groups of potential victims are, rather than simply where people live. This will then show us where crime is higher or lower than expected, given the number of people who are in the area at the time and might be victimised. This will have obvious impacts for crime reduction policies and the project will work with the police and crime-reduction experts to make the best use of the results.

2. Health. The second project will calculate peoples' exposure to air pollution based on where they actually spend their time, rather than where they live. Normally, peoples' home location is used to estimate how susceptible they are to air pollution, but this ignores the fact that many people will be exposed whilst away from home (e.g. going to work, travelling to the shops, etc.). By more accurately estimating peoples' exposure, this project could have substantial impacts on EU/UK air quality laws and lead to an overall improvement in national health.

In summary, this project will make use of new 'big' data and advanced computer simulation to better understand how people move around cities. It will then apply this new knowledge to try to better understand rates of crime and to assess the impacts of air pollution on peoples' health.

The spatio-temporal dynamics of urban movement patterns are poorly understood at high resolutions. This strongly disadvantages the organisations responsible for managing public resources that rely on accurate estimates of daily population density. This research will utilise new forms of data, coupled with advanced simulation modelling, to create much a more accurate picture of urban dynamics. Both the ESRC and the Government have recognised the importance of engaging with this `big data' paradigm, making this research highly topical. Beneficiaries include:

West Yorkshire Police;
Safer Leeds;
Jill Dando Institute of Crime Science;
Metropolitan Police;
Leeds Data Thing;
Department of Health Committee on the Medical Effects of Air Pollutants (COMEAP);
World Health Organisation;
Environmental Protection UK (EPUK).

Initially, local organisations will receive the greatest impact through direct engagement (for details see Pathways to Impact). West Yorkshire Police will receive benefit as new crime rate estimates will have a impact on their crime reduction initiatives. For the first time measures of crime can be estimated much more accurately through knowledge about the underlying population of potential victims which will indicate the places (and times) where crime is significantly higher than would otherwise be expected and, importantly, where it is significantly lower. Also, there are a number of non-profit organisations interested in how big data research will shape society in the future. One group in particular, Leeds Data Thing, hold regular events to discuss new data and will undoubtedly benefit from the methodological outcomes of the research.

In the longer term, the project will have broader impacts to the Leeds Community Safety Partnership (Safer Leeds) and their partner organisations. Safer Leeds have a broad agenda for improving community safety through the coordination of numerous bodies. Following preliminary engagement meetings the organisation is excited about the possibilities of applying the research outputs to application areas such as the identification of bicycle accident hotspots, predicting graffiti and litter locations and optimising the standing locations of accident & emergency vehicles. Understanding urban dynamics will have significant impact on each of these areas and has the potential to substantially improve resources allocation. These are ideal areas for collaborative PhD studentships that the PI will attempt to attract funding for. There are also national crime research organisations who will receive indirect impact through this research. The Jill Dando Institute for Crime Science will benefit from the new approach, developed by this project, for measuring the population at risk of crime. New methods that can more accurately take account of the spatio-temporal behaviour of potential victims will undoubtedly influence the research and the crime science classes that they run.

The exploratory project that will apply the research outputs to the study of health and air quality will also have national and international impact in the longer term. New estimates of exposure to polluted air, based on mobile rather than residential population estimates, will be invaluable to COMEAP (Department of Health), the World Health Organisation Global Health Estimates and the charity EPUK . These organisations assess the impact of air quality on health and, in particular, estimate the number of yearly deaths as a result of exposure to polluted air. Estimates based on the residential (night time) population will substantially under-estimate personal exposure and downplay the affects of air pollution on health. Hence the results of this study could fundamentally influence the ways that these organisations assess exposure and quantify the impacts of air pollution. Ultimately, this has the potential to result in more stringent air quality targets at an EU level.