TY - JOUR
T1 - Characterizing the Spatial Structure(s) of Cities “on the fly”
T2 - the Space-Time Calendar
AU - Tranos, Emmanouil
AU - Arribas-Bel, Daniel
PY - 2017/10/6
Y1 - 2017/10/6
N2 - Our understanding of the spatial structure of cities has been traditionally shaped by the availability of static urban data. In the last few years, thanks to improvements in geospatial technology as well as computing storage and power, there has been an explosion of geo-referenced data, which monitor cities and urban activities in real time. Although this shift in the data landscape promises to change and augment the way we measure, understand, and act on cities, it poses significant methodological challenges and uncovers substantial gaps in the analytics required to leverage its power. In this paper, we propose a novel approach – the Space-Time LISA Calendar – that improves our understanding of both fast and slow spatial dynamics of urban activity. Combining explicitly space-time analytic techniques with a visualization approach that efficiently displays statistical output, our results allow to characterize the nature of the different areas that make up a city and to spot trends and changes, potentially as they occur in the real world. We illustrate the advantages of the Space-Time LISA Calendar using a dataset derived over two years from cell-phone activity in the city of Amsterdam (The Netherlands). Our findings highlight its power to uncover not only the expected patterns, but also additional unexpected ones that would have remained hidden under a more traditional approach.
AB - Our understanding of the spatial structure of cities has been traditionally shaped by the availability of static urban data. In the last few years, thanks to improvements in geospatial technology as well as computing storage and power, there has been an explosion of geo-referenced data, which monitor cities and urban activities in real time. Although this shift in the data landscape promises to change and augment the way we measure, understand, and act on cities, it poses significant methodological challenges and uncovers substantial gaps in the analytics required to leverage its power. In this paper, we propose a novel approach – the Space-Time LISA Calendar – that improves our understanding of both fast and slow spatial dynamics of urban activity. Combining explicitly space-time analytic techniques with a visualization approach that efficiently displays statistical output, our results allow to characterize the nature of the different areas that make up a city and to spot trends and changes, potentially as they occur in the real world. We illustrate the advantages of the Space-Time LISA Calendar using a dataset derived over two years from cell-phone activity in the city of Amsterdam (The Netherlands). Our findings highlight its power to uncover not only the expected patterns, but also additional unexpected ones that would have remained hidden under a more traditional approach.
U2 - 10.1111/gean.12137
DO - 10.1111/gean.12137
M3 - Article
SN - 0016-7363
JO - Geographical Analysis
JF - Geographical Analysis
ER -