Thoughts on Geovisualization of Human Activity… (Kwan 2004)

The immediate discussion of the historical antecedents for temporal GIS by Swedish geographers uses the 24-hour day as a “sequence of temporal events” but I wonder why this unit of measurement was chosen as opposed to 48-hours or a week to illustrate the periodicity of temporal events, which may not be captured at the daily scale. It is interesting to note the gendered differences that are made visible by studies of women’s and mens spatio-temporal activities. As the authors note, “This perspective has been particularly fruitful for understanding women’s everyday lives because it helps to identify the restrictive effect of space-time constraints on their activity choice….” I am curious about how much additional data researchers must collect to formulate hypotheses about why women follow certain paths to work or are typically present at certain locations at certain times. I am also curious about how this process is different when trying to explain the spatiotemporal patterns observed in men’s travel behaviour.

One of the primary challenges identified by the authors is the lack of fine-grain individual data relating to peoples’ mobility in urban environments, such as in transportation systems or their daily commutes. This paper was written in 2004 and now, with the rapid increase in streaming, GPS from mobile devices, and open big data sets for most large cities, this is less of a concern. The big challenge these days is probably in parsing the sheer quantity of data with appropriate tools and hypotheses to identify key trends and gain usable insights about resident’s travel behaviour.

The methodology used by the researchers for their study of Portland relied on self-reported behaviour in the form of  a two-day travel study. There are many reasons why the reported data might be unreliable or unusable, especially given the fallibility  of time estimation and tendency to under or over report travel times based on mode of transport, mood, memory of the event, etc. That being said, this is probably the most ethical mode of data collection and asks for explicit consent. I would be interested to know how the researchers cross referenced the survey data with their information about the Portland Metropolitan Region, as well as the structure of the survey.




Comments are closed.