For a month now, we’ve been asked to stay at home to stop the spread of Covid-19. New data visualisations now show how well we’re following that instruction.
The Science Media Centre has funded data visualisations showing when and where we’ve travelled over the past month. Geographer Chris McDowall has made videos following population movements in Auckland, Christchurch, and Wellington since the beginning of lockdown.
Data Ventures, the commercial arm of Statistics New Zealand, has been collating anonymised information from cellphone companies on hour-by-hour movements of our population. McDowall’s videos show the slowing of travel in the lead-up to lockdown, and how our cities are now virtually stagnant save for the pulse of essential workers continuing their day-to-day movements.
“Every shape represents a Statistics New Zealand area, which in urban areas roughly correspond to suburbs,” said McDowall. “Data Ventures collates anonymised information from cellphone towers and uses this information to approximate how many people are in each area in a given hour. This data is a window in travel as people (and their phones) move between suburbs.”
The visualisations also use population estimates and modelling, to account for those without cellphones. Encouragingly, Data Venture’s numbers line up with Google and Apple movement data. “The difference between cellphones and Google mobility data is that for Google you need to turn on location data, but cellphone pinging is universal,” said Robert Chiu, head of data at Data Ventures. “Cellphones cover about 82% of the population.”
The maps offer a narrative account of the data that helps both the public and scientists understand how our movements change under the alert levels. “The cyclic ‘breathing’ effect that you see near the start of these videos is the workforce leaving the suburbs and commuting to commercial and industrial hubs,” said McDowall.
“As we move through alert levels two, three and four the activity slows down. Over Easter, there is minimal movement as the population largely stays in one place. There is still some activity in major commercial spaces, but most places enter a jittery stasis.” Data Venture’s latest report says 90% of us have stayed at home over the past 30 days.
Most of the areas that remain red throughout lockdown are commercial spaces where essential services such as supermarket shopping and doctor’s visits are conducted. Data Venture’s numbers show a 70% drop in retail activity over the past 30 days, including over weekends; days that were once meant for shopping and recreation are now as meaningless as the rest of them.
Looking at these maps, it begs the question: can we use these methods for contact tracing? Fortunately for your privacy, cellphone pinging can pinpoint people to a suburb but nowhere more specific. For example, it’s possible to tell that someone is in Thorndon, but not if they’re in the National Library. However, tracking population movement at this level is useful for scientists predicting how a pandemic might spread.
Professor Malcolm Campbell, deputy director of GeoHealth at the University of Canterbury, made it clear that the public’s privacy is something he takes seriously. “This isn’t individuals’ [data],” he said. “The level of detail is such that I wouldn’t really have concerns about privacy.” He said that his own data — anonymised like everyone else’s — is in the visualisation, which he’s comfortable with. “I’m happy it’s in this data set and I’m remarkably relaxed about that kind of concern.”
McDowall said the maps are impressionistic. “It’s giving the vibe of the thing rather than the percentile. It’s a broader sweep, but I think that there is a value in that.”