When governments first scrambled to control the spread of COVID-19 in early 2020, they faced a crisis with very little precedent. To protect their populations, many governments tried containment measures like lockdowns despite limited data about whether such measures would be effective against SARS-CoV-2. One year on, insights from spatial and temporal data are now allowing policymakers to assess the effectiveness of lockdown measures while making evidence-based forecasts for possible future outbreaks.
In Singapore’s case, one might ask: how can we measure the impact of lockdowns like the government’s ‘circuit breaker’ policy? This challenge was taken on by Xiuju Fu, a Senior Research Scientist at A*STAR's Institute of High Performance Computing (IHPC). With collaborators from Delft University, Brno University of Technology and Singapore’s Ministry of Health, Fu and her team looked at how mobility-curbing measures affected variables like COVID-19 exposure risk and air quality.
However, collecting spatio-temporal mobility data about walking, driving or public transport is difficult, so the researchers turned to a near real-time, open-source dataset—the availability of car-park lots—as a proxy for mobility. “Compared to other mobility data, driving mobility data from residential car-park records is readily available, accurate and continuously updated, which saves us time-consuming data collection and validation,” Fu explained.
Using an extended Bayesian spatial-temporal model that considered data uncertainties, Fu’s team found that the ‘circuit breaker’ immediately reduced driving mobility by 13.4 percent, reaching a peak reduction of 36.4 percent by April 12. This reduction in mobility was associated with an average reduction of potential COVID-19 exposure risk by 37.6 percent and a 55.4 percent average reduction in transportation-related emissions.
These changes happened after a six-day post-lockdown ‘lag effect,' where a preceding phase with near-exponential rates of increase in COVID-19 cases steadily slowed down and eventually made a turnaround. “Our results indicate that the circuit breaker not only keeps residents safe but also leads to environmental changes with reduced transportation pollution,” Fu said.
This information is useful for policymakers because it not only shows that mitigation measures can control the explosion of COVID-19 cases, but also when and how this might happen. The methods used in the study can help inform governments and citizens alike of the appropriate kinds of mitigation strategies and when to use them. “For example, residents working from home can even acquire more online information through predictive analytics on crowd distribution and select the most suitable time to go out to minimize transmission risk,” Fu said.
A*STAR researchers contributing to the study are from A*STAR’s Institute of High Performance Computing (IHPC).
