The Effects of Particulate Matter Alert on Urban Park Visitation in Seoul, Korea: Using Segmented Regression.

To reduce the health burden from particulate matter (PM), the Korean government implemented a nationwide PM 10 (particles less than 10 µg/m 3 in diameter) alert system in 2015. The policy was intended to reduce PM exposure by advising people to refrain from outdoor activities on highly polluted days. The present study aimed to estimate the effect of the PM 10 alert system on people's daily outdoor activity patterns using urban park (specifically, Children's Grand Park) visitation data from Seoul, South Korea, from 2014-2019. Segmented regression was fitted to estimate whether the number of visitors to the park decreased on the days with PM 10 alerts. PM 10 concentration of 80 µg/m 3 , the cut-off point for a "Bad" alert, was set as a threshold, and discontinuity at the threshold and change in the relative risk after the threshold was tested. Time series regression was used to estimate the dose-response line between the ambient PM 10 concentration and the daily number of park visitors. The number of park visitors decreased by 11.8% (relative risk: 0.881, 95% confidence interval: 0.808, 0.960) when a "Bad" alert was issued (PM 10 level above 80 µg/m 3 ) compared to when the alert level was "Normal" (PM 10 level less than 80 µg/m 3 ). The present study found evidence that the PM 10 alert influenced people's daily outdoor activities in Seoul, Korea. As the main purpose of the PM alert is to encourage people to refrain from outdoor activities, evaluating the relationship between PM alerts and behavior patterns can help to grasp the effectiveness of the policy. Further efforts should be made to investigate whether the observed behavioral change leads to reductions in health outcomes caused by PM.