Evaluating local-area transport interventions through mobile phone GPS data: a framework and case study of low traffic neighbourhoods
提出利用手机GPS数据评估地方交通干预措施的新框架，并以伦敦低交通街区为例，发现居民减少驾车、增加步行，同时非居民访客增多，但部分交通被转移到周边住宅区。
There is increasing adoption of local-area transport interventions worldwide, as part of a strategy to encourage more sustainable mobility. However, the impacts of such interventions are complex and wide-ranging and often difficult to monitor using conventional instruments, as they include various travel modes (e.g. car, bus, cycle, and walk), different population groups (e.g. residents, visitors, and non-local travellers), and multiple scales of influence (both social and spatial). This paper presents a new comprehensive evaluation framework to assess local-area scheme interventions that leverages the continuous, high-granularity spatial coverage of mobile phone GPS data to overcome these limitations. This framework is then applied to examine the impact of a set of Low Traffic Neighbourhood (LTN) Schemes in North-East London, in which through traffic is removed from predominantly residential areas, using road closures and traffic filters. Using Propensity Score Matching and a Difference-in-Differences model, it explores issues related to changes in mobility and activity patterns, at both individual and spatial levels. The findings reveal that, in this sample of LTNs, residents have reduced their driving outside LTNs and spend more time walking and engaging in other activities within their LTNs. It also enhances local attractiveness, as evidenced by increased footfall among non-residents during weekdays. Moreover, in this case the LTNs have reduced motor traffic flow on both internal and surrounding roads, challenging the notion that traffic is merely displaced to closely adjacent main roads. However, the LTNs have resulted in former pass-through drivers taking longer routes, potentially transferring motor traffic and increasing road usage in surrounding residential areas. This study demonstrates the value of using mobile phone GPS data and causal inference methods to provide a broader and richer evaluation of local-area transport interventions.