重庆市新型冠状病毒肺炎流行时空特征及其与人群活动性的关系

本文收集了重庆市2020年1月21日—2月24日确诊的545例新型冠状病毒肺炎（COVID-19）病例个案信息,结合1500万手机用户在疫情期间的信令轨迹大数据,分析了其疫情的时空演化特征以及人群活动的变化规律,并从复杂网络的角度揭示了疫情分布的异质性,从疫情传播与人群活动之间的关系揭示了异质性的原因。研究发现：① 重庆市疫情在时间上,经历了以输入病例为主、输入和本地病例共存、以本地病例为主3个阶段,病例实时再生数（R_t）初期较高,随着防控措施的实施,逐步减小;空间上,病例分布呈现显著聚集性,病例高聚集区主要分布在以万州区为核心的渝东北地区和以主城区为核心的渝西南地区;② 疫情发生后,重庆市人群移动总量减少为疫情前的53.20%,减少主要集中在主城区以及其他各区县的中心城区,而郊区、农村的人群移动变化不大,甚至有所增加;③ 人群活动与病例发生之间存在不同程度的相关性,具体为：每日人群移动总量与病例实时再生数、一个平均潜伏期（7 d）后的每日新增病例数的相关系数为0.97、0.89,揭示了人群活动与病例增长的时间相关性;各街道（乡镇）人群移动总量与其累计确诊病例数、本地感染病例数之间的相关系数为0.40、0.35,揭示了人群活动与病例空间分布的相关性;病例高聚集区与人群移动网络社区对应,且与网络社区内人群活动较强的区域吻合,揭示了重庆市疫情传播的本地聚集特征。大数据与疫情信息的聚合分析证实,切断人群移动网络社区之间的连接,并遏制疫情高风险社区内部的传播是在城市内部疫情防控的有效措施。

Spatial-temporal Characteristics of COVID-19 in Chongqing and Its Relationship with Human Mobility

Based on the epidemiological investigation data of 545 COVID-19 cases and mobile phone trajectory data of 15 million users during the epidemic( from 21 January, 2020 to 24 February, 2020), this paper analyzed the spatial-temporal characteristics of COVID-19 and the human mobility changes in Chongqing. Furthermore,the correlation relationship between them was explored to explain these characteristics and changes. The results show that:(1) The epidemic pattern in Chongqing can be divided into three stages( i.e. imported cases stage,imported cases plus local cases stage, and local cases stage) and the real time reproduction number(Rt) was high at early stage, but declined significantly after prevention and control measures were taken;The spatial distribution of cases presented a significant clustering, and the high clustering areas were mainly distributed in the northeastern and the southwestern of Chongqing;(2) After the epidemic, the total amount of human mobility decreased to 53.20% and the decrease was mainly concentrated in the main urban area, while that of in the suburbs and rural areas did not change, or even increased;(3) The relationship between human mobility and case occurrence lies in two aspects: The correlation coefficient between daily human mobility and Rt, daily increased number of cases after an average incubation period(7 d) were 0.98, 0.87, revealing the time correlation between human mobility and case growth;The correlation coefficient between total amount of human mobility and total number of cases, number of local cases in each street(township) were 0.40, 0.35, revealing the correlation between human mobility and spatial distribution of cases. The cases clustering area corresponds to the network community of human mobility, revealing the local clustering transmission is the major transmission model. By aggregating the big data and the epidemic data, we suggests that cutting off the connection between different human mobility network communities and blocking the local transmission inside the high risk communities is an effective measure for the prevention and control of epidemics in cities.